The focus of both UJ and Capitec was the interpretation rule, however, and in order to appreciate their implications, it is necessary to understand the current approach adopted by our courts to the task of contractual interpretation in general. This approach (applied initially to statutory interpretation, but extended to contracts in subsequent judgments) is summarised in the 2012 decision of the Supreme Court of Appeal in Natal Joint Municipal Pension Fund v Endumeni Municipality (“Endumeni”), which established that the terms of a contract must be interpreted holistically, and (even in the absence of ambiguity) the approach must, from the outset, have regard to the “triad” of text, context and purpose. The approach is objective, rather than subjective, and involves attributing meaning to the words used in a document, rather than a search for the intention of the parties or the legislature. The point of departure is the language of the provision itself (‘in the light of the ordinary rules of grammar and syntax’), but with due regard to the broader context of the provision, which is established by reading the document as a whole, in light of the circumstances attendant upon its coming into existence; the background to its preparation and production; the material known to those responsible for its production; and the apparent purpose to which it is directed. A sensible interpre-tation is to be preferred to one that leads to insensible or unbusinesslike results, or which undermines the apparent purpose of the document, but judges must resist the temptation to substitute what they regard as reasonable, sensible or businesslike for the words actually used: they must not “make the contract” for the parties. 

​

What Endumeni left unclear, however, was what extrinsic evidence of context remained inadmissible in interpreting the terms of the contract. Subsequent decisions of the SCA prior to UJ and Capitec held that (i) evidence of the parties’ intentions and their prior negotiations remained inadmissible, even if the language of the agreement was ambiguous, but (ii) evidence of the conduct of the parties in implementing the agreement is admissible, since their conduct in adopting the same approach provides clear evidence of “how reasonable business people situated as they were and knowing what they knew, would construe the disputed provision”. Notwithstanding these decisions, the SCA remained critical of the tendency of courts to admit too much extrinsic evidence, which, in its view, resulted in the written text of the contract being “relegated” in the process of interpretation. 

​

In UJ, the Concourt had to determine whether the rights of a lessee under a long-term lease were so personal as to render them incapable of cession (a so-called delectus personae), and whether extrinsic evidence was admissible for the purpose of making this determination. The matter involved a lease concluded in 1996, between the University of Johannesburg (UJ) and the Auckland Park Theological Seminary (APT), in terms of which APT leased premises owned by UJ for a period of 30 years, for the purpose of building a theological seminary, against payment of a once-off rental of R700,000. UJ and APT had also concluded a separate cooperation agreement, in terms of which students registered for theological degrees at the seminary would be taught certain courses by UJ and others by APT. In 2011, APT ceded its rights under the lease to a third party for an amount of R6.5 million, to establish a religious-based school for primary and secondary education. Although the lease did not contain a “no cession or assignment” clause, UJ regarded the cession as a repudiation of the lease on the basis that the rights granted were personal to APT, cancelled the lease, and applied for APT to be evicted from the premises.

​

Although a single judge and full bench of the Gauteng High Court ruled in favour of UJ (finding that the lease constituted a delectus personae), the SCA took the view that the rights of a lessee under a long-term lease are by their very nature not personal, and the introduction of extrinsic evidence to show otherwise was thus inadmissible. The Concourt disagreed, finding that the long-established test for a delectus personae is whether the contract is so personal that it can make any reasonable or substantial difference to the other party (in this case, UJ), whether the cedent or the cessionary is entitled to enforce it – this is a matter of contractual interpretation (not an issue to be determined simply by the nature of the right in the abstract), and hence, the Edumeni approach should be followed from the outset. The parties would thus have to adduce extrinsic evidence, which the Concourt held “could include the pre-contractual exchanges between the parties leading up to the conclusion of the contract”. While such evidence should be used ‘as conservatively as possible’, it is essential for the court to consider evidence of context and purpose, and where reasonable people may disagree on admissibility, the court should err on the side of admitting the evidence, which can thereafter be disregarded on the basis of weight – it is important not to conflate admissibility and weight of evidence. Applying this approach to the facts, the Concourt held that various contextual factors (such as the statutory context, the fact that the premises could only be used for tertiary education, the references to the co-operation agreement in the negotiations between the parties) would have shown that the lease was, in fact, personal to APT. The Concourt thus found that UJ had validly cancelled the lease for repudiation. 

​

The Concourt in UJ appears to have found that only the first leg of the parol evidence rule (i.e. the integration rule) survives in our law, by holding that the rule “is concerned with cases where the evidence in question seeks to vary, contradict or add to (as opposed to assist the court to interpret) the terms of the agreement. The decision of the SCA in Capitec supports this, although the SCA also added some important commentary on the extent to which extrinsic evidence should be permitted for purposes of interpreting the terms of the agreement. 

​

The dispute in Capitec involved the interpretation of a subscription agreement concluded between Coral Lagoon and Capitec, in terms of which Coral Lagoon subscribed for 10 million shares in Capitec in order to enable Capitec to fulfil its BEE obligations. Clause 8.3 of the agreement provided that, should Coral dispose of any of its shares in Capitec to any person that does not comply with the BEE Act and Codes, Capitec would be entitled to compel Coral Lagoon to acquire an equal number of Capitec shares in order to restore its black ownership. In implementing this clause subsequent to the conclusion of the agreement, the parties read it to mean that Capitec’s consent was required for any sale by Coral Lagoon of its Capitec shares (although this did not appear from the wording of the clause itself). When Coral Lagoon subsequently undertook to transfer its Capitec shares to the Transnet Second Defined Benefit Pension Fund (‘the Fund’) as part of a settlement agreement, the transfer was thus made subject to Capitec’s consent. When Capitec refused to give its consent, Coral Lagoon approached the High Court for an order compelling Capitec to give such consent on the basis that withholding such consent was unreasonable in the circumstances, and contrary to good faith. 

The High Court granted the order, finding that the way in which the parties had previously interpreted and implemented clause 8.3 meant that ‘justice can only be dispensed if the matter is approached on the basis that Capitec’s consent is required for the sale”, and that Capitec had breached its duty to act in good faith and reasonably by refusing its consent. The decision was heavily criticised by the SCA on appeal, which found that the departure point must remain the words of the clause, to be read with regard to the context and the purpose (i.e. applying the Endumeni test). Importantly, the SCA noted that the Endumeni approach does not ‘license judicial interpretation that imports meanings into a contract so as to make it a better contract, or one that is ethically preferable’. In this case, neither the text nor the context, nor the purpose of clause 8.3 prevented Coral Lagoon from selling its shares without Capitec’s consent, and the SCA thus held unanimously that such consent was not required for the transfer of Coral Lagoon’s shares to the Fund. 

​

In reaching its decision, the SCA provided some useful guidelines on the admissibility of extrinsic evidence for purposes of interpreting a contract, noting that the parol evidence rule is likely to become a residual rule that does little more than to identify the written agreement (its content), the meaning of which must be determined. In allowing extrinsic evidence, however, the courts’ aversion to receiving evidence of prior negotiations and what the parties intended or understood the contract to mean remains an important limitation on what could be said to be relevant to context or purpose, and previous decisions were also not an invitation to “harvest evidence, on an indiscriminate basis, of what the parties did after they concluded their agreement.” Ultimately, interpretation continues to begin with the text and structure of the document, and context is “not a licence to contend for meanings unmoored in the text and its structure”.

​

The lesson for us as commercial lawyers is that the words of the contract continue to matter, and we would be well served to make the context and purpose of those words as clear as possible in the document itself. 

​

Oosthuizen and Hutchison are Executives, Corporate Commercial | ENSafrica.