IN THE SUPREME COURT OF APPEAL

OF SOUTH AFRICA

Case number : 416/99

In the matter between :

THOROUGHBRED BREEDERS’ ASSOCIATION

OF SOUTH AFRICA Appellant

and

PRICE WATERHOUSE Respondent

CORAM : NIENABER, MARAIS, OLIVIER, FARLAM JJA and BRAND AJA

HEARD : 2 - 3 MAY 2001

DELIVERED : 1 JUNE 2001

Auditor - duties of vis-à-vis client - negligent failure by auditor, in auditing the books of a company, to appreciate the significance of (and pursue) certain unusual features and discrepancies in the company’s books of account - had it done so it would have led to the discovery that the company’s financial manager was engaged in large-scale thefts from the company - company suing auditor for damages for breach of contract for the losses it suffered because the thefts continued for a further period after the audit - company itself careless in failing to properly supervise the activities of its financial manager, knowing that he had previously been convicted of and imprisoned for theft - auditor not so informed - causation and remoteness - test for - whether the Apportionment of Damages Act 1956 applies to breach of contract - additional interest payable on increased overdraft - whether claimed as special damages or as mora interest - interest on the amount awarded as damages - from when and on what scale - costs on attorney and own client scale.

JUDGMENT

NIENABER JA/

NIENABER JA :

[1] Introduction

This is a case of a client suing its auditor for damages for breach of contract. The complaint is that the auditor, in the course of a routine annual audit, failed to realise that the client’s own financial manager, a man with a criminal record for theft of which the client but not the auditor was aware, had been systematically stealing from it in the past and, if undetected, would be likely to continue to do so in the future.

[2] The appellant, the plaintiff in the court below, hereinafter referred to simply as “TBA”, is a statutory juristic person registered in terms of s 18 of the Livestock Improvement Act, 25 of 1977. It is a non-profit making association of breeders of thoroughbred horses, as its name implies. Membership of TBA is a precondition for the registration of horses to be entered into the stud book maintained by the Jockey Club of South Africa, the body administering horse racing in this country. TBA has some 700–800 members, scattered throughout the country, who are predominantly farmers. TBA furthermore acts as auctioneer and sales agent for the sale of thoroughbred horses and it derives its income principally from membership fees and from the commissions it earns on auctions and sales.

[3] In 1991 TBA appointed one John Mitchell as its financial manager. During his three months probationary period it was discovered that he had been convicted in 1985 of theft and sentenced to a period of imprisonment of eight years of which he served approximately 18 months. Even though he withheld this information from TBA when applying for the position (and indeed supplied false information to it about his employment during the period of his incarceration) it was nevertheless resolved by TBA’s council (a) to confirm his employment, (b) to monitor his activities in future, (c) not to disclose his past history to other members of the staff, and accordingly (d) not to minute the discussion and the resolution.

[4] Mitchell’s performance was in fact monitored for the first eighteen months or so of his employment and proved to be entirely satisfactory. Indeed, he was highly regarded by his co-employees and the council alike for his competence in significantly improving the manner in which TBA’s financial affairs were conducted. By all accounts he was the dominant figure in TBA’s entire management team.

[5] Since the early 1970’s Messrs Richardson Reid acted as auditors to TBA and prepared its annual financial statements. The responsible partner was a Mr Reid. It was part of the modus operandi of Richardson Reid to do what was described in the trial as a “reperformance” of the annual year-end bank reconciliation. This was described by TBA expert witness, Wainer, as “a redoing of the task that has already been done by the entity being audited.”

[6] During 1990 Richardson Reid amalgamated with Price Waterhouse Meyernel (as it was known at the time the trial commenced), a partnership of public accountants, hereinafter referred to simply as “PW”. PW was the defendant in the Court below and is the respondent in this Court. Reid continued to be responsible for the audit of the TBA’s books of account but, following PW’s new procedures, he no longer caused a reperformance of the year-end bank reconciliations to be done. What Reid did not know, since he was never told and could not have discovered it from the appropriate minutes, was that Mitchell had a record of proven dishonesty.

[7] The last set of financial statements prepared under the supervision of Reid was for TBA’s financial year ending October 1993. The actual field work was done by three audit clerks in January 1994. Their work was initially reviewed by the audit manager, Greyling, but ultimately by Reid himself. The core of TBA’s case against PW is that there were a series of discrepancies in TBA’s books of account which ought to have alerted the auditing team, but did not, that something was amiss; that if these matters had been pursued as they should have been, Mitchell’s misdemeanours would have been discovered in January 1994; and that the thefts he committed thereafter with the consequent losses to TBA would have been averted.

[8] It was only during November 1994 that it was discovered that Mitchell had consistently stolen large sums from TBA. He had done so by a process described in the trial as “teeming and lading” or “rolling over”. This consisted of misappropriating cash or cheque payments made by members or customers of TBA and using them to “clear” earlier defalcations, thereby ostensibly “balancing” the books for the time being. When Mitchell was confronted he immediately confessed to the fact but not necessarily to all the details of his embezzlement. At that stage it was believed that the thefts had commenced after October 1993 i.e. after the period on which PW had focused for purposes of its last audit. PW, more particularly Reid, was thereupon commissioned to investigate and prepare a full report on the extent of Mitchell’s thefts. This he did. He uncovered a massive series of thefts which he quantified as being of the order of R 1 697 929.28. Mitchell was dismissed. TBA obtained summary judgment against him based on Reid’s report. (An amount in the vicinity of R100 000 was recovered from him.) At the same time and at TBA’s insistence he was prosecuted, convicted and again imprisoned.

[9] TBA’s complaint

During the course of these investigations it became apparent that Mitchell had also stolen considerable sums during 1993, later estimated to be about R300 000. It was this disclosure, that Mitchell’s thefts predated the 1993 audit, that became the source of the present proceedings against PW. PW, it was common cause, was contractually bound to exercise reasonable care in the execution of its audit and not to do the work negligently. The allegation is that it failed in that respect; that had the work been done properly, Mitchell’s theft would have been uncovered in January 1994; and that all the direct losses suffered by TBA due to Mitchell’s subsequent thefts and his inability to repay were accordingly for the defendant’s account. The factual basis of TBA’s quantification of its claim, much to PW’s indignation, was Reid’s special report commissioned by TBA after the thefts were initially discovered in November 1994.

[10] PW’s defence

PW in its plea denied that it was negligent and accordingly that it committed a breach of contract vis-à-vis TBA. In the alternative it denied that any breach it may have committed was a cause of TBA’s loss. The true cause of the TBA’s loss, so it alleged, was TBA’s own negligence first, in employing Mitchell; secondly, in retaining him as its financial officer after discovering that he had a criminal record; thirdly, in failing to inform PW thereof; and fourthly, in failing, due also to its own admittedly inadequate and lax internal controls, to properly supervise and control Mitchell’s activities. It was furthermore alleged by PW in its plea that the parties at all material times contemplated and agreed that PW’s appointment was made on the basis that it would not be liable to TBA “for any loss suffered by the latter as a result of the defendant’s breach of contract, if the plaintiff’s own negligence was the primary cause, or alternatively a material cause, or alternatively a cause of its loss”. As a final alternative it was pleaded that TBA’s claim was liable to be reduced because TBA was itself negligent, because its negligence was a contributory cause of its loss and because the Apportionment of Damages Act, 34 of 1956 (“the Act”) was applicable to its cause of action.

[11] The issues and the Court a quo’s findings thereon

The first major issue before the Court a quo (Goldstein J sitting in the Witwatersrand Local Division of the High Court) was whether PW was negligent (and hence committed a breach of contract) in not being sufficiently alert in conducting its October 1993 audit. The Court a quo held that it was and since it was common cause that TBA had suffered a loss (the quantification of which was eventually settled between the parties), the next major issue was whether such negligence was the cause of such loss. And that issue posed the next two, namely, whether TBA was not itself negligent and if so, whether such negligence was not the true cause of its loss. The Court a quo found that TBA was indeed negligent and that its negligence was the real and dominant cause of its loss. Despite that finding Goldstein J held that TBA was not non-suited because the claim was partly rescued by the Act, even though it was not framed in delict but in contract. It further held that TBA was eighty percent to blame for its own loss and that the damages it was otherwise entitled to recover had accordingly to be reduced by that percentage. The judgment has been reported as Thoroughbred Breeders’ Association of South African v Price Waterhouse 1999 (4) SA 968 (W).

[12] The parties had earlier agreed on the quantification of Mitchell’s misappropriations of cash receipts and cheques payments as being R1 389 801.90. The Court a quo deducted R143 403,44, in respect of moneys stolen before 31 October 1993. In the result the Court a quo calculated that TBA was entitled to R1 246 398,46 less eighty percent, resulting in an award of R249 279,69. Both parties were in agreement before us that the amount of R143 403.44 was wrongly deducted and that, consistently with the Court a quo’s apportionment of liability, the sum should have been R277 960.38 (being twenty percent of R1 389 801.90).

[13] TBA disputed, as a matter of law, the finding that the Act was applicable to a claim founded on breach of contract and it disputed, as matters of fact, both the finding and the Court a quo’s assessment of the proportion of TBA’s own negligence. There were other issues as well. PW had instituted a counterclaim for the payment of its agreed fees for the preparation of the special report referred to earlier. This was initially disputed by the TBA on the basis that its “obligation to pay is reciprocal to [PW’s] obligation to produce a report which is accurate and true”. The Court a quo found that there was no merit in this defence and the point was not pursued by TBA. Nor did TBA dispute its obligation to pay the agreed fees on the ground that these were foreseeable consequences of PW’s alleged breach of the routine audit contract. The agreed fee of R74 000 is accordingly to be deducted from the amount, if any, to be awarded to TBA. Other issues which remain alive are whether PW was liable to TBA for the interest TBA was obliged to pay its bank on the amount by which, due to Mitchell’s undiscovered theft, its overdraft was inflated; and from what date and at what rate interest on its claim was otherwise to be calculated. And finally there was the question of costs. Save for a special order against PW in respect of the time spent in proving the quantum which was eventually agreed (in respect of which there is still an outstanding issue, namely, whether such costs should have been awarded on the scale as between attorney and client or attorney and own client), the Court, in the exercise of its discretion, initially resolved to make no order as to costs in respect of the trial.

[14] The orders made

After a trial which commenced in May 1997 and concluded, with interruptions, some seventeen months later and generated a record of close to 6 500 pages, the Court a quo on 7 July 1999 and at 1038B-F of the report made the following order:

“1. The defendant is ordered to pay the sum of R249 279.69 to the plaintiff together with interest thereon at Nedbank’s prime rate to its most favoured customers and reckoned from date of judgment to date of payment.

2. The plaintiff is ordered to pay to the defendant the sum of R74 100 together with interest thereon at 15,5% per annum from 6 February 1996 to date of payment.

3. The defendant is ordered to pay the plaintiff’s costs, including the costs of two counsel, of 14 days’ trial spent on quantum on the scale as between attorney and client and including the costs of the argument on the admissibility of J W Mitchell’s admissions at his criminal hearing.

4. The defendant is to pay the plaintiff’s costs, including the costs of 2 counsel, of consultation with Mr D Dorfan and the latter’s qualifying fees.

5. Save as aforesaid there will be no order as to costs.

6. …”

[15] Both parties sought leave to appeal. At the same time the Court a quo considered the aftermath of a secret payment into court of R281 680 which PW had made on 16 April 1997 (before the commencement of the trial) in terms of Supreme Court rule 34(12). Leave to appeal was duly granted to both parties but in the light of the secret payment and allowing for a spatium deliberandi until 24 April 1997, the Court a quo amended its earlier order and on 22 September 1999 (at 1038H-1039D) substituted for it the following order:

“1. …

2. The order of 7 July 1999 is amended by the deletion of paragraphs 3, 4 and 5 and the substitution therefor of the following:

‘3. The defendant is ordered to pay the plaintiff’s costs, including the costs of two counsel and the qualifying fees of the plaintiff’s expert, Harvey Elliot Wainer, incurred on or before 24 April 1997.

4. The plaintiff is ordered to pay the defendant’s costs, including the costs of two counsel and the qualifying fees of the defendant’s expert witness, Mr. Guy Smith, incurred after 24 April 1997 subject to the contents of paragraph 5 below.

5. The defendant shall not be entitled to recover costs in respect of 14 days’ trial spent on quantum, including the costs of the argument on the admissibility of J W Mitchell’s admissions at his criminal hearing.’

3. …

4.1 …

4.2 …

4.3 …”

[16] This, then, is the appeal against certain aspects of that order.

[17] The first and perhaps foremost issue is whether PW breached its agreement with TBA to act as its auditor and hence to audit its financial statements for the financial period ending 1993. That question presents itself in two parts: (a) what were the terms of the agreement between the parties; and (b) did PW breach any of those terms?

[18] The terms of the agreement

The agreement in respect of the audit for 1993 was concluded when PW was routinely appointed in March 1993 as TBA’s auditors by its Annual General Meeting. This has been a recurring annual item on the agenda since the 1970’s when Richardson Reid was first appointed as auditors to TBA and Reid became the partner primarily responsible for the TBA account, both before and after Richardson Reid was absorbed into PW in 1990. The actual terms of the appointment were never formalised as such. The relationship between the parties was accordingly governed by such terms as were customary in South Africa at the time between a client and his auditor.

[19] There has been a fair measure of recent learning in various cases in various jurisdictions about an auditor’s contractual duties and responsibilities vis-à-vis his client. We were referred to a host of authorities of which I list the more prominent ones as a source for future reference: In re London and General Bank (No. 2) [1895] 2 Ch 673 (CA); In re Kingston Cotton Mill Co (No 2) [1896] 2 Ch 279 (CA); In re City Equitable Fire Insurance Co Ltd [1925] 1 Ch 407 (CA); Fomento (Sterling Area) Ltd v Selsdon Fountain Pen Co Ltd [1958] 1 All ER 11 (HL); Re Thomas Gerrard & Son Ltd [1967] 1 Ch 455; Pacific Acceptance Corporation Ltd v Forsyth and Others (1970) 92 WN(NSW) 29; Tonkwane Sawmill Co v Filmalter 1975 (2) SA 453 (W); Alexander and Others v Cambridge Credit Corporation Ltd (1987) 9 NSWLR 310; Fletcher v National Mutual Life [1990] 3 NZLR 641 (HC Auckland); AWA Ltd v Daniels t/a Deloitte Haskins & Sells (1992) 7 ACSR 759; Galoo Ltd (in liq) and Others v Bright Grahame Murray (a firm) and Another [1995] 1 All ER 16 (CA); Dairy Containers Ltd v NZI Bank Ltd [1995] 2 NZLR 30 (HC Auckland).

[20] For present purposes it is not necessary to delve into these authorities or to discuss controversies such as whether and to what extent an auditor in conducting an audit is bound nowadays to actively search for, detect, pursue and, if needs be, prevent fraud and other illegal acts (in contradistinction to mere inaccuracies, irregularities and errors) that may have been committed during the year under review. It is not so necessary because it was common cause between the parties in this case that the plaintiff would conduct its audit “in accordance with generally accepted auditing standards” (GAAS) and “with due professional care required of an auditor in public practice and would not act negligently”. It was so alleged by TBA and so admitted by PW. Being admitted, there is no need to enquire whether these were tacit terms arising ex consensu or by operation of law. Nor is it necessary to enquire whether TBA succeeded in proving the additional term which it alleged (but PW disputed) that PW would “properly verify investments shown in the accounts of the plaintiff”. Whether the failure to do so (which was common cause) constituted the breach of a particularised term or a particular manifestation of the breach of a general term is not conclusive. However it is presented, PW would have been in breach if it had been careless in the execution of any aspect of its mandate, measured against the general standards prevailing in the profession at the time.

[21] Those standards are to be gathered from relevant legislative enactments (in this case the Public Accountants’ and Auditors’ Act, 80 of 1991), from the profession’s own codifications of an auditor’s duties (such as the material issued by the South African Institute of Chartered Accountants), from authoritative publications and legal decisions, here and abroad, and from expert evidence presented to the court. AU130 of the statement issued by the South African Institute of Chartered Accountants states generally:

“Due professional care must be exercised during the examination and in the preparation of the report.

This standard requires the auditor to perform his work with due care and imposes a responsibility upon each person within an auditor’s organisation to observe the required standards of field work and reporting. Exercise of due care requires critical review at every level of supervision of the work done and of the judgment exercised by those assisting in the examination.

The application of due care concerns what the auditor does and how well he does it. For example, due care in the matter of working papers requires that their content be sufficient to provide an important support to the auditor’s opinion and to show compliance with generally accepted auditing standards.”

In this respect the evidence of the expert witnesses, Wainer for TBA and Reid and Smith for PW, was perhaps more directly in point than the other sources mentioned. The issue before the trial court, once again, was not whether it was incumbent on PW, as a particularised term of its agreement with TBA, to unmask Mitchell as a thief. The issue was whether PW was careless and thus negligent and therefore in breach of the uncontested general term in not recognising and reacting to certain irregularities and anomalies in TBA’s books of account which TBA alleges should reasonably have alerted it to Mitchell’s deviousness. TBA’s allegation, which was hotly disputed by PW, was that the work on the financial statements for the period ending 31 October 1993 was negligently done in that respect and that PW was accordingly liable for the consequences flowing from its failure to pick up and follow Mitchell’s trail. The first enquiry with which this case is accordingly concerned is whether PW was negligent in two respects in particular: (a) in not verifying the promissory note included in the item “investments” in the Futurity race programme; and (b) in not appreciating the significance in the bank reconciliations of the items described as “outstanding deposits”. I deal in sequence with each of these alleged manifestations of negligence and hence of breach of contract.

[22] The promissory note

The TBA 1993 financial statements reported separately on the Futurity race programme. That was a programme designed to provide additional stake money collected from all breeders and participating owners to be distributed to such breeders and owners whose horses finished within the first four in any of a series of designated races. The funds so collected were retained in a separate bank account and in certain selected investments. The financial statement relating to this programme showed investments of R423 665. The working papers of Ms Smit, who was responsible for the field work on this part of PW’s audit, indicated that this investment consisted of only two items, one being the balance of a secured mortgage debenture of R315 165 and the other a promissory note issued by the Department of Trade and Industry (“DTI”) and dated 7 February 1993. (The face value of the note, issued by DTI to Snoek Wholesalers (Pty) Ltd, and which TBA obtained at the discounted price of R108 500 was R138 864.) A note in Smit’s working papers referred back to PW’s 1992 working papers, which meant that she must have appreciated, if she applied her mind to it, that the date 7 February 1993 could not have been the date of issue but could only be its date of maturity and that the difference of R30 364 between the face value and the discounted value remained unaccounted for. Smit did not demand to see the original of the promissory note. Had she done so it could not have been produced to her. The reason was that TBA no longer held it. Mitchell had earlier arranged with DTI to replace the note which in the meantime had gone stale with a cheque for R138 864 which he had deposited on 3 May 1993 in the TBA sales account (not the Futurity programme bank account) and which was dealt with in the books as if it constituted a payment received from another TBA debtor, one Wilensky. In this manner Mitchell contrived to cover, in TBA’s books, payments previously made by Wilensky to TBA but which he had misappropriated. The net effect was that Wilensky was no longer reflected as a debtor but that TBA had lost the R108 500 which it had initially invested (as well as the difference between that sum and the face value of the note), although it was still reflected in the financial statements of the Futurity race programme as a current asset.

[23] TBA’s complaint is that it was negligent of Smit, knowing as she did that TBA’s system of internal control was lax (which was common cause), not to have verified the continued existence of the document when it was patent that it had matured and had thus gone stale since the previous audit in 1992. Had she done so, instead of simply relying on what had been done in 1992, she would have discovered that the document no longer existed. That discovery would have set in motion a chain of enquiry, so it is alleged, which would inevitably have led to Mitchell’s exposure and the prevention of his future defalcations. This is not, therefore, the sort of case where suspicions were aroused but not pursued; this is a case, says TBA, where suspicions were not aroused when, reasonably, they should have been aroused and pursued.

[24] Reid, PW’s partner who supervised the audit and testified on its behalf, conceded that the existence of the document should have been verified and that an explanation for its absence ought to have been asked for. If Smit had not asked for an explanation “then she did not perform her work properly”. Smit was not called as a witness by PW although she was still in PW’s employ at the time of the trial. The inference is inescapable that she did not ask Mitchell for an explanation. Reid testified that if she had done so “I cannot imagine what explanation … was or was not offered, if any”. Mr Smith, an experienced and respected auditor, who also gave expert evidence on behalf of PW, likewise confirmed that in the circumstances, more especially the fact that the maturity date had obviously expired, Smit should have insisted on seeing the document and should have demanded an explanation from Mitchell if it could not be produced. Nevertheless this failure, according to Smith, did not amount to negligence on PW’s part. One reason was that the promissory note was not in auditing terms a material item; and, that being so, it did not imperatively call for separate “verification” by the auditor concerned. (The other reason is discussed in para 42.)

[25] Materiality in this context is understood to refer to an item or matter which is significant in relation to the substantial correctness of the financial statements as a whole and which would ordinarily be calculated to influence a client (or any other regular reader of the financial statements) in his assessment thereof. As it was put by Reid in evidence: “Well, material means important in the context of the presentation of the annual financial statement.” AU010.03 of the statement issued by the South African Institute of Chartered Accountants reads:

“The elements of “materiality” and “audit risk” underlie the application of all the standards, particularly the standards of fieldwork and reporting. The concept of materiality is inherent in the work of the auditor. There should be stronger grounds to sustain the auditor’s opinion with respect to items which are relatively more important and with respect to those in which the possibilities of material error are greater than with respect to those of lesser importance or those in which the possibility of material error is remote.”

[26] From an auditing perspective the promissory note, according to Smith, was not a material item, that is to say, an audit risk which demanded separate attention. He gave a number of reasons for saying so:

1. Quantitatively, being an asset of only approximately R100 000 and representing less than 1% of TBA’s total asset holdings of R16m, it was relatively trivial.

2. The fact that it was separately treated in the financial statements as part of the Futurity race programme and reflected “an operating profit” of R79 000, whereas it was in truth a loss situation, was also not of any great consequence. That is so, firstly, because PW did not purport to render a separate report on the Futurity account on its own but reported on TBA’s financial statements as a whole; nor did TBA, which was in the habit of transferring assets freely from one fund to another, administer the Futurity programme separately. Secondly, the financial statements did not in any event reflect the true situation of TBA (including the Futurity account) because of a deliberate change in the accounting policy of TBA in 1993, when it was decided to write back the provision made in 1992 for doubtful debtors and to make no further provision therefor in the 1993 statement. This had the effect, so Smith contended, of inflating its 1993 “operating profit” by R385 388 and the value of its 1993 assets by R991 907. In the result PW felt constrained to qualify its 1993 audit report in that respect. The consequence was that the ostensible operating profit of R79 000 in the Futurity programme paled into insignificance and would not, in the light of PW’s qualification of the financial statements, have impressed any reasonably informed reader.

3. The inclusion in the financial statements of the promissory note as an asset did not represent “an audit risk” and diminished the need for verification. Once it had been raised as an asset in TBA’s books of account, it could only be removed by either a payment or by writing it off as a separate item. Since neither eventuality was likely to occur its loss by theft was bound to be discovered at the next audit. Consequently it was not vital that it should have been detected in the current one.

4. Finally, materiality is a matter for the judgment of the auditor concerned. Smit, and those above her, Greyling, the audit manager, and Reid, the partner, adjudged that it was not necessary to verify the existence of the promissory note. That being so it cannot be said that Smit’s judgment call was negligent in the circumstances.

[27] Smith’s opinion that the promissory note was not material and that the failure to verify it was accordingly of no significance in the broad scheme of things was contradicted by PW’s own principal witness, Reid, and by TBA’s expert witness, Wainer. That concession makes the discussion about materiality somewhat academic but it was persisted in in argument and must accordingly be considered. In my opinion and for the reasons that follow Smith’s assertion that the non-verification of the promissory note was not material, must be rejected. I deal with the points made by Smith in the same order.

[28] As to (a) and (b): The assets of the Futurity programme were separately identified in the financial statements. A loss (which should have been reflected) instead of a profit (which was), would doubtless have been heeded by management and members as something which required investigation and attention. It would have tended, if reported as a loss, to have instilled a sense of disquiet rather than one of comfort. Nor does the change in TBA’s accounting policy, however ill-advised it may have been, assist PW. It is a red herring. A greater falsity in the books for which TBA was responsible does not nullify a lesser one for which PW was responsible. That TBA changed its policy as regards the treatment of debtors could not therefore relieve PW of the obligation of reflecting a true state of affairs in the financial statements.

[29] As to (c): It is true that the unavailability of the promissory note would in due course, probably at the next audit, have been detected. On the other hand, it was not detected during the 1993 audit and, if it was missed once, it could have been missed twice. In the meantime, because of its non-detection, Mitchell was enabled to misappropriate moneys in excess of a million rand. An item which might otherwise not have been regarded as an audit risk (and hence as material) may become material precisely because it stands out, or ought to do so to the alert auditor, as being anomalous, irregular, unusual or illegal and as such as demanding of further investigation. As it is stated in PW’s own audit manual:

“Sometimes a matter may not be significant in itself but may have implications for other matters which are material.”

That would be particularly so when such an item may be indicative of a recurring irregularity or of a flaw in the system or of dishonesty. This, according to TBA, was such a case. I agree. Materiality should not be judged in isolation. It does not depend merely on the magnitude or not of the item relative to the whole but also on its actual or potential implications relative to other items or relative to the future. The stale promissory note was an anomaly which it was common cause between the witnesses for both sides, called for further investigation. In the absence of a satisfactory explanation from management the stale promissory note could be a pointer to other irregularities in TBA’s books of account. As such it was material.

[30] As to (d): It may well be that materiality is initially a matter for the judgment of the auditor himself. That does not invest him with immunity should his judgment afterwards be adjudged to have been so conspicuously wrong as to be symptomatic of carelessness on his part. In this case, moreover, it is by no means clear from her working note whether Smit consciously exercised any judgment at all not to verify the promissory note. This was conceded by Reid. She simply reproduced the item from the 1992 statements and either overlooked or ignored the warning signal that it was outdated. This is in contrast to the manner in which the promissory note was treated in the 1992 audit where it received meticulous attention. By way of further contrast other items in the 1993 audit with a value far less than R100 000 were in fact closely scrutinised.

[31] The Court a quo’s treatment of this aspect (at 1005H-J of the report) was odd. It expressed a preference for Smith’s view, notwithstanding Reid’s concession to the contrary, that the non-verification of the promissory note was not material. It then stated:

“Furthermore, the TBA is an association not for profit. An operating loss of about R30 000 in place of a profit of R79 893 does not seem of great significance when one considers that the TBA has the substantial number of 800 members.

It seems to me too that there is a fatal flaw in TBA’s case relating to materiality of the promissory note, and that is that no evidence has been led on Mitchell’s financial position as at 31 October 1993. If he were able to repay the amount stolen by him on that date, the figures in the operating profit would have been unaffected and Mitchell’s debt would have replaced the promissory note as an asset.”

The first paragraph contains a non sequitur. The second entails a disregard of both the significance of a discovery that the note was missing and of the probabilities. As to the former, the significance of such a discovery would have been that Mitchell’s inability to produce the note coupled with an inability to provide a plausible explanation as to how it was accounted for (a matter dealt with in para 42), would have led to his unmasking as a thief. His ability, if any, to repay the amount misappropriated would not have altered that and would have been irrelevant to the materiality of the misappropriation. As to the latter, in the light of Mitchell’s proven gambling addiction, and his thefts which commenced in March 1993 and appear to have increased exponentially until November 1994 it is surprising to learn that he might have been able to make good his defalcations in October 1993. There was not an iota of evidence that pointed in that direction. The Court then continued (at 1006B-D):

“It follows from the aforegoing that I am of the view that TBA did not discharge its onus in regard to materiality. In reaching this conclusion I do not overlook that Reid says that he would have verified the promissory note and that Smith too, despite his evidence against materiality, says that he would have done so. These factors are, however, not weighty enough. It follows, too, that I cannot find that Smit should have verified the promissory note even if there was no indication of anything amiss in regard thereto.”

Having first found that Smit need not have verified the promissory note, the Court then proceeded to find that she should at least have examined it and that she was negligent in not doing so. This finding is a little difficult to reconcile with the earlier finding that the promissory note was not a material item. If it was not a material item there was no duty on Smit to examine it and consequently she would not have been negligent in failing to do so.

[32] For the reasons stated earlier I am of the view that the Court a quo erred in finding that the promissory note was not a material feature of the financial statements requiring verification. It follows that there was a duty on PW to do so and that its failure was negligent in the circumstances and constituted a breach of contract. I deal with the consequences of that finding later in this judgment in conjunction with the next issue viz, whether PW committed a breach of contract in not appreciating the significance of what was described during the trial as the “outstanding deposits”.

[33] The outstanding deposits:

The audit clerk who was deputed to perform the field work on this part of the audit was a Mr A. Ford. Like his counterpart, Ms Smit, he was still, at the time of the trial, in the employ of PW but was not called to give evidence on its behalf. His working papers reveal that he had reviewed the bank reconciliations and that he reported that “the overall results were satisfactory”. His working papers were reviewed by Greyling, the audit manager, who was likewise not called to give evidence, and by a partner of PW, Reid, who was.

[34] TBA operated several accounts with its bank. One of them was the sales bank account, known as “TBA Sales”. Ford’s duties in terms of his working programme included the duty to:

“examine monthly bank reconciliations and investigate any unusual reconciling items, outstanding deposits, etc.”

This involved a comparison and reconciliation of TBA’s cash book and its monthly and year-end bank statements. Ford’s note, in elaboration of what he had “done”, reads:

“Work done

Examined monthly bank recons for Authorisation of Bank recons by John Mitchell and for large and unusual items.

Results

1. John Mitchell (accountant) sighnes (sic) the bank recons monthly as evidence of his reviews (not all recons were sighned) (sic).

2. Unpaid cheques were not written back.

3. Overall results were satisfactory.”

Mitchell in fact did not sign the monthly bank reconciliations. Ford also omitted to mention the “outstanding deposits”. Those referred to a series of entries in the monthly bank reconciliations routinely done by Ms Gloria Seamen, an employee of TBA who worked under Mitchell. These bank reconciliations reveal that payments received by TBA and recorded in the cash book, and for which receipts were issued, were not, as one would have expected, immediately banked. So for example the bank reconciliation as at 25 April 1997 reflected three payments of R23 700, R52 700 and R25 300 respectively for which receipts had been issued but which had not been deposited at the bank. The first two items were still undeposited a month later and the amount of R23 000, which in the meantime had been listed with other items described as “outstanding deposits”, was still recorded as “outstanding in the bank reconciliation of September 1993”. It was only “cleared” (as were all the other “outstanding deposits” totalling as at that date, R148 903,44), at the end of September 1993, in time for the year-end reconciliation of October 1993 which Mitchell knew would be reviewed by the auditor. As the Court a quo remarked (at 978G-H):

“Of course, a long-outstanding deposit is a cause for suspicion. Money, and especially cash, ought to be deposited as soon as possible. If this does not occur a reasonable suspicion arises that the money has been stolen.”

The Court a quo regarded it (at 979J-980A) as an:

“overwhelming probability arising from Ford’s working papers that he must have overlooked the outstanding deposits despite his having perused the bank reconciliations, or at least, that he failed to recognise their significance if he saw them.”

[35] A further feature, not commented on by Ford, was the manner in which the outstanding deposits were ‘cleared’. This was done inter alia by two cheques of R25 000 and R18 000 respectively signed by Mitchell and drawn on the TBA sales account in favour of itself. This could conceivably have been a legitimate means of reconciling the bank statements with the cash book on paper but only if there were vouchers substantiating that the payments which had not been banked had been used in the course of TBA’s operations. It is common cause that there were none. In the circumstances it is difficult to escape the conclusion that the clearing of the outstanding deposits in this manner was simply a ruse by Mitchell to pull the wool over the auditors’ eyes.

[36] According to Wainer the outstanding deposits as an item was not the only anomaly that Ford should have picked up but did not. It is common cause that if Ford had done a reperformance of the 1993 year-end bank reconciliations he would, in the words of Wainer, have identified

“numerous deposits in the cash books which were not in agreement with the bank statement entries. The differences identified indicate deposits in the cash book not appearing in the bank statements; deposits in the bank statement not appearing in the cash book, and deposits where the amount in the cash book is not in agreement with the amounts per the bank statement ...”

There is a difference between Wainer and Smith as to whether it was necessary for PW to have done such a reperformance. According to Wainer it should have been done because TBA was a relatively small and tightly-knit entity with admittedly poor internal controls. Smith disagreed. The Court a quo preferred the view expressed on this point by Wainer. It also analysed and commented on other discrepancies, not noted by Ford, which would have become apparent from a reconciliation of the deposits in the cash book and the bank statements during the period September to October 1993, and which caused it to conclude (at 981I-982A):

“None of the above discrepancies, anomalies or coincidences appear from Ford’s working papers. On the contrary, as we have seen, his working papers indicate that he was satisfied that none such existed. The inference is accordingly inescapable, in the absence of an explanation from him, that his investigation of unusual reconciling items was superficial and inadequate – a matter to which I shall return below.”

[37] It is an issue between the parties whether PW was obliged to do a reperformance of the year-end bank reconciliation. TBA pleaded that it was a specific term of the agreement that such a reperformance should have been done. Had such term been proved it would have been the end of the matter because its breach was common cause and its consequences were clear: Mitchell’s manipulations would have been exposed forthwith. But the existence or not of such a specific term was not a matter that ever exercised TBA’s mind. Until 1990 when Richardson Reid was absorbed by PW it was the practice to reperform the year-end bank reconciliation. This practice was discontinued (in favour of PW’s own “standard procedure”) when PW took over Richardson Reid (but was, incidentally, reinstated in the case of small entities as a result of PW’s experience in this case). TBA left it to PW to decide how to conduct its audit. Its agreement with PW was in the broad terms that it should do its work with proper and reasonable care. In the circumstances it cannot be said that TBA discharged the onus of establishing the specific term pleaded. Nor can it confidently be said that the failure to do such a reperformance was a manifestation of work carelessly done. It is in any event not necessary to resolve this issue between Wainer and Smith for whatever differences there may be between them on that point, all the experts, including Reid, were agreed that Ford was remiss in not examining the outstanding deposits and in not insisting on an explanation from management. It was conceded in argument that Ford did not do his work properly in that regard. His failure to do so was not picked up and corrected by either Greyling or Reid who were supposed to supervise his work. To that extent it appears to be common cause that PW did not give that particular aspect of its audit the attention it deserved and accordingly that it was negligent and hence in breach of its contract with TBA.

[38] It follows from what has been said above that TBA succeeded in proving that PW was negligent and hence committed a breach of its contract with TBA in at least the two respects outlined above, the failure to appreciate the significance of and to react to the non-existent promissory note and the failure to appreciate the implications of the recurring item “outstanding deposits” in the bank reconciliation statements.

[39] It is central to TBA’s case that if PW had been more astute in either of these respects it would have prompted what was described as “a train of enquiry” which, if done competently, would in the ordinary course of the audit inevitably have revealed that Mitchell was systematically siphoning off funds for which he would have been unable to account. This is disputed by PW. But what can hardly be disputed is that Mitchell was embezzling vast sums of money from TBA. When ultimately confronted he immediately capitulated. Reid’s own special report established the extent of Mitchell’s larceny and Mitchell, having confessed, pleaded guilty and was sentenced to a term of imprisonment on the strength thereof. If the earlier thefts had been uncovered in January 1994 there can, moreover, be little doubt that Mitchell, given his past record, would have been dismissed or at the very best for him prevented from committing the massive thefts that occurred after February 1994 and that he would have been unable to repay the missing amounts.

[40] The quantum of TBA’s losses was finally agreed between the parties to be R1 389 801.90. That sum includes thefts committed prior to February 1994 but since it was established that Mitchell made good his earlier thefts by later ones in order to “clear” the earlier ones in TBA’s books of account, the full amount was in principle recoverable from PW. This was not contested by PW.

[41] Causation

The next issue is whether TBA succeeded in showing that the breach of contract it proved caused the losses it sustained. This is disputed by PW on three main grounds:

(a) Mitchell would have been able to deflect any enquiries directed to him;

(b) The loss suffered by TBA, having regard to the test for causation in a claim for damages for breach of contract, was too remote;

(c) The dominant cause of TBA’s loss was its own negligence.

I deal in turn with each of these topics.

[42] Would the discrepancies have been explained away?

Reid and even more so Smith were adamant that if Smit or Ford or Greyling or even Reid himself had asked Mitchell about any of the matters that should have concerned them he would have been able to satisfy them in January 1994 that nothing was fundamentally amiss. That is because they would not have been predisposed to suspect dishonesty let alone larceny, on Mitchell’s part. This was not, after all, a forensic audit and PW’s personnel, not having been briefed about Mitchell’s past, had no cause to approach any explanation he may have furnished with scepticism and suspicion. By all accounts Mitchell was an intelligent, skilled and plausible person who commanded the unqualified respect of his co-employees. He was able to keep them at bay and he would likewise have been able, so it was contended, to explain any matter relating to TBA’s books to the satisfaction of any reasonable auditor, at least until the next audit. (As it happens his defalcations were discovered in November 1994 when the scale of his gambling losses reached such proportions that they could no longer be covered up by a manipulation of TBA’s books of account.) But when asked what possible explanation Mitchell could have given which in February 1994 would have satisfied a reasonable auditor Reid was unwilling to speculate and Smith’s speculations were unconvincing. Mitchell may have been able to hoodwink his colleagues; it by no means follows that he would likewise have been able to bluff an observant auditor investigating apparent irregularities in TBA’s books of account.

[43] But perhaps most significantly neither Smit nor Ford was called by PW to explain whether either of them paid any attention to the anomalies described earlier, whether it occurred to them to confront Mitchell, whether they asked him for any explanations and what explanations could conceivably have been given which would have satisfied them. In the absence of such evidence it is idle to conjure up possible explanations, as Smith sought to do. What is at least clear is that there were no obviously plausible explanations, for if there had been any Smith would have thought of them.

[44] According to the Court a quo Mitchell might yet have been able to explain away the missing promissory note but not the outstanding deposits. As to the promissory note the Court a quo said (at 1011D-G):

“I return to Smit and Mitchell and the promissory note. Smith speculates that Mitchell may well have proffered the following explanation to Smit: that he had in error not presented the note on due date, that he had thereafter contacted the Department of Trade and Industries and requested a replacement cheque, that he had returned the note to the department and been promised such a cheque when funds became available, that such had not yet occurred but that in the new financial year the department would have the funds and would pay the TBA. Mitchell might even have produced a letter from his computer, directed to the department and confirming these arrangements. Mitchell was by all accounts, Smith points out, a plausible liar and given his seniority Smit would have been entitled to believe him and do nothing further. There is no basis for me to reject this evidence of Smith on a balance of probabilities and I shall accept that Mitchell may have been able to overcome the hurdle of the promissory note when it was presented to him on or just prior to 20 January 1994. The matter does, of course, not end there.”

As to the outstanding deposits the Court a quo said (at 1012G-I):

“Of course, he [Ford] could have confronted Mitchell himself but I think, overwhelmingly on the probabilities, he would have consulted his immediate superior, Smit, and explained the disturbing facts to her. She would then have recalled the missing promissory note and Mitchell’s explanation of it. In my view, she would then, as a reasonable diligent auditor, have decided that it was necessary to question Mitchell on the outstanding deposits and to revisit the problem of the promissory note. Before doing so she ought, in my view, to have reperformed or caused Ford to reperform the October bank reconciliation to see how the outstanding deposits were cleared.”

And if that had been done, so the Court finds, she and Ford would have become aware of the serious discrepancies which are apparent when the cash book is compared to the deposits on the October 1993 bank statement. The Court then concluded (at 104A-B):

“In my view, this was clearly a case with so much calling for explanation that the level of suspicion of the reasonably astute auditor ought to have been so high that a thorough investigation was called for; the matter required in fact a ‘probe ... to the bottom’.”

(Of course, as stated earlier, if such a probe had been attempted Mitchell’s thefts would have been uncovered.)

[45] I agree with the conclusion which can in my opinion be reached by a more direct route. On the supposition that Smit discovered the non-existence of the promissory note, that Ford appreciated the significance of the outstanding deposits, that they reported these peculiarities to their superiors and that PW realised that something was seriously out of kilter, the scale thereof would have driven the auditors firstly, to seek corroboration for any possible explanation Mitchell might have furnished them and secondly, to reconcile the cash book, the receipts, and the deposits reflected on the bank statements, especially the October 1993 bank statement. Had they done so it would have revealed the other discrepancies detailed by the Court in its judgment. The thefts would then have been discovered. Further thefts would have been averted. TBA would not have suffered its loss.

[46] Does the loss flow from the breach?

Factual causation being a given, was the loss not too remote? The traditional approach for determining remoteness in a contractual context was restated in 1977 by Corbett JA in Holmdene Brickworks v Roberts Construction Company 1977 (3) SA 670 (A) 687D-F in the following terms:

“To ensure that undue hardship is not imposed on the defaulting party … the defaulting party’s liability is limited in terms of broad principles of causation and remoteness, to (a) those damages that flow naturally and generally from the kind of breach of contract in question and which the law presumes the parties contemplated as a probable result of the breach, and (b) those damages that, although caused by the breach of contract, are ordinarily regarded in law as being too remote to be recoverable unless, in the special circumstances attending the conclusion of the contract, the parties actually or presumptively contemplated that they would probably result from its breach (Shatz Investments (Pty.) Ltd. v. Kalovyrnas, 1976 (2) S.A. 545 (A.D.) at p. 550). The two limbs, (a) and (b), of the above-stated limitation upon the defaulting party’s liability for damages correspond closely to the well-known two rules in the English case of Hadley v. Baxendale, 156 E.R. 145, which read as follows (at p. 151):

‘Where two parties have made a contract which one of them has broken, the damages which the other party ought to receive in respect of such breach of contract should be such as may fairly and reasonably be considered either arising naturally, i.e., according to the usual course of things, from such breach of contract itself, or such as may reasonably be supposed to have been in the contemplation of both parties, at the time they made the contract, as the probable result of the breach of it.’

As was pointed out in the Victoria Falls case, supra, the laws of Holland and England are in substantial agreement on this point. The damages described in limb (a) and the first rule in Hadley v. Baxendale are often labelled “general” or “intrinsic” damages, while those described in limb (b) and the second rule in Hadley v. Baxendale are called “special” or “extrinsic” damages.

It was suggested in argument that in the present case the damages claimed were special or extrinsic and had to be considered in terms of the test laid down in limb (b) above. As a corollary to this the Court was invited to resolve the controversy as to whether in this connection the “contemplation principle” or the “convention principle” should prevail (see Shatz’s case, supra at pp. 552-4, in which the point was left open). In my opinion, however, for the reasons which follow, it is limb (a) that is relevant and I see no need to accede to counsel’s invitation.”

[47] It is apparent from the above dictum that “contemplation” is the minimum desideratum common to both so-called limbs or sub-rules. The controversy referred to in the dictum, which was identified in Shatz Investments (Pty) Ltd v Kalovyrnas, 1976 (2) SA 545 (A) at 552 and which remains unresolved to this day, relates to limb (b) and not to limb (a): it is whether “the rationale of special damages is the parties’ convention and not merely their contemplation” (Shatz at 552C), that is to say, whether the contemplation of the parties must be shown “virtually to be a term of the contract” (at 552D). One of the disputes in this case was precisely whether TBA’s claim was to be classed under limb (a) (termed the “default position” by Cartwright in 1996 Cambridge L J 488 490 513) or limb (b). TBA pleaded its case in the alternative under both limbs but contended that it was properly to be accommodated under limb (a). PW on the other hand contended that it was immaterial under which limb the claim was advanced because even at the lesser threshold of limb (a), requiring proof of the parties’ actual or presumed contemplation of the loss as a “probable” result of the breach, TBA must fail.

[48] According to PW the kind and extent of loss suffered by TBA could never have been in the contemplation of the parties, either actually or presumptively, as a probable result of the type of breach committed by PW. PW’s breach of contract, so it was contended, if any, was that it was negligent not in the actual performance of its audit but in failing to appreciate the full implications of certain admitted irregularities. TBA sustained a loss because its financial manager stole its money and his thefts remained undetected; that loss could not be said to have been of the kind “that flows naturally and generally” from PW’s breach of contract; such thefts are not according to the usual course of things the probable result of a mere oversight; it might conceivably be a foreseeable risk but it was never a probable result. That was the argument.

[49] It is not altogether clear whether the word “probable” in the phrase “which the law presumes the parties contemplated as a probable result of the breach” in the dictum from the Holmdene Brickworks case, quoted earlier, (previously endorsed by Innes CJ in Victoria Falls & Transvaal Power Co Ltd v Consolidated Langlaagte Mines 1915 AD 1 22 (“likely”) and by Curlewis JA in Lavery & Co Ltd v Jungheinrich 1931 AD 156 169 (“probable”)) is to be understood in the sense of “more likely to occur than not”. Corbett JA was not in the quoted phrase formulating the rule (i.e. damages flowing naturally and generally from the breach) but only its supposed rationale (the presumed contemplation of the parties) and it is far from certain that he meant to introduce “high probability” as a further limiting factor under the first sub-rule. In this field South Africa has tended to follow the contours of the English law. English law deferred to Pothier (Obligations sec 159-162) and South African law deferred to Pothier and the English law (compare Hadley v Baxendale (1854) 9 Ex 341 354; Emslie v African Merchants Ltd 1908 EDC 82 90-91; Victoria Falls and Transvaal Power Co Ltd v Consolidated Langlaagte Mines, supra, 22; Lavery & Co Ltd v Jungheinrich, supra, 165-166; Erasmus (1975) 38 THRHR 363-364; Cartwright op cit Cambridge L J 488; Zimmermann, The Law of Obligations 829-830). In England the degree of likelihood required for purposes of the contemplation test has in recent years attracted close attention. These developments are discussed in some detail in the standard text books (such as McGregor on Damages, 16^th ed, para 248-274; Chitty on Contracts, 28^th ed, para 27-039-051; Cheshire, Fifoot & Furmston, Law of Contract, 13^th ed, 611-617; Treitel, The Law of Contract, 8^th ed, 855-859; Atiyah, The Law of Contract, 3^rd ed, 318-323 and 15 Stair Memorial Encyclopaedia para 903-905), with particular reference to what was said in Koufos v C Czarnikow Ltd (The Heron II) [1969] 1 AC 350 (HL) and the cases following it, such as Balfour Beatty Construction (Scotland) Ltd v Scottish Power plc 1994 SC 20 (HL). (See, too, the helpful exposition in Kerr, The Principles of the Law of Contract, 5^th ed 700-709). The formulae used ranged from ‘real danger’ or ‘very substantial’ to ‘easily foreseeable’, ‘liable to result’ or ‘not unlikely’ (Treitel, op cit, 857). The Heron II, supra, was referred to in both Shatz’s case, supra, and Holmdene Brickworks, supra, but in neither case, unlike this one, was the exact shading or nuance of meaning of any consequence. Even so, it is not necessary to trace the minute developments in the English decisions in this case for I believe that McGregor in para 264 of the work cited has fairly captured the essence of current English thinking on the point when he stated:

“The important factor is therefore whether the particular type of loss which occurs is within the contemplation of the contracting parties as a serious possibility …”

Or, as it was put by Goff J in The Pagase [1981] 1 Ll R 175 182:

“…but the thread running through the speeches [in the Heron II] is that the damages must have been within the contemplation of the defendant, not in the sense that they were probable (which would be too strict a test) but rather in the sense that there was a serious possibility of their occurrence or that they were not unlikely to occur.”

That approach, postulating as it does not a likelihood (at the upper end of the scale) of the harm complained of occurring but (at the lower end) a realistic possibility thereof, appears to me to be sensible and sound. Parties cannot contemplate what they cannot foresee. In the end it will usually turn on the degree of foreseeability of the kind of harm incurred (compare McElroy Milne v Commercial Electronics Ltd [1993] 1 NZLR 39(CA) 43 45). What matters to the law is of course not infinite but reasonable foreseeability. Leaving aside atypical situations (such as, for instance, a circumstance which was foreseeable by only one of the parties or only at the time of breach and not also at the time of contract), what is required to be reasonably foreseeable is not that the type of event or circumstance causing the loss will in all probability occur but minimally that its occurrence is not improbable and would tend to follow upon the breach as a matter of course.

[50] I cannot agree that the loss suffered by TBA did not in that sense flow naturally and generally from PW’s breach (compare Bruce NO v Berman 1963 (3) SA 21 (T) 23G-24E). Both the non-existent promissory note and the outstanding deposits should have struck Smit and Ford, if they had done their work properly, as odd. There would have been a reason for it. That reason could have been either innocent or not innocent – although any possible innocent explanation was clearly not an obvious one or Reid and Smith would immediately have suggested it in evidence. If the reason was not an innocent one it was an indication either of neglect or of dishonesty on the part of one or more of TBA’s personnel. Dishonesty was consequently one of three conceivable and predictable reasons. It was a realistic possibility. PW’s breach consisted not only of the failure to read the warning signs but also in its failure to probe them further. Had it done so Mitchell’s past thefts would as a matter of course have been uncovered and his future ones avoided. A competent auditor would know that the failure to recognise, identify and engage a problem of this kind may well lead to a prospective loss of this nature for his client which cannot be redeemed from the thief. This is what happened. TBA’s loss, in the words of Corbett JA in Holmdene Brickworks, supra, at 687D-F, did “flow naturally and generally” from PW’s breach.

[51] That being so it is not strictly speaking necessary to revisit, in general, the dichotomous orthodox approach of this Court to remoteness in contract. Nor is this the occasion, as it was not the occasion in both Shatz Investments, supra, 554F-G and Holmdene Brickworks, supra, 688A, to review limb (b) in particular. (The “convention principle” embraced by Wessels JA in Lavery & Co v Jungheinrich, supra, at 176 has long been discredited in England. Compare The Pegase, supra, 182-183; Cartwright, op cit, 492.) But it may be worth noting that this Court’s approach to legal causation within other disparate fields such as crime, delict, insurance and latterly, perhaps, estoppel, has undergone considerable evolution in recent years by the development of a new model for causation sometimes termed the flexible or supple test. (Compare S v Mokgethi en Andere 1990 (1) SA 32 (A) 39I-41A; International Shipping Company (Pty) Ltd v Bentley 1990 (1) SA 680 (A) 700H-701F; Smit v Abrahams 1994 (4) SA 1 (A) 15B-18H; Stellenbosch Farmers’ Winery Ltd v Apostolos Vlachos t/a Liquor Den case number 117/99, not yet reported.) In Standard Chartered Bank of Canada v Nedperm Bank Ltd 1994 (4) SA 747 (A) at 765A-B the new test was described, again by Corbett CJ, as:

“… a flexible one in which factors such as reasonable foreseeability, directness, the absence or presence of a novus actus interveniens, legal policy, reasonability, fairness and justice all play their part.”

[52] When the matter, which was deferred for future consideration in Shatz’s case, supra, does eventually come before this Court as a pertinent issue, it may be appropriate to employ the learning developed in those cases to good advantage. With breach of contract, as in delict and estoppel but unlike insurance (which entails the interpretation of the terms of the policy – compare Napier v Collett and Another 1995 (3) SA 140 (A)), the exercise would involve measuring the consequences of wrongful conduct by a composite legal yardstick. Admittedly there is an important factor present in contract and absent in the other categories mentioned and that is the competence of the parties to regulate, limit or expand by arrangement amongst themselves the consequences of any prospective breach (compare Kerr, op cit 648). Such arrangements can and must of course be accommodated in any flexible test. A conjectured application of the flexible test will not mean that the collected wisdom of past cases is summarily to be discarded. Both limbs of the current conventional test can readily be blended into an integrated test as being relevant factors to be taken into account. The fact that both parties had particular consequences in mind when they concluded their agreement will still be conclusive. There may be instances where the time of breach will be more appropriate than the time of contract. The circumstances of each case will determine where the emphasis belongs. Reasonable foreseeability, one imagines, will govern most but not all cases (compare Holmdene Brickworks, supra, 688G-H; Smit v Abrahams, supra, 17 D-F; Kerr, op cit 718). Ultimately it may be practical common sense based on the judicial officer’s years of experience – and not dogma – that has to cut the Gordian knot. (Compare Alexander v Cambridge Credit Corporation Ltd (1987) 9 NSWLR 310, 315B-C; 358B-C.) As has recently been said by Lord Steyn in a slightly different context in Smith New Court Securities v Scrimgeour Vickers (Asset Management) [1996] 4 All ER 769 (HL) 794j-795b:

“The development of a single satisfactory theory of causation has taxed great academic minds … But, as yet, it seems to me that no satisfactory theory capable of solving the infinite variety of practical problems has been found. Our case law yields few secure footholds. But it is settled that at any rate in the law of obligations causation is to be categorised as an issue of fact. What has further been established is that the ‘but for’ test, although it often yields the right answer, does not always do so. That has led judges to apply the pragmatic test whether the condition in question was a substantial factor in producing the result. On other occasions judges assert that the guiding criterion is whether in common sense terms there is a sufficient causal connection … There is no material difference between these two approaches. While acknowledging that this hardly amounts to an intellectually satisfying theory of causation, that is how I must approach the question of causation.”

[53] The only fundamental difference between the current and the suggested approaches is that there will be a more expansive solvent more suitable for all circumstances (compare Kerr, op cit 716). The exclusive criteria of the past will become auxiliary criteria in the future. There may of course be other repercussions relating to matters such as pleadings of which time alone will tell but those are not the problems of today but of tomorrow.

[54] Was TBA itself negligent?

The primary cause for TBA’s loss was of course the dishonesty of its own financial manager, Mitchell, who stole and continued to steal its money. PW laid great stress on TBA’s own alleged negligence, rather than on PW’s own breach, as being the true, real, effective, dominant or overwhelming cause of TBA’s loss. PW accused TBA of being negligent in not protecting its own assets and in not forestalling its own loss in one or more of several respects with which, before returning to the issue of causation I deal in the paragraphs that follow. (Since TBA owes itself no legally enforceable duty to protect its assets against the risk of theft by its own personnel, it is perhaps more accurate to speak of carelessness in this connection rather than negligence.)

[55] The first complaint is that TBA confirmed Mitchell’s appointment in the sensitive position of financial manager when, during his period of probation, TBA discovered that he had been convicted of theft, had served a term of imprisonment and had lied to it in his application for the position. But was it irresponsible or, as it was put on PW’s behalf, “reckless” of TBA to do so? In our view not necessarily so. It may have been a calculated risk, but it was not careless in itself. It would only have been negligent in the broad sense if it was reasonably foreseeable that Mitchell, far from being reformed as he claimed to be, would steal from it in future. There was no warrant in the facts for making that assumption. What is of course true, as a matter of common experience, is that a man with a criminal record for theft is more likely to steal than someone with no such precious convictions. Consequently an employer should be even more cautious in giving such a person unsupervised access to sizable sums of money. The “negligence”, if any, would have consisted not in his appointment as such but in TBA’s failure to introduce sufficient hedging mechanisms to prevent a repetition of Mitchell’s past misdeeds. The point is dealt with further in para 59 below. Indeed, it was a humanitarian act on TBA’s part to confirm Mitchell’s appointment on the basis of his potential. In turn that meant that the information about his criminal past had to be kept hidden from the other members of staff. That is also the reason why the discussion of the matter and the decision of council were not minuted or disclosed to its auditor or insurers. (TBA paid a price for its magnanimity: a potential claim in terms of its fidelity insurance policy was for that reason not even made.) Mitchell’s permanent appointment, even if open to criticism, was anterior to and was thus overtaken by PW’s later breach of contract. PW, as auditors, were entrusted with the auditing of TBA’s financial statements. There was admittedly no contractual obligation on PW to detect and prevent illegal acts (as opposed to bona fide irregularities and errors) on the part of management. But TBA’s complaint against PW was not that it failed to detect and prevent Mitchell’s thefts but that it failed to read the signs of possible illegal activities within the ranks of its management team and that it failed to respond thereto when it would have been reasonable for it to have done so. The complaint is that PW, had it interpreted the signs correctly, would have been able to intercept Mitchell’s later thefts. In our view the confirmation of Mitchell’s appointment as such, notwithstanding his criminal record, was therefore not per se “negligent” and was not a relevant cause of TBA’s loss.

[56] The second complaint was that TBA’s financial management of its affairs and its business practices and controls were notoriously inadequate, thereby enabling Mitchell to exploit the weaknesses in its system to his own dishonest advantage. It disregarded, so it was contended, the provisions of its own constitution and bye-laws, such as clause 11.2 which required all cheques to be signed by two of its officials. It was common cause that this did not happen and that this was one of the means by which Mitchell was able to siphon off money from TBA. The difficulty with this argument is that PW was fully aware that TBA’s controls were lax. It undertook its mandate to audit TBA’s financial statements on that very basis. As such there was in our opinion a duty on PW to be extra attentive; and it was not open to it to complain of a form of neglect which it was contracted to take into account.

[57] The next complaint is that TBA did not advise PW when it was commissioned to do the audit that Mitchell, who was in charge of its financial affairs, had a criminal record involving theft and dishonesty. Was it careless of TBA not to inform PW of Mitchell’s past and if so, was this the crucial or at the very least a relevant cause of TBA’s loss? The first observation to be made is that the omission had no effect on PW’s breach of contract as such. PW committed its breach unaware of Mitchell’s track record and purely on its own terms. If, as has been held earlier, PW should in the course of its ordinary audit have uncovered Mitchell’s past thefts without the benefit of any knowledge of his past record, then it is irrelevant for purposes of its breach whether PW was so informed or not. If the information had been given to PW in advance it may conceivably have enhanced PW’s capability of discovering Mitchell’s thefts. But that had no effect on its contractual obligations. Since PW was obliged in the particular circumstances of this case to discover the thefts when it was, so to speak, blindfolded to the risk posed by the employment of Mitchell, it is no defence for it to say that it would have discovered the thefts if only it could see. The omission to inform PW would only have been relevant to PW’s breach if it would have made its task more onerous, not more easy. This failure so to inform PW accordingly had no effect on PW’s breach as such.

[58] But of course the question is not whether TBA was negligent vis-à-vis PW but whether it was “negligent” vis-à-vis the protection of its own interests and the avoidance of its own loss. The proper question for purposes of causation is thus whether TBA should have appreciated that the loss could and would have been avoided if TBA had briefed PW about Mitchell’s criminal past. At best for PW it was argued that the failure on TBA’s part to disclose information concerning Mitchell’s dishonesty precluded PW “from properly assessing the risk” and “from planning their audit accordingly”. But the evidence fell far short of establishing that if Smit and Ford had been told (even on condition of complete confidentiality) they would have approached their task differently and they would have uncovered the thefts. Smit and Ford were not called to say that. Had it been PW’s case that TBA should have appreciated this and that it would probably have discovered Mitchell’s defalcations if only it had been told of his prior criminal record in advance of the audit being undertaken, the failure so to inform PW may indeed be said to have been the effective cause of TBA’s loss. But that was not the case pleaded by PW nor was it the thrust of its evidence. There was a passing suggestion that PW would not have accepted the commission as auditors if it had been told of Mitchell’s previous misdemeanours, but that allegation, if accepted, would not in itself have been a self-sufficient defence to TBA’s claim based on PW’s breach of contract.

[59] The final and perhaps major complaint was that TBA, knowing of Mitchell’s criminal past, placed him in charge of the administration of TBA’s accounts when TBA’s business had an annual turnover of R40m, much of it in cash. By the admission of TBA’s own witness, Bladergroen, it “had the fox looking after the hen coop”. TBA’s real “negligence” therefore consisted, so it was alleged, (a) in placing Mitchell in a situation where he could freely manipulate TBA’s affairs to his own advantage and (b) in failing to properly monitor and supervise his activities. Peter Fenix, a member of TBA’s council, who held the “finance and administration” portfolio, was initially entrusted with the responsibility of supervising Mitchell’s performance. This he did for approximately 18 months until March 1993, when he was succeeded by Allem. Allem testified that he was not aware that he was supposed to scrutinise Mitchell’s work. Hawkins, TBA’s chief executive officer during much of that period, had been told of Mitchell’s past but by all accounts (except his own, for he was not called to testify) lacked the personality, the expertise and the experience to keep a firm hold on Mitchell’s activities. It is not without significance that Mitchell’s first known defalcation commenced when Fenix’s term as council member ended. I agree with the criticism expressed by PW and accepted by the Court below that TBA’s supervision of Mitchell, given his background, was demonstrably inadequate and constituted carelessness on its part. Such lack of supervision continued after the 1993 audit and until Mitchell was eventually exposed.

[60] The cumulative effect of the relevant factors mentioned above is that it can fairly be said that TBA’s conduct fell short of the standard and degree of care and attention which an organisation of this nature ought to have exercised over its own management and ought to have devoted to its own affairs.

[61] Was TBA’s carelessness the sole or dominant cause of its loss?

Both parties were careless. Can it be said that TBA’s carelessness was the exclusive cause of its loss? I do not think so. This is not the sort of case where harm can be said to have been caused by either one or the other of two competing causes, one for which a plaintiff and the other for which the defendant was responsible. On a finding to that effect, a plaintiff, bearing the onus to prove causation, must lose if he fails to prove that it was the cause for which the defendant was responsible. This case is better typified as one where two unrelated determinants converged in causing the loss complained of. Whether and to what extent it is necessary to disentangle and apportion between them their respective degrees of carelessness in relation to that loss is a matter to which I return later in this judgment.

[62] The Court a quo made the following finding (at 1024B-D):

“Applying these principles of our common law and approaching the matter mindful of the cases, I conclude that both causes operated significantly in bringing about the result complained of. Price Waterhouse’s failure to perform their contractual duties as auditors was an important cause of the loss. But, in my view, the highly irresponsible employment of a convicted thief in so vulnerable an area of the TBA’s business and with so little check on his behaviour was the predominant, effective or real cause of the loss suffered. It follows, applying the common law, that the TBA’s claim must fail because its own negligence in employing Mitchell as it did was the causa causans of its loss.”

It was submitted on behalf of PW, on the assumption that both parties were to blame for the loss and in support of that finding, that its own negligence was so slight and TBA’s failure to guard against the loss it suffered was so dominant that PW’s breach was to be disregarded as a sufficiently significant cause of the loss.

[63] In my view both the dictum and, following it, the submission are wrong both in fact and in law. It is wrong as a conclusion of fact since it cannot as a matter of practical common sense be said that PW’s negligence was so minimal in comparison to TBA’s carelessness as to be nullified as an effective cause of the loss. It is wrong as a proposition of law since it seeks to convert an approach which is more appropriate to the law of delict to the law of contract where it is not appropriate.

[64] In the law of delict where there is culpa on both sides the so-called “all or nothing principle” has been applied since Roman times. This is dealt with in extenso by Zimmermann, op cit, 1010-1013 1030 and 1047-1048. At 1030 it is stated:

“The fault of the plaintiff/victim was, in a way, ‘set off’ against that of the defendant/wrongdoer, with the result that ‘culpa culpam abolet’. Hence the expression of compensatio culpae or culpa compensation that came to be used to label the uncompromising approach to the problem of contributory negligence. Whether every contributory fault on the part of the victim – even culpa levissima – was originally taken to deprive him of his remedy is not quite clear. In the later usus modernus, at any rate, the issue appears to have been decided on the basis of a preponderance of fault: only if he had displayed the same or a greater degree of negligence than the wrongdoer did the victim lose his claim. Where, on the other hand, his negligence was less significant, when compared with that of the wrongdoer, his claim for damages remained completely unaffected.”

In South Africa, under the influence of English law (compare Zimmermann and Visser, Southern Cross 575-6), the all or nothing approach prevailed and its application was, in the words of Boberg, The Law of Delict, vol 1 653, “uncompromising”. He continued:

“A plaintiff who was part author of his own loss could recover nothing at all. No provision existed for comparing the negligence of the parties and awarding proportionate compensation. The plaintiff’s fall from grace, no matter how venial, cost him his remedy, and the defendant – through possibly far more negligent than the plaintiff – went scot-free. The defence was a complete one.”

[65] A similar clear-cut statement is absent in the law of contract. There is a conspicuous dearth of express authority in the Roman-Dutch law either admitting or denying the existence of a defence of preponderance of own fault to a claim for damages for breach of contract. None was quoted to us by counsel and we were unable to find any ourselves, as to the applicability or non-applicability of an all- embracing “all or nothing principle”, or any variant thereof, in a contractual setting. Nowhere is it expressly stated that a plaintiff who sued a defendant for negligently performing his contract but who was himself careless was thereby non-suited, except of course where his culpa was held to be the sole cause of his loss. On the other hand, there is also no direct authority to the effect that such a plaintiff was entitled to full payment notwithstanding his proven lack of care. Not surprisingly there is likewise no authority for the intermediate situation i.e. that a plaintiff’s claim is to be reduced in those circumstances in proportion to his own lack of precaution in preventing or minimising his loss.

[66] The defence of a preponderance of fault on the part of the plaintiff, on which the Court a quo appears to rely, is incongruent within the field of contract. Where a plaintiff can prove that the breach of the defendant was a cause of the loss (as opposed to the cause thereof) he should succeed even if there was another contributing cause for the loss, be it an innocent one, the actions of a third party (compare Nedcor Bank Ltd t/a Nedbank v Lloyd-Gray Lithographers (Pty) Ltd 2000 (4) SA 915 (SCA) para 10-12), or, logically, the carelessness of the plaintiff himself in failing to take reasonable precautions to avoid it. A defendant who commits a breach of contract does so independently of any of the extraneous factors mentioned above. All the requirements for his liability will have been fulfilled. In the absence of a contrary term in the agreement itself or of legislative intervention excluding or reducing his claim, he should therefore be held fully liable, regardless of whether the plaintiff’s culpa was the dominant or pre-eminent cause of the loss. What was said for Australia in Alexander v Cambridge Credit Corporation Ltd and Another, supra, 315B, applies, I believe, with equal force to South Africa:

“It is irrelevant to inquire whether the defendants’ default was the dominant, effective or real cause of the plaintiff’s loss. If the evidence is suggestive of multiple causation, the inquiry to be made is whether the defendants’ default was a cause of the plaintiff’s loss: Fitzgerald v Penn (1954) 91 CLR 268 at 273.”

And again, at 357G-358A:

“In my opinion the above cases do not establish the proposition that a plaintiff in an action for breach of contract must prove that the breach of contract was the real and efficient or dominant cause of the loss which he suffered. In the law of tort it is well-established that it is sufficient that the wrongful act or omission of the defendant is a material cause of the plaintiff’s injury or damage. In principle the same rule must apply in the law of contract unless the terms of the contract require the sole or dominant cause to be determined. In Simonius Vischer & Co v Holt & Thompson [1979] 2 NSWLR 322, Samuels JA, with whose judgment on this point Moffit P and Reynolds JA agreed, said (at 346) that in an action for breach of contract against an auditor it was ‘sufficient for the plaintiffs to establish that the defendants’ breaches were a cause of the loss notwithstanding that there may have been other concurrent causes’.”

[67] A plaintiff who sues for damages for breach of contract for a loss allegedly sustained through the negligence of the defendant but who was himself careless in relation to the non-avoidance of such loss may therefore be non-suited: (a) if there was a term in the contract to that effect; (b) if the plaintiff’s own carelessness is held to be the sole cause of the loss, either in its totality or, to that extent, in relation to a particular segment thereof; or (c) if the defendant’s negligence was, comparatively speaking, so negligible or minimal as to be discountable as a significant cause of the loss, which, strictly speaking, is simply an instance of (b).

[68] In the present case PW did indeed seek to invoke a contractual term to that effect. It pleaded:

“8.3 The parties at all material times contemplated and made and accepted the defendant’s appointment on the basis that the defendant would not be liable to the plaintiff for any loss suffered by the latter as a result of the defendant’s breach of contract, if the plaintiff’s own negligence was the primary cause, or alternatively a material cause, or alternatively a cause of its loss.”

No serious attempt was, however, made in evidence or in argument to press for the acceptance of the term so pleaded and it may for present purposes be disregarded.

[69] The Court a quo’s finding that TBA’s carelessness was “the predominant, effective or real cause of the loss sustained” involved a qualitative and quantitative comparison between the one party’s breach of contract and the other party’s lack of precaution. Nevertheless there was common ground capable of comparison: The failure by both parties to prevent the loss by the exercise of due care. What the Court a quo in effect found was that TBA’s carelessness was so gross that PW’s negligence paled into insignificance and was accordingly neutralised as a causative factor. I disagree with that factual finding. Both sets of carelessness contributed to the loss. The defendant’s negligence was a cause of the loss. TBA accordingly succeeded in proving the causative element of its cause of action.

[70] The Apportionment of Damages Act

Having wrongly concluded, as stated in para 62 above, that TBA must lose the Court a quo proceeded to say (at 1024D-E):

“This conclusion does not end the matter for the question which now arises is whether the TBA’s claim is not at least partly saved by the provisions of chap 1 of the Apportionment of Damages Act 34 of 1956 (the Act), ...”

My approach is somewhat different. The issue is not whether the Act can salvage something from a lost cause but whether a good cause is to be abated. The foremost question on that approach is whether the Act is applicable to a claim for damages for breach of contract where the breach consists of the negligent performance of a professional duty. To that question I now turn.

[71] Chapter 1 of the Act which is the portion thereof which is relevant for present purposes, provides as follows:

“1. Apportionment of liability in case of contributory negligence.– (1) (a) Where any person suffers damage which is caused partly by his own fault and partly by the fault of any other person, a claim in respect of that damage shall not be defeated by reason of the fault of the claimant but the damages recoverable in respect thereof shall be reduced by the court to such extent as the court may deem just and equitable having regard to the degree in which the claimant was at fault in relation to the damage.

(b) Damage shall for the purpose of paragraph (a) be regarded as having been caused by a person’s fault notwithstanding the fact that another person had an opportunity of avoiding the consequences thereof and negligently failed to do so.

(2) Where in any case to which the provisions of subsection (1) apply, one of the persons at fault avoids liability to any claimant by pleading and proving that the time within which proceedings should have been instituted or notice should have been given in connection with such proceedings should have been instituted or notice should have been given in connection with such proceedings in terms of any law, has been exceeded, such person shall not by virtue of the provisions of the said subsection, be entitled to recover damages from that claimant.

(3) For the purposes of this section “fault” includes any act or omission which would, but for the provisions of this section, have given rise to the defence of contributory negligence.”

[72] I have had the considerable benefit of reading in draft the opposing judgments of Marais JA, Farlam JA and Brand AJA, on the one hand and Olivier JA on the other. The two judgments, I would suggest, comprehensively cover the entire field and it would serve little purpose for me to traverse the same territory. My sympathies and inclination are wholly on the side of the views expressed by Olivier JA. There is, I believe, for the reasons stated by him, a pressing need for legislative intervention in a situation such as the present where the defendant’s breach of contract is defined in terms of his negligent conduct but the plaintiff, by his own carelessness, contributed to the ultimate harm. But having said that, I am afraid that I have reluctantly come to the conclusion that this particular piece of legislation does not fulfil that function. I state my reasons for saying so with a minimum of elaboration.

[73] The core concept in chapter 1 is “the defence of contributory negligence” which is foreshadowed in the words “shall not be defeated” in ss 1(a) and which is expressly referred to in ss 3. Section 1 envisages:

a) a claim by a claimant against a defendant for damages (the loss) he sustained as a result of the fault (i.e. causative negligence) of the defendant;

b) fault (causative negligence) on the part of the claimant himself which partly caused the loss;

c) a defence of contributory negligence, based on such fault, which, but for the provisions of s 1, would have availed the defendant against the claim of the claimant.

[74] As was stated in paras 63-67 above the extraneous defence of culpa compensatio was known to the common law in the law of delict but not in the law of contract. In the law of contract the claim of the claimant would not have been “defeated” by his own culpa. (Of course, it would have been a defence available to a defendant, even in a contractual setting, if the claimant’s carelessness was the sole cause of the loss – but that would ex hypothesi not have been a case where the damage was caused “partly by his own fault and partly by the fault of any other person”.) That remained the position at the time the Act was promulgated in 1956. The intention of the legislature as to the scope and range of the Act must be determined in the light of the situation prevailing at the time it was enacted. At that time the concepts of both contributory negligence and “last opportunity” were unknown to a claim based on breach of contract. That being so, it seems to me to follow that the Act was designed to address and correct a particular mischief that was identified as such within the law of delict; that it was confined to that particular mischief; and that the corresponding problem that might arise within the law of contract was never within the legislature’s compass. The express wording used in the Act does not fit a contractual claim. In my view the comfort of the Act was accordingly not available to PW in this case to counter or curtail TBA’s claim for damages.

[75] It follows from that approach that it is not open to this Court to seek to determine whether and to what extent TBA’s claim should be reduced “having regard to the degree in which the claimant was at fault in relation to the damage”. I turn, then, to the other issues remaining alive between the parties.

[76] Interest

As was stated in para 13 above it remained an issue between the parties on what basis and to what extent PW was liable to TBA for the payment of interest on its claim, which was eventually quantified at R1 389 801.90 and accepted by the parties as being the amounts stolen by Mitchell during the period February to November 1994.

[77] TBA’s main claim for interest was calculated at the rate of interest levied by its banker, Nedbank, on every individual amount stolen by Mitchell from the date upon which that particular amount was stolen. Because of the thefts TBA’s overdraft, so it was contended, was inflated and the excess attracted interest at the higher rates charged by a banker to its customer on overdraft. The factual basis relied upon for this claim is that during the financial year under consideration, i.e. 1994, TBA consistently operated its bank account in overdraft and that, but for the misappropriations by Mitchell, its overdraft would have been reduced by the exact amount stolen on the very day it was so stolen. In those circumstances, so TBA maintained, it suffered damages not only in the form of the capital amounts stolen but also in the form of additional interest for which it became liable to its banker on an overdraft that would otherwise have been reduced. TBA’s case was that these additional damages flowed naturally from PW’s breach of contract, alternatively, that the additional interest which it was thus liable to pay on overdraft was a matter which fell squarely within the contemplation of the parties, particularly since PW acted as TBA’s auditors and as such was fully aware of the existence of its perennial overdraft.

[78] Thus formulated, TBA’s claim was one not for ordinary mora interest but for additional damages in the form of interest. But as a further alternative TBA contended that it was entitled to mora interest at the rate charged by Nedbank on the appellant’s overdraft from a date upon which the quantum of TBA’s damages became reasonably ascertainable by PW. In this regard various dates were mooted by TBA, the latest being the date of Reid’s final report on Mitchell’s defalcations, i.e. 24 January 1995.

[79] PW’s contention, on the other hand, was that TBA was entitled to no more than mora interest on the capital amount, calculated in accordance with the provisions of the Prescribed Rate of Interest Act 55 of 1975 (“the latter Act”). In terms of this Act TBA would be entitled to interest on the capital sum calculated at the prescribed rate of 15,5% per annum from the date of demand or summons, whichever date is the earlier. Since no demand prior to summons was proved, the date for the commencement of the calculation would therefore be the date upon which summons was served, that is 3 November 1995.

[80] The relevant provisions of the latter Act are:

“1. Interest on a debt to be calculated at a prescribed rate in certain circumstances.- (1) If a debt bears interest and the rate at which the interest is to be calculated is not governed by any other law or by an agreement or a trade custom or in any other manner, such interest shall be calculated at the rate prescribed under subsection (2) as at the time when such interest begins to run, unless a court of law, on the ground of special circumstances relating to that debt, orders otherwise.”

(2) “The Minister of Justice may from time to time prescribe a rate of interest for the purposes of subsection (1) by notice in the Gazette.

(3) …

2A. Interest on unliquidated debts.-(1) Subject to the provisions of this section the amount of every unliquidated debt as determined by a court of law, or an arbitrator or an arbitration tribunal or by agreement between the creditor and the debtor, shall bear interest as contemplated in section 1.”

(2) (a) Subject to any other agreement between the parties the interest contemplated in subsection (1) shall run from the date on which payment of the debt is claimed by the service on the debtor of a demand or summons, whichever date is the earlier.

(b) …”

[81] The Court a quo made an award of interest, quoted in para 14 above, from the date of judgment at Nedbank’s prime rate, which is substantially higher than the 15,5% per annum prescribed in terms of the latter Act. The award left both sides dissatisfied, not without justification, since it does reveal some confusion. If the award was one for mora interest there is no reason why, having regard to s 2A of the latter Act, interest should only run from date of judgment and not from date of summons (compare Adel Builders (Pty) Ltd v Thompson 2000 (4) SA 1027 (SCA) para 10). By the same token, and having regard to s 1(2), there is no reason why mora interest should be calculated at Nedbank’s prime rate instead of at the prescribed rate. It is therefore clear that the interest award as formulated by the Court a quo cannot stand. What order should it have made?

[82] The approach to a claim for interest was formulated by this Court in Bellairs v Hodnett and Another 1978 (1) SA 1109 (A) 1146H-1147C in these terms:

“As previously pointed out, mora interest in a case like the present constitutes a form of damages for breach of contract. The general principle in the assessment of such damages is that the sufferer by the breach should be placed in the position he would have occupied had the contract been performed, so far as this can be done by the payment of money and without undue hardship to the defaulting party. Accordingly, such damages only are awarded as flow naturally from the breach or as may reasonably be supposed to have been in the contemplation of the contracting parties as likely to result therefrom (Victoria Falls and Transvaal Power Co. Ltd. v. Consolidated Langlaagte Mines Ltd., 1915 A.D. 1 at p.22). In awarding mora interest to a creditor who has not received due payment of a monetary debt owed under contract, the Court seeks to place him in the position he would have occupied had due payment been made. The Court acts on the assumption that, had due payment been made, the capital sum would have been productively employed by the creditor during the period of mora and the interest consequently represents the damages flowing naturally from the breach of contract. The practice of awarding such interest at the legal rate of 6 per cent obviates the need to prove in every case what the capital sum would naturally and probably have earned had it thus been productively employed. A party wishing to recover a higher rate of interest would, in the absence of any alteration in this practice, have to establish by way of evidence as to current rates of interest on investment, etc. (such as was adduced in, for instance, the Enteka case, supra) that the loss naturally and probably suffered by him through the non-employment of his capital exceeded the accepted legal rate.”

[83] If TBA’s claim for interest is advanced as a claim for damages it must, as the law now stands, be accommodated under either limb (a) or limb (b) of the dictum in Holmdene Brickworks, supra, discussed in para 46 and following above. I do not believe that the additional interest for which TBA became liable to its banker can for the purpose of limb (a) be said to flow naturally from PW’s breach. Thinking away, as one must, the particular knowledge PW had that TBA was habitually operating on overdraft, it does not tend to follow as a matter of course that where an auditor is negligent his client will (a) operate on overdraft (b) apply the exact amounts stolen from him because of an oversight on the part of his auditor to the reduction thereof and (c) be unable factually to recover such monies from the thief.

[84] The next question is then whether TBA’s claim can be accommodated under limb (b). Here TBA is perhaps on firmer ground. PW’s working knowledge of TBA’s chronic overdraft is germane. Even so, and regardless of whether the “contemplation” or the “convention” approach is to be adopted, I do not believe that it can confidently be said that both parties were intent on that consequence. Be that as it may, there is in any event a more elementary reason why TBA’s claim for interest qua damages, be it general or special, cannot succeed and that is that it lacks the factual underpinning to support it. It is crucial to TBA’s case in this regard that it became liable to its banker for the additional interest now claimed. The only evidence proffered related to the actual rates of interest that were in fact charged by Nedbank. I agree with the submission made on behalf of PW that in order to substantiate its claim TBA at the very least had to prove: (a) the existence of the overdraft agreement between it and its banker; (b) the specific terms of that agreement, including the applicable terms as to the basis upon and the rate at which interest would be charged on the various levels of the overdraft; to which I would add (c) that had it not been for Mitchell’s thefts its overdraft would have been reduced by the exact amounts of the thefts (and would not, for instance, have been employed for another purpose such as an investment). TBA’s only witness on this issue, Bladergroen, testified that the terms of TBA’s overdraft agreement with Nedbank were embodied in a document which Bladergroen undertook to produce, but never did. The factual basis for the claim was accordingly lacking. It follows that TBA’s claim for interest qua damages cannot succeed (compare Bellairs v Hodnett and Another, supra, 1147C-H).

[85] That leaves the claim for mora interest on the residual basis. Our courts accept without requiring special proof that a party who has been deprived of the use of his capital for a period of time has suffered a loss. At the same time it is accepted that in the normal course of events such a party will be compensated for his loss by an award of mora interest (see e.g. Bellairs v Hodnett and Another, supra, 1145D-H). The first issue in this regard relates to the rate at which such interest should be calculated. PW contends for the rate prescribed pursuant to the provisions of s 1 of the latter Act, i.e.15,5% per annum. TBA’s opposing contention is that interest should, in the exercise of the court’s discretion provided for in s 1 of the latter Act, have been awarded at the higher rate charged by TBA’s banker on its overdraft. In my view TBA has failed to make out any case for the exercise of such a discretion in its favour. At best for TBA it is accordingly entitled to interest at 15,5% per annum.

[86] The only remaining issue regarding TBA’s claim for mora interest relates to the date from which such interest should be calculated. TBA’s contention is that the commencement date should be a date earlier than the date of summons because the quantum of its damages was readily ascertainable by PW at such earlier date. I disagree. In the first place the quantum was by no means capable of easy and ready proof and the fact that Reid reported on it cannot be held as an admission by PW against itself. In the second place it fails to recognise the fundamental principle that however liquidated a plaintiff’s claim for damages may be, mora interest can only be calculated from the date when mora commenced. This principle is formulated by Solomon JA in West Rand Estates Ltd v New Zealand Insurance Co Ltd 1926 AD 173 at 182:

“Here … the amount of loss in respect of each item of the claim was ascertained by agreement between the parties before issue of summons … It follows therefore that by our law interest began to run on the amount of defendant’s liability from the date of mora. And that brings me to consider the question of what that date is.”

and at 183:

“There is no satisfactory reason for following any other practice, and we think that we should now definitively lay down the rule that mora begins to run from the date of receipt of the letter of demand. It of course follows that, where there has been no letter of demand, there would be no mora until summons has been served on defendant.”

(see also Standard Chartered Bank of Canada v Nedperm Bank Ltd 1994 (4) SA 747 (A) 778A-B). It is idle to contend that PW was in mora on the very date each amount was stolen by Mitchell or for that matter on the date Reid submitted his report. For these reasons TBA should have been awarded interest on the capital amount of its claim, calculated at the rate of 15.5% per annum from date of summons, being 3 November 1995, to date of payment.

[87] Costs

The initial costs order by the Court a quo was made on 7 July 1999 and is quoted in para 14 above. It can be separated into two parts. The first part related to the 14 days of trial during which TBA set out to prove the quantum of its claim. With regard to this period of 14 days PW was ordered to pay TBA’s costs on the scale of attorney and client. The second part of the order provided that, save for the quantum-related period of 14 days dealt with in the first part, there would be no order as to costs, i.e. that each party was liable for its own costs.

[88] After 7 July 1999 it was disclosed to the Court that PW had made a “without prejudice” tender as contemplated in rule 34 on 14 April 1997 i.e. shortly before the commencement of the trial. This is referred to in para 15 above. Since the amount tendered exceeded the damages awarded in the first order PW asked the Court to reconsider its original costs order. The Court acceded to PW’s request and as a result the revised order which is quoted in para 15 above was substituted. In terms of the revised order PW was ordered to pay TBA’s costs incurred up to 24 April 1997 while TBA was ordered to pay PW’s subsequent costs, save for the costs incurred by PW during the stipulated quantum-related period of 14 days. With reference to this 14 days period PW was deprived of its costs.

[89] On the view I hold on the merits of the appeal the tender of 24 April 1997 is no longer of any consequence. I did not understand PW’s counsel to contend that in these circumstances this Court is to deviate from the general principle that costs should follow the result. Accordingly, PW is to pay the trial costs incurred by TBA.

[90] The only remaining issue regarding the costs in the Court a quo relates to the scale of the costs awarded to TBA during the quantum-related period of 14 days. Though both sides maintained that the Court a quo erred when, in its original order of 7 July 1997, it awarded those costs to TBA on a scale as between attorney and client, their respective contentions as to what the Court should have ordered differ diametrically. TBA contended that the Court should have awarded those costs not only on an attorney and client scale but on the even more punitive scale as between attorney and own client. PW’s contention, on the other hand, is that the special order of costs was unwarranted and that the costs should have been awarded on the ordinary party and party scale.

[91] From the judgment of the Court a quo it is apparent that it did not appreciate the import of TBA’s request for attorney and own client costs. What also appears from the judgment is that the reason for its special costs order was the finding that in all the circumstances PW acted “grossly unreasonably” in only conceding quantum after the matter was bitterly contested for fourteen days of the trial. In weighing up these opposing contentions it is unnecessary to dwell in detail on the trial court’s stated reasons for its finding of gross unreasonableness on PW’s part. Although it may be that I would not have made a similar order, sitting as a court of first instance, it is a discretionary decision with which this Court, in the absence of a misdirection, will be slow to interfere. PW’s case is not that the Court a quo misdirected itself. In all the circumstances there is in my view no basis for interfering with the Court a quo’s conclusion that a special costs order was justified.

[92] This brings me to TBA’s counter-submission that the costs pertaining to the quantum-related period of 14 days should have been awarded on the even more punitive scale of attorney and own client. Since the Court a quo appears not to have appreciated that TBA’s request was for an order in its refined form this Court is at liberty to reconsider the issue de novo. Having done so, I believe that the grounds advanced by TBA in support of its submission fall short of justifying a cost order which is even more punitive than the one already made in its favour. That being the case, and in the absence of full argument on the matter, it is not called for to express a firm view on whether an order in that form is a competent one. I may in passing mention that the Full Bench of the Cape Provincial Division has recently expressed the considered view, in Law Society of the Cape of Good Hope v Windvogel 1996 (1) SA 1171 (C), that an order for attorney and own client costs is not appropriate since it is not generically different from an order for attorney and client costs. On the other hand, there are considered decisions to the contrary in other Divisions (see e.g. Cambridge Plan AG v Cambridge Diet (Pty) Ltd and Others 1990 (2) SA 574 (T); Fidelity Bank v Three Women (Pty) Ltd and Others [1996] All SA 368 (W) and Ben McDonald Inc and Another v Rudolph and Another 1997 (4) SA 252 (T)). Moreover, in Sentrachem v Prinsloo 1997 (2) SA 1 (A) 22B-D and Cape Pacific Ltd v Lubner Controlling Investments (Pty) Ltd 1995 (4) SA 790 (A) 807C-D, this Court appears to have accepted in principle, but without pertinent consideration, that an order for attorney and own client costs would in appropriate circumstances be competent. This remains yet another issue for future consideration by this Court (see generally, Cilliers, Law of Costs, 3^rd ed para 4.08.)

The following order is made:

1. The appeal succeeds with costs including the costs of two counsel.

2. The following order is substituted for the orders made by the Court a quo:

‘1. The defendant is ordered to pay the sum of R1 389 801.90 to the plaintiff together with interest thereon at the rate of 15,5% per annum from 3 November 1995 to date of payment.

2. The plaintiff is ordered to pay to the defendant the sum of R74 100 together with interest thereon at the rate of 15,5% per annum from 6 February 1996 to date of payment.

3. The defendant is ordered to pay the plaintiff’s costs, including the costs of two counsel and the qualifying fees of the plaintiff’s expert witnesses, Wainer and Dorfan. Save for paragraph 4 below, the plaintiff is awarded such costs on a party and party scale.

4. The defendant is ordered to pay the plaintiff’s costs, including the costs of two counsel, of 14 days’ trial spent on quantum on the scale as between attorney and client.’

…………………

P M NIENABER

JUDGE OF APPEAL

MARAIS JA, FARLAM JA, BRAND AJA:

[1] With one reservation we concur in the judgment of Nienaber JA. The reservation relates to what has been referred to as the “supple” test in assessing what damages are recoverable. While the approach has attractions, we have not explored its possible disadvantages in sufficient depth to enable us to affirm its soundness. We too prefer to leave the question for another day. We have also had the advantage of reading the informative judgment of Olivier JA. Regrettably, we are unable to agree with his conclusion that the Act is applicable in this case.

[2] At the risk of stating what should be obvious we remind ourselves that the enquiry is what the legislature intended when it enacted the Act. The exercise is one of interpretation of a statute and does not involve “reforming” either the law of delict or the law of contract. Nor does it involve developing the common law in the manner contemplated by s 39(2) of the Constitution of the Republic of South Africa Act 108 of 1996. The manner in which the common law might now be reformed or developed in terms of that provision is logically irrelevant to the interpretation of the Act. What is relevant to its interpretation is what the common law was commonly understood to be in 1956 when the legislature decided to alter it by legislation. It is so of course that s 39(2) of the Constitution enjoins us “when interpreting any legislation” to “promote the spirit, purport and objects of the Bill of Rights”, but that postulates that the interpretation which it is proposed to place upon legislation is indeed one which would demonstrably promote an identifiable value enshrined in the Bill of Rights and also one of which the legislation is reasonably capable. To this aspect of the matter we shall return.

[3] We have no doubt that in 1956 the common understanding of the common law of South Africa was that contributory negligence (a well-known term of art evolved in and peculiar to the law of delict) was a concept alien to the law of contract. It is true of course that the negligence of a plaintiff was not always entirely irrelevant in a contractual claim. It might be regarded from a purely causative point of view as the only or true cause of the plaintiff’s loss or it might form the basis for a successful contention that the plaintiff had unreasonably failed to mitigate the loss. However, in such cases no question of contributory negligence and hence of apportionment of liability could arise. The plaintiff either succeeded or failed in the claim or, in the case of unreasonable failure to mitigate part of the loss, had the claim reduced to the extent that the loss could have been mitigated. But factual causation was the key concept, not respective degrees of culpa.

[4] Where, as here, the stated object of the Act in its long title is “To amend the law relating to contributory negligence and the law relating to the liability of persons jointly or severally liable in delict for the same damage, and to provide for matters incidental thereto”, one is immediately led to think that it was the law of delict which the legislature had in mind to amend. A fortiori when one recalls the dissatisfaction which existed at the time in regard to the common law principle in the law of delict which put a plaintiff entirely out of court if he or she was concurrently negligent, even if the degree of the defendant’s negligence was much greater. It led to considerable sophistry in the search for a last opportunity to avoid loss which could be said to have been available to the defendant in order to rescue the plaintiff from being non-suited because of his or her contributory negligence. But even where such an opportunity could be identified, it was generally thought to be unfair that the plaintiff’s negligence should then be entirely ignored and the defendant be held liable for the full amount of the loss. There was thus a notorious mischief in the common law of delict to which the Act would be expected to be intended to put an end. No comparable state of public and professional disaffection with the extent to which the law of contract catered for negligence on the part of a plaintiff existed and, in the absence of a clear indication in the Act that it was intended to amend the law of contract in that respect, one would hesitate to conclude that it did.

[5] As one reads on, one finds provisions which plainly show that the law of contributory negligence in the common law of delict was being amended but precious little, if anything, to show that negligence in the law of contract was also being addressed. Indeed, one finds indications to the contrary. One need waste no time on the former. It has never been doubted that the Act was intended to amend, at least, the law of delict in so far as it related to contributory negligence. But what of the latter?

[6] First there is the choice of language. We have already observed that contributory negligence is a term of art which had its fons et origo in the law of delict. It is a concept entirely foreign to the law of contract. Yet that is the expression which the legislature has employed in the long title of the Act, as the heading to Chapter I, and in s 1(3).

[7] Next there is s 1(1)(b). It provides that for the purposes of paragraph (a) of s 1(1), damage shall be regarded as having been caused by a person’s fault notwithstanding the fact that another person had an opportunity of avoiding the consequences thereof and negligently failed to do so. It was obviously intended to abolish the last opportunity rule which was also a common law rule associated with the law of delict.

[8] Then there are the consequences of accepting that the Act was intended to apply to contracts. The plaintiff sues the defendant for damages for breach of contract. No negligence is entailed in the breach; the obligation is absolute and fault is irrelevant. As a fact the plaintiff’s negligent conduct has contributed to the happening of the loss-causing event. There being no damage caused partly by the defendant’s fault, s 1 (1)(a) cannot apply and the defendant cannot claim an apportionment of liability. If, on the other hand, negligence was entailed in the breach of contract, the defendant would be entitled to an apportionment of liability. The inequity is manifest: the defendant who is at fault may invoke the plaintiff’s negligence to reduce the claim; the defendant who is not at fault, may not. That bizarre result is not satisfactorily explained by saying that the contractual obligation of the latter was “absolute” whereas that of the former was relative and dependent upon the existence of negligence. The former was under just as absolute an obligation not to be negligent. Why would he or she have been allowed to invoke the Act after breaching the contract by being negligent, but the latter not be entitled to do so after breaching the contract in a manner which does not entail negligence? The distinction seems absurd. What is more, there is a clear distinction between a contractually imposed obligation to take care in discharging a stated obligation and a contractually imposed obligation to perform a stated obligation failure to perform which will constitute a breach of contract whether or not the breach was in fact due to lack of care. In the case of the latter, is the defendant who is blameless in breaching the contract to be liable in full but the defendant who is negligent in breaching the contract to have the benefit of the Act? It is difficult to accept that, if the legislature did intend the Act to apply to contracts generally or even to only some contracts, it would have enacted so blunt an instrument and left fundamental questions such as these unanswered.

[9] In our view, the random and inconsistent results of the indiscriminate application of the Act to contracts negative the existence of any intention on the part of the legislature to have the Act apply to contracts, whether or not they are contracts which require to be performed in a manner which is not negligent.

[10] There are still further indications that such is the correct conclusion. Section 1(1)(a) provides that a claim “shall not be defeated by reason of the fault of the claimant”. In the law of delict that was of course something which could and did happen. A claim valid in all respects could be defeated by the plaintiff’s contributory negligence. In the law of contract it could not. The negligence of a plaintiff could not “defeat” his claim. The point was made by Watermeyer J in OK Bazaars (1929) Ltd and Others v Stern and Ekermans 1976 (2) SA 521 (C) at 528F. If his own negligence was held to be the true or real cause of his loss and he was non-suited on that account it was implicit that there never was a justifiable claim against the defendant. If he negligently failed to mitigate his loss, that too did not “defeat” his claim. It disabled him from pursuing a valid claim if the entire loss could have been avoided and merely reduced it if part of the loss could have been avoided.

[11] Section 1(3) provides that “ ‘fault’ includes any act or omission which would, but for the provisions of the section, have given rise to the defence of contributory negligence”. Here again it is necessary to repeat that the law of contract knew no “defence of contributory negligence”. Moreover, “fault” must obviously be confined to negligence. The context of the Act shows that to be so. Dolus is a form of fault in the wide sense but it is obviously not included. The legislature did not exclude it by name because the context of the Act showed plainly enough that it was to be excluded. If it be suggested that fault is always involved in a breach of contract and therefore contracts are covered by the Act the suggestion would be wrong. Contracts may be breached in circumstances where no fault can be identified. If it be said that at least those contracts breach of which entails fault are covered, the questions raised in para [8] arise. All these considerations point inexorably to the conclusion that the law of contract was far from the mind of the legislature when it enacted the Act and that it did not intend the Act to amend the law of contract.

[12] We cannot agree with the approach of the Court a quo to the interpretation of the Act. It entailed isolating s 1(1)(a) and attempting to accommodate contractual claims within what was said to be the plain language of the provision. Such an approach ignores the colour given to the language by the context of the Act read as a whole and by the long title, the use of the expression contributory negligence, and the other considerations raised in this judgment. The days are long past when blinkered peering at an isolated provision in a statute was thought to be the only legitimate technique in interpreting it if it seemed on the face of it to have a readily discernible meaning. As was said in University of Cape Town v Cape Bar Council 1986 (4) SA 903 (A) at 914D-E “I am of the opinion that the words of s 3(2)(d) of the Act, clear and unambiguous as they may appear to be on the face thereof, should be read in the light of the subject-matter with which they are concerned, and that it is only when that is done that one can arrive at the true intention of the Legislature”. The well-known passage in the dissenting judgment of Schreiner JA in Jaga v Dönges NO and Another; Bhana v Dönges NO and Another 1950 (4) SA 653 (A) at 662G-663A was also quoted with approval. It is of course clear that the context to which reference is made in the latter case must include the long title and chapter headings. (Cf Swart v Cape Fabrix (Pty) Ltd 1979 (1) SA 195 (A) at 202C.)

[13] The decision of Watermeyer J and Steyn J in the OK Bazaars case, supra, that the Act does not apply to claims in contract has stood for 26 years. Its correctness has not been challenged in our courts. The legislature has amended the Act on three occasions since the decision and, if the decision did indeed frustrate its desire to amend the law of contract, we find it very strange that remedial legislative steps were not taken on any of those three occasions. It is so that it was not a decision of this Court which would have bound every court in the land but it was a decision of two judges in a provincial division which would bind single judges in the Cape Provincial Division and magistrates throughout the country. That the legislature has acquiesced in the decision for nigh on 30 years is significant. The remarks made in Kergeulen Sealing and Whaling Co Ltd v CIR 1939 AD 487 at 505 also have some bearing on the matter. In that case the decision of two judges of the Cape Provincial Division in a prior case which had remained unchallenged for over 20 years and on which the commercial community had presumably acted on the assumption of its correctness was described as one which “one would hesitate now to disturb”. Such a consideration can obviously not be conclusive. Its weight will depend upon the circumstances of the case and the extent to which existing legal relations affecting many members of society may be retrospectively nullified by disturbing the commonly accepted interpretation. We say “retrospectively” because that would be the effect of a decision by a court reversing the previously held view. It has never been suggested that South African courts have the power to limit the operation of such judge-wrought changes in the law. (See Hahlo & Kahn, The South African Legal System and its Background, p 145 n 13 and pp 249-50.) We should add that, in any event, we consider the reasoning of Watermeyer J to be sound.

[14] We do not find it helpful to examine how courts elsewhere have interpreted their own domestic legislation dealing with contributory negligence. Their legislation falls to be interpreted against the background of a common law regime which, while similar in important respects to our own, is or may be in other respects very different. The language of the legislation is also not identical to our own. None the less and because the legislation elsewhere clearly influenced ours and the interpretation of it might be thought to be persuasive we shall indicate briefly why we do not find it to be so.

[15] We do not think that the difference in wording between the English definition of “fault” in section 4 of the Law Reform (Contributory Negligence) Act, 1945 8 & 9 Geo. 6, c.28 and the definition of “fault” in s 1(3) of our Act indicates an intention on the part of the South African legislature to make the Act apply in contractual cases. The English definition of “fault” is:

“ ‘fault’ means negligence, breach of statutory duty or other act or omission which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence.”

Our definition is in s 1(3) of the Act and reads as follows:

“(3) For the purpose of this section ‘fault’ includes any act or omission which would, but for the provisions of this section have given rise to the defence of contributory negligence.”

[16] In English law in a case where both a plaintiff employee and a defendant employer are not negligent but are in breach of their statutory duties under the factories legislation and the regulations made thereunder the damages suffered by the plaintiff will be subject to apportionment under the 1945 Act because the employee’s non-negligent breach of statutory duty is “fault” specifically covered by the definition in s 4 of that Act: see Boyle v Kodak Ltd [1969] 1 WLR 661 (HL). In that case Lord Diplock pointed out (at 672B) that a new branch of the law of civil wrongs was developed in England by judicial decisions based on the Factories Act (9 & 10 Eliz 2, c. 34) and its predecessors and by regulations made thereunder.

[17] Although our law recognizes an action for damages for breach of a statutory duty where the statute was intended to give a right of action (see McKerron, The Law of Delict, 7^th edition, p 276), where it does not the courts may yet hold that the breach may be evidence of negligence. Compare Rawles v Barnard 1936 CPD 74 at 77 and Olitzki Property Holdings v State Tender Board and the Premier of the Province of Gauteng (SCA, 28.3.2001, as yet unreported, at para [13]).) Our courts have not by judicial decision built up a new branch of the law of civil wrongs relating to breach of statutory duties imposed by legislation akin to the English Factories Acts. It may well be that for that reason our legislature decided to omit from the definition of “fault” in the Act a reference to breach of statutory duty such as was found in the English 1945 Act. Furthermore, liability for delict in our law is based in general on fault, unlike in English law where in an appreciable number of torts strict liability exists. That would explain why our legislature omitted any reference to other acts or omissions which give rise to strict liability and was content to make apportionment available only in cases where fault in its ordinary sense was present, subject only to the inclusion of contributory negligence for the reason given in the next paragraph. Moreover, as Lord Diplock said (at 674H) in Boyle v Kodak Ltd, supra, it is difficult to apportion the respective shares of responsibility for damage of parties who were not blameworthy in any way and who are only regarded as being at “fault” because of the application of strict liability to their case. We therefore do not think that the differences between the definitions of “fault” appearing in the English and South African Acts indicate an intention on the part of our legislature to make the Act apply not only in the context of actions ex delicto but also of those ex contractu.

[18] The fact that our definition is introduced by the word “includes” and not “means”, as is the English definition, is explicable in our view on the simple ground that because a plaintiff is said to be guilty of contributory negligence where he is careless in safeguarding his own person or property, even if his carelessness puts no-one else’s person or property at risk, it was considered advisable to make it clear that the blameworthy conduct of the plaintiff which was to form the basis of the apportionment was to include contributory negligence in this sense. Aquarius (Watermeyer CJ) in his article Causation and Legal Responsibility (1941) 58 SALJ 232 at 248 made this aspect of contributory negligence clear when he said:

“It is negligence in the sense of a failure to look after his own interests, and not necessarily negligence in the sense of a breach of a duty to take care which is owed to another.”

[19] In para 11 (a) and again in para 12 of his judgment Olivier JA states that, by virtue of the decision of this Court in Lillicrap, the approach followed in England and New Zealand is not open to us. We have difficulty in seeing the relevance of the decision in Lillicrap to the solution of the problem. The approach followed in England and New Zealand involves drawing a distinction between three categories of breach of contract and an acceptance of the proposition that their Acts only apply to the third category, being the category of concurrent contractual and delictual liability. This approach is dictated by the definition of fault in their Acts, more particularly, the requirement in the definition that the defendant’s conduct must give rise to a liability in tort. It follows that if the defendant is only liable in contract and not in tort there is no “fault” on the part of the defendant and the English Act cannot apply. That is the very essence of the Glanville Williams theory. However, as is emphasised by Olivier JA in para 5 (a) and again in 9 (f), the expression “which gives rise to liability in tort” does not form part of the definition of “fault” in our Act. The existence of concurrent liability in delict and contract therefore appears to be irrelevant when construing our Act.

[20] However, even if it were open to us to adopt the approach of the English courts and accept that the Act could apply to what was identified as the third category of case in Forsikringsaktieselskapet Vesta v Butcher [1986] 2 All ER 488, namely, where concurrent liability in both contract and delict exists, this is not such a case. It falls within the second category of case identified by Hobhouse J at 508 (and approved by the Court of Appeal: see [1989] AC 852 at 865D-E, 867F-G and 875F-G) in which an apportionment on the basis of contributory negligence is not available: see further Raflatac Ltd v Eade [1999] 1 Ll R 506 (QB) at 510.

[21] The decision of the Court of Appeal in Vesta was given by a majority (O’Connor LJ and Neill LJ). The third member of the court, Sir Roger Ormrod, dissented, saying (at 879 A-B) that he remained unconvinced that “contributory negligence, as such, at common law had any relevance in a claim in contract”.

[22] On the other hand one of the factors mentioned by O’Connor LJ (at 867 F-G) in favour of the view that there is a power to apportion in a category (3) case even though the claim is made in contract was “that contributory negligence was a defence in category (3) cases pleaded in contract before 1945”. He continued: “The argument is supported by railway cases and banking cases.”

[23] The majority in the High Court of Australia in Astley v Austrust Ltd (1999) 197 CLR 1 based their decision in part on a finding that prior to 1945 contributory negligence did not operate as a defence to a breach of contract. Further support for that view is to be found in an article by N E Palmer and P J Davies, “Contributory Negligence and Breach of Contract – English and Australian Attitudes Compared” published in (1980) 29 ICLQ 415, who state (at 418-9) that contributory negligence was never a defence to an action for breach of contract at common law and refer to a decision of the Court of Appeal, Becker v Medd (1897) 33 TLR 313, in which it was specifically held that a claim in contract could not be defeated by proof of negligence on the plaintiff’s part.

[24] The point need not detain us further in this case because, whatever the position was in English law, it was not suggested that contributory negligence by the plaintiff was a defence to a contractual claim in our law.

[25] During the course of the argument counsel for the respondents referred us to the texts of draft bills which were published in the Government Gazette before the Act was passed by Parliament. These texts showed, so it was submitted, that although it was originally proposed to limit the operation of the Act to delictual claims this intention was departed from in the text which was eventually passed by the Legislature. Counsel for the appellant responded to this material by placing before us the text of the Hansard report of the proceedings in the House of Assembly which clearly showed, particularly from the speech of the Minister of Justice, who introduced the Second Reading of the Bill, that the discussions related solely to delictual claims. In view of the fact that we have without reference to this material come to the conclusion that the Act only applies to delictual claims it is unnecessary for us to decide whether material of this kind can be looked at by a court when legislation falls to be interpreted and, in particular, whether the decision of the House of Lords in Pepper v Hart [1993] AC 593 (HL) is in accordance with our law.

[26] It remains to observe that we are unable to discern in the Bill of Rights any societal value which is imperilled by the conclusion that the Act does not apply to claims based on breach of contract and that we do not consider the Act to be reasonably capable of a contrary interpretation in the light of all the indicia to the contrary which exist.

[27] We conclude, therefore, that there can be no reduction of the damages proved to have been suffered by the appellant. Whether there were other ways at common law in which the respondent could have exploited the negligence of the appellant does not fall to be considered. Those that were suggested and covered by the pleadings have been dealt with in the judgment of Nienaber JA.

[28] By drawing attention to some of the implications of boldly applying the Act to cases in contract (even if only to those where a breach entails negligence), we do not wish to be thought to be hostile to the very idea of extending the operation of the Act to contract cases by legislation. All that we would caution against is a decision to do so without a full appreciation and consideration of all its implications. These matters are, so we understand, being considered by the Law Commission and it will doubtless take into account all the relevant implications, including those touched upon in the case of Austrust, supra, at paras 47 and 48.

R M MARAIS

JUDGE OF APPEAL

I G FARLAM

JUDGE OF APPEAL

Olivier JA

[1] I am in respectful agreement with the judgment of my colleague, Nienaber JA, that there was a failure by both TBA and PW to prevent the loss suffered by TBA by the exercise of due care. I also agree with his approach to the questions of interest and costs. Unfortunately, I disagree with my learned colleagues that s 1 (1) (a) of the Apportionment of Damages Act, 34 of 1956 ('the Act') is not applicable in the present case. I readily concede that the question whether the Act is applicable to contractual claims is controversial. In the end the opposing judicial views may well depend on differing philosophical and jurisprudential points of departure.

A minority judgment is usually short and to the point. In the present case I consider it necessary to explain my views somewhat more fully, also because the Act is presently under review by the South African Law Commission and a somewhat more complete overview may be helpful to it.

[2] While it is undisputed that the Act applies to claims based on delict, the question of the applicability of the Act and its counterparts to contractual clauses has elicited strongly opposed views and judgments in England, Australia, New Zealand and Canada and in our country. All the Commonwealth countries just mentioned share apportionment legislation. But the wording of the various acts differ; so do the lenses through which we look at the statutes. In particular, South African lawyers are required to keep our common law - the Roman-Dutch law - in mind as a background factor in interpreting our legislation. What is important, though, is the quest for recognition of the underlying principles and philosophies developed by other courts as an aid in clarifying our own thoughts. I will first examine the position in our common law and, for the sake of clarity, distinguish between delictual and contractual claims.

[3] Apportionment in delictual actions

(a) Originally, in Roman law, due to the procedural formula applicable to the Aquilian action, a strict all-or-nothing approach prevailed. Reinhard Zimmermann, The Law of Obligations : Roman Foundations of the Civilian Tradition, (1990), at 1010 concludes :

'If somebody suffered harm through his own fault, he was denied recovery, unless the tortfeasor had acted intentionally (in which case he could recover his full damages). The strict principle of all-or-nothing was predetermined by the procedural formula. The judge only had the alternative to condemn in the full amount or to absolve the defendant - tertium non datur.'

This principle appears from two texts in the Digesta dealing with general principles applicable to the Aquilian action, D 9.2.9.4 in fine (Ulpianus) and D 9.2.31 (Paulus). But two peculiar cases have elicited involved debates through the ages. The first is the case of the athlete who is training at javelin throwing, discussed in D.9.2.9.4. A slave is passing by and is injured by the javelin. If this happens on a proper sports field, the athlete is not held liable; if outside a recognised sports field, he is liable. Is this a case of contributory negligence? Or, rather, volenti non fit injuria? The second case is that of the barber who sets up his chair in the immediate vicinity of a playing field. While shaving a slave, the barber's hand is hit by a ball thrown or kicked by one of the players. The slave is injured. The text (D.9.2.11 pr) mentions three opinions, none of which applies an apportionment of damages. Mela says the one who is negligent, is liable. But who is negligent : the player or the barber? Mela does not say. Proculus thinks the barber is negligent for setting up his chair in a dangerous place. Ulpian states that it is rightly said that the slave only had himself to blame, because

' … it is no bad point in reply that if someone entrusts himself to a barber who has his chair in a dangerous place he has only himself to blame for his own misfortune.'

(See Zimmermann op cit 1011 - 1013 for a discussion of the texts.) We do not know how the Romans actually solved the problem. Mela and Proculus clearly think that the answer lay in the field of culpa. Does Ulpian invoke the volenti defence? What emerges is that the problem of concurrent causation of the loss by a plaintiff and a defendant was not solved by what Flemming, Torts, p 244 calls 'the abracadabra of causation', but by having regard to fault or wrongfulness.

(b) There is, however, another text which later assumed more importance than the discussions by the classical scholars of the cases of the javelin thrower and the barber. It is D.50.17.203, a text ascribed to Pomponius and which appears in the 50^th book of the Digest, dealing with general rules and principles. The text lays down :

'If anyone incurs loss which is his own fault, he is not regarded as incurring loss'

(c) This principle was used by medieval lawyers to begin to develop a theory applicable to cases of concurrence of fault in the field of delict. Zimmermann op cit 1030, citing Lauterbach (1618 - 1678) and a gloss to D.9.2.9.4, explains that the approach of the Roman law was retained, but it was now more clearly explained in terms of fault:

'The fault of the plaintiff / victim was, in a way, "set off" against that of the defendant/wrongdoer, with the result that "culpa culpam abolet". Hence the expression of compensatio culpae or culpa compensation that came to be used to label the uncompromising approach to the problem of contributory negligence. Whether every contributory fault on the part of the victim - even culpa levissima - was originally taken to deprive him of his remedy is not quite clear. In the later usus modernus, at any rate, the issue appears to have been decided on the basis of a preponderance of fault; only if he had displayed the same or a greater degree of negligence than the wrongdoer did the victim lose his claim. Where, on the other hand, his negligence was less significant, when compared with that of the wrongdoer, his claim for damages remained completely unaffected.' (My emphasis)

(d) On the continent of Europe, the Romanistic principle of D.50.17.203 and the idea of culpa compensatio, as described above, prevailed ( see Zimmermann op cit 1047).

(e) The South African law of delict, cut off from its historical roots

in the 19^th and early 20^th centuries

' … became completely entrapped in the "abracadabra" of the causal approach to contributory negligence. Ultimately, therefore, only the legislator was able to save the day.' (Zimmermann op cit 1049)

Our courts simply adopted the English law according to which the contributory negligence of a plaintiff was a complete bar to relief in an action in tort, rather than the relative fault principle of our common law. This was lamented by Watermeyer CJ in Pierce v Hau Mon 1944 AD 175 at 195 :

'The law relating to the subject of contributory negligence which is applied by our Courts has been taken over from English law and it is seldom that any Roman-Dutch authority is referred to. In fact there is plenty of authority in Roman law (see Grueber, Lex Aquilia (2.7.4, p. 228 et seq.) and also in Roman-Dutch law (see Voet (9.2.17; 9.2.22)), and the principle of culpa compensatio was referred to by De Villiers, C.J., in Lennon's case (1914, A.D. 1), by Kotze, J.A., in Jacobs v Union Government (1919, A.D. 325) and by Gardiner, A.J.A., in the case of Union Government v Lee (1927, A.D. 202). It may be that if Roman - Dutch authorities had been more fully referred to in earlier South African cases that our law of contributory negligence might have developed on different lines from the English law. However, if we take the English law on the subject as it now is, and as it had been adopted in our Courts, we shall find that there are still doubts and difficulties about its application in certain classes of cases.'

(f) Our courts, not enchanted with the 'all-or-nothing' approach, adopted the 'last opportunity' rule : the party who had the last opportunity to avoid the loss is liable. This was a manifestation of the proximate cause theory of causation.

(g) In the Commonwealth countries, in the field of torts, the law in respect of contributory negligence was changed by legislation first in 1924 in Ontario, later in England by the Law Reform (Contributory Negligence) Act of 1945 and in other Commonwealth countries. Some of these statutes were specifically designed and phrased to be applicable to delictual claims only; others were worded in general terms. Our legislator followed suit in 1956 with the Act which in its terms differs from that of the other Commonwealth statutes. Overall the Act was not well-drafted - see the scathing remarks of Holmes J in Taylor v South African Railways and Harbours 1958 (1) SA 139 (D) at 142 A - B. Since 1956 our courts have developed a substantial body of jurisprudence as regards the application of the Act to delictual claims. In that field our law has now, broadly speaking, stabilised in a system which seems to be equitable.

[4] Apportionment in contractual actions

(a) We must now return to the field of contract and ascertain how the phenomenon of concurrent fault was dealt with in our common law. The subject is a difficult one, mainly because of the absence of clear texts or well-developed rules.

(b) Culpa played a significant role in the Roman and Roman-Dutch law relating to breach of contract.

In post-classical Roman law all claims for breach of contract were given content ex aequo et bono. From now on, through medieval law, usus modernus and Roman-Dutch law :

'What mattered was simply whether the debtor had complied with his contractual obligations and, if not, whether his failure to perform (properly) was attributable to his fault; hence the emphasis throughout the various periods of the ius commune on the subjective requirements for liability for breach of contract and the attempts to analyse, refine and systematize the various degrees of culpa (in the broad sense of the word).' (Zimmermann op cit 807 et seq; Ramsden Supervening Impossibility of Performance in the South African law of Contract. 1985 19 et seq).

(c) The important point is that the basic requirements for contractual and delictual liability in our common law did not differ fundamentally. Both kinds of liability were based, in the end, on culpa. The incidence of onus may have been different, and the quantum of damages may have been different, but there was a basic unitary approach. No wonder that, as Zimmermann (op cit at 808 footnote 176) points out, during the time of the usus modernus liability arising as a consequence of deficient performance 'tended to be based on the lex Aquilia rather than on contractual principles.'

(d) I must pause here to refer to and emphasise the fundamental difference between our common law roots and that of the English law in relation to the role played by culpa as a requirement for an action on contract. Zimmermann op cit 814 puts it as follows :

'Contrary to the tradition of the ius commune, the debtor's liability [in English law] does not depend on fault. The reason is, of course, that the common law regards all contractual promises as guarantees:

"[W]hen [a] party by his own contract creates a duty or charge upon himself, he is bound to make it good, … notwithstanding any accident by inevitable necessity." '

The harsh and uncompromising rule of English contract law led to the creation of fictional 'implied' terms and 'implied' conditions to assist the debtor. But it has also led to the view that as fault is not relevant in contract cases, the principle of apportionment could not become relevant - and this explains, in my view, the omission in the English Act of 1945 of a reference to actions based on contract.

(e) The question remains : how did our common law deal with cases where the plaintiff, suing on contract, was also at fault in respect of the loss suffered by him? Did the principle of culpa compensatio or the last

opportunity rule or apportionment of liability apply? We simply do not know. The old authorities do not give us any guidance. Neither our old writers nor our early reported cases are helpful in this respect. There simply is no authority for the proposition that the contributory negligence of a plaintiff who sues in contract was ever considered to be relevant; nor for the opposite point of view. The principle of our common law that both delictual and contractual liability depend in various ways on culpa was never expressly rejected; nor the common law principle, recognised so clearly in delictual claims, that the greater culpa of the plaintiff neutralises the lesser culpa of the defendant. We could find no pertinent authority for the proposition that the principles relating to delict applied did not apply to contractual claims.

[5] The legislation

I now turn to the legislative intrusion, especially in England and in our country in respect of contributory negligence in delict and contract.

(a) In England, the question of apportionment of damages is dealt with in the Law Reform (Contributory Negligence) Act 1945. S 1 reads as follows :

‘(i) Where any person suffers damage as a result partly of his own fault and partly of the fault of any other person or persons, a claim in respect of that damage shall not be defeated by reason of the fault of the person suffering the damage, but the damages recoverable in respect thereof shall be reduced to such extent as the court thinks just and equitable having regard to the claimant's share in the responsibility for the damage :

Provided that -

• this subsection shall not operate to defeat any defence arising under a contract;

• where any contract or enactment providing for the limitation of liability is applicable to the claim, the amount of damages recoverable by the claimant by virtue of this subsection shall not exceed the maximum limit so applicable.’

S 4 of the English Act contained a definition of fault, viz

‘ “fault” means negligence, breach of a statutory duty or other act or omission which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence.’(My emphasis)

In South Africa, the Apportionment of Damages Act 34 of 1956 ("the Act") came into operation on 1 June 1956. Its long title states that its purpose is

'To amend the law relating to contributory negligence and the law relating to the liability of persons jointly and severally liable in delict for the same damage, and to provide for matters incidental thereto.'

Chapter 1 bears the title 'Contributory Negligence' and Chapter 2 the title 'Joint or Several Wrongdoers'. For our purposes Chapter 1 (consisting of only one section) is of direct and immediate importance; Chapter 2 becomes relevant only insofar as it may be said to illuminate the ambit and scope of Chapter 1.

(b) Chapter 1 reads as follows :

‘1. Apportionment of liability in case of contributory negligence. -

(1) (a) Where any person suffers damage which is caused partly by his own fault and partly by the fault of any other person, a claim in respect of that damage shall not be defeated by reason of the fault of the claimant but the damages recoverable in respect thereof shall be reduced by the court to such extent as the court may deem just and equitable having regard to the degree in which the claimant was at fault in relation to the damage.

(b) Damage shall for the purpose of paragraph (a) be regarded as having been caused by a person’s fault notwithstanding the fact that another person had an opportunity of avoiding the consequences thereof and negligently failed to do so.

(2) Where in any case to which the provisions of sub-section (1) apply, one of the persons at fault avoids liability to any claimant by pleading and proving that the time within which proceedings should have been instituted or notice should have been given in connection with such proceedings in terms of any law, has been exceeded, such person shall not by virtue of the provisions of the said sub-section, be entitled to recover damages from that claimant.

(3) For the purposes of this section ‘fault' includes any act or omission which would, but for the provisions of this section, have given rise to the defence of contributory negligence’

(c) Is Chapter 1 applicable to cases where the plaintiff who sues for damages caused by a breach of contract by the defendant is himself causally negligent in respect of the said damage?

[6] South African case law

In the past, two Provincial Divisions of what is now the High Court have said 'no'. This is the first occasion that the question has come before this Court. The Provincial decisions are those in Barclays Bank D.C.O. v Shaw 1965 (2) SA 93 (O) (‘Shaw’) and O.K. Bazaars (1929) Ltd. and Others v Stern and Ekermans 1976 (2) SA 521 (C) (‘OK Bazaars’).

(a) In Shaw the plaintiff sued his bank on contract for damages in the sum of R999. The plaintiff had issued a bearer cheque for R1, which amount was unlawfully altered by a bearer to R1000. The latter amount was negligently paid out by the bank. The defendant argued that the plaintiff was also negligent and that s 1 of the Act should be applied. The court (quite wrongly) held that the plaintiff’s claim was not one for damages. It was said (at 99 E), clearly obiter and without any researches into the authorities, that the Act was historically not intended to apply to claims based on contract.

(b) In the second case mentioned above, O K Bazaars sued the defendant, a firm of land surveyors, for damages for breach of a contract to survey a property in order to determine the correct boundaries and site limits thereof, and to prepare an up-to-date site diagram. The plaintiff alleged that the contract incorporated an implied term, requiring the defendant to exercise reasonable skill in the performance of its obligations. It then alleged that the defendant, in breach of its obligations, failed to set out the correct boundaries, etc, as a consequence of which the plaintiff suffered the damage alleged. In its plea the defendant alleged, inter alia, that the plaintiff was partly at fault in relation to the occurrence of the damage, it being negligent in a number of respects, set out by the defendant. It relied on s 1 of the Act, claiming an apportionment. The plaintiff excepted to this part of the plea on the basis that the Act deals only with delictual claims and not with

claims based on breach of contract.

(c) The Court held that the plaintiff’s claim was based solely on breach of contract (at 525 H). The Court (per Watermeyer J; Steyn J concurring) held that s 1 of the Act does not apply to such a claim, for the following reasons :

1 It was argued that the word ‘fault’ in s 1 of the Act, in so far as it refers to the defendant, was wide enough to include a breach of contract (at 528 A - B). But, held Watermeyer J, (at 528 B) fault normally connotes a degree of blameworthiness; a contract can be breached by a party through no fault of his own. If s 1 is then construed as covering claims based upon breach of contract, should it be held to apply to certain breaches of contract only and not to others?

2 The history of the Act shows that it was intended to apply to delictual actions only. Prior to the passage of the Act, contributory negligence on the part of the plaintiff had the effect of completely defeating his or her claim. To alleviate this harsh consequence the ‘last opportunity’ rule was developed, but even this was not satisfactory. Chapter 1 of the Act was designed to overcome this state of affairs (at 528 C - E).

3 The aforesaid object of the Legislature seems to be borne out by the words ‘shall not be defeated by reason of the fault of the claimant’ in s 1 (1). Watermeyer J said :

‘Although in a claim based upon breach of contract negligence on the part of the plaintiff might be relevant in determining whether or not the damages claimed flowed from the defendant’s breach, it would not be apposite to say that such negligence (fault) “defeated” the plaintif’s claim. The plaintiff’s claim would fail because he did not show that the damages flowed from the breach.’(at 528 F)



The learned judge also stated that whilst this reasoning may not be entirely conclusive, it seems to be far more likely that the Legislature had in mind the well-known defence of contributory negligence to a delictual claim.

4 A further indication is that contributory negligence was not normally one of the recognised defences to a claim based upon a breach of contract (at 528 H).

5 The meaning of s 1, if it is ambiguous, has then to be found by applying the canons of construction, which all indicate that the section is not applicable to actions based upon breach of contract, inter alia, that the legislature knows the existing state of the law; that an ambiguous statute should be interpreted in such a way as to conform to the existing law, and that in cases of obscurity the long title may be looked to. The learned judge remarked that the long title makes it clear that the Act is one to amend the law relating to contributory negligence (at 529 A).

6 Inasmuch as prior to the passing of the Act contributory negligence was not one of the recognised common law defences to a claim based upon a breach of contract it seems unlikely that, had the legislature intended to introduce a radical change in the law, it would have done so in an oblique way and without using clear language to express such an intention (at 529 F - G).

7 An alternative argument was raised by the defendant. It was that even if s 1 of the Act did not apply to all claims for breach of contract, then it should at least be construed as covering claims for breaches of contract which import a duty not to be negligent (at 529 G - H). Counsel for the defendant relied on a number of English cases, viz Sayers v Harlow Urban District Council (1958) 2 All E. R. 344 ; Quinn v Burch Bros. (Builders) Ltd. (1965) 3 All E.R. 801, (1966) 2 All E.R. 283 and De Meza and Stuart v Apple, Van Straten, Shena and Stone (1974) 1 Lloyd’s Law Reports 508 (Q.B.). Watermeyer J held that the first case mentioned above appears to have been brought in tort, the second was decided on the basis of causation, and the last was unconvincing. Apart from these considerations, Watermeyer J held that the English common law is not the same as ours and that there are material differences between the English Act and our Act. The alternative was thus also rejected.

(d) The present state of our case law is that laid down in O K Bazaars: the Act is not applicable to contractual claims, not even where the contract imports a duty not to be negligent. Where the plaintiff has clearly elected to sue in contract, ‘ ... it does not lie in the defendants mouth to say that the defendant might also have been liable to the plaintiff in delict.’ (per Watermeyer J at 527 A).

(e) Counsel for the defendant in this Court, PW, invited us to revisit the question now under consideration and to subject O K Bazaars to close scrutiny. They asked this Court to consider the latest trends in England and New Zealand, especially in view of the fact that the question of apportionment of damages is dealt with in these two countries in legislation largely similar to our Act. Counsel contended that in those two countries an apportionment can take place, in certain circumstances, even where the claim is based on a breach of contract. Counsel for the plaintiff, on the other hand, referred to the Australian law, where, they contended, an apportionment cannot take place if the claim is based on a breach of contract.

[7] Comparative Law : England

(a) Conflicting views were expressed by English judges as to whether the Act of 1945 was applicable to claims based on breach of contract. Eventually, the matter came before the Court of Appeal in Forsikringsaktieselskapet Vesta v Butcher [1989] 1 AC 852 (CA) (‘Vesta v Butcher'). The matter came on appeal from a judgment by Hobhouse J. For our purpose, it is sufficient to state that the plaintiff, Vesta, having correctly settled a claim against it, instituted action against the first defendant, an underwriter, for indemnification by virtue of a policy of reinsurance. In the alternative, Vesta claimed damages against the second and third defendants (insurance brokers), alleging that they failed to obtain a valid contract of reinsurance and furthermore that they failed to inform the first defendant that the insured could not comply with a clause requiring it to provide a 24-hour watch over its operations, thus allowing the first defendant to escape liability. The second and third defendants denied liability. Hobhouse J found that the third defendants were in breach of their duty towards the plaintiff, but that the plaintiff was contributorily negligent in not making sure that the matter of the 24-hour watch problem had been solved by the third defendants. He held that the Act of 1945 was applicable and assessed the respective degrees of fault as 75% to the plaintiff and 25% to the third defendants.

(b) The decision of Hobhouse JAs regards the applicability of the Act of 1945 to claims based on breach of contract was upheld by the Court of Appeal (O’Connor L J; Neill L J and Sir Roger Ormrod). Hobhouse J in the Court a quo approached the question now under consideration as follows : (see [1986] 2 All ER 488 at 508)

‘The question whether the 1945 Act applies to claims brought in contract can arise in a number of classes of case. Three categories can conveniently be identified. (1) Where the defendant’s liability arises from some contractual provision which does not depend on negligence on the part of the defendant. (2) Where the defendant’s liability arises from a contractual obligation which is expressed in terms of taking care (or its equivalent) but does not correspond to a common law duty to take care which would exist in the given case independently of contract. (3) Where the defendant’s liability in contract is the same as his liability in the tort of negligence independently of the existence of any contract.’

3. Hobhouse J held that Vesta v Butcher fell into category (3).

He said, at 509 :

‘The category (3) question has arisen in very many different types of case and the answer is treated as so obvious that it passes without any comment. It is commonplace that actions are brought by persons

who have suffered personal injuries as the result of the negligence of the person sued and that there is a contractual as well as tortious relationship. In such cases apportionment of blame is invariably adopted by the court notwithstanding that the plaintiff could sue in contract as well as in tort. The example normally cited in the present context is the decision of the Court of Appeal in Sayers v. Harlow Urban District Council [1958] 1 W.L.R. 623, which concerned a contractual visitor to premises (a lady who had paid to use a public lavatory). The Court of Appeal said it did not matter whether the cause of action was put in tort or in contract and proceeded to apportion blame awarding her three-quarters of her damages. This was a decision on a category (3) case. The power to make an apportionment was part of the ratio decidendi and is binding on me. There are innumerable similar decisions to the same effect which could be cited, very many by appellate courts.’

(d) O’Connor L J came to the conclusion that the claim of Vesta against the brokers fell into category (3) and was, therefore, subject to apportionment. Neill L J, now convinced that he was wrong in the decision given by him in A.B. Maintrans v Comet Shipping Co. Ltd. [1985] 1 W.L.R. 1270 (to the effect that apportionment could not be applied where the claim was one in contract), agreed that Vesta’s claim fell into category (3) inter alia

‘Where the broker’s liability in contract is the same as their liability would have been in tort. Accordingly, I would agree that as the claim against the brokers could have been framed in this action as a breach of a duty of care in tort any damages awarded to Vesta can properly be reduced and apportioned in accordance with the Act of 1945.' (at 875 F - G).

(e) Sir Roger Ormrod, the third member of the bench, held that the context of the Act of 1945 and the language of s 1 made it clear that the Act is concerned only with tortious liability. The power to apportion only arises where the defendant is liable in tort and that concurrent liability in contract, if any, is immaterial (at 879 C - D). However, notwithstanding that Vestas claim against the brokers was framed as one on contract,

‘ ... the existence of the contract created a degree of proximity between Vesta and the brokers sufficient to give rise, on ordinary principles, to a duty of care and, therefore, to a claim in negligence. Consequently, I agree with Hobhouse J that this is a case for apportionment of damages.’ (at 879 E - F).

(f) Vesta v Butcher went on appeal to the House of Lords. The appeal only dealt with the liability of the re-insurers (they were held liable) and the alternative claim against the brokers fell away and consequently the question of apportionment was not considered.

[8] Comparative law : New Zealand

(a) In New Zealand, where the Contributory Negligence Act 1947 follows the wording of the English Act of 1945, the case of Dairy Containers Ltd v N Z I Bank Ltd; Dairy Containers Ltd v Auditor-General [1995] 2 NZLR 30 (H C Auckland) (‘Dairy Containers’) raised the very difficulties with which we are now confronted. In that case, the Auditor-General was the auditor of the Dairy Containers Ltd (‘DCL’) by virtue of a contract between them. DCL sued the Auditor-General for damages, relying on a breach of contract by the latter, alleging a number of negligent acts and omissions. Thomas J held that the Auditor-General had been negligent and had thus committed a breach of contract. The Auditor-General argued that the damages awarded against him should be reduced having regard to DCL’s contributory negligence, inter alia, in failing to provide any clear direction or supervision in respect of a major part of the companys business (at 80 line 30 et seq).

(b) In discussing the approach to be taken to the test for contributory negligence the learned judge touched upon the heart of the matter, viz the policy underlying the approach to the apportionment question. He said (at 76 line 29 et seq):

‘Difficult though the exercise may at times be, the Act requires a Court to recognise the plaintiff’s failure to meet the standard of care required of it for its own protection where that failure is partly the cause of the loss. It is an attempt to ensure that liability coincides with the responsibility of the parties for the damages in issue. As Cook P said in Mouat v Clark Boyce (at p 563), it would be strange if after all these centuries the common law, using that word in its widest sense, had been able to produce only instruments of remedy so blunt and inefficient that apportionment of responsibility where it rightly belongs is impossible. The President did not believe that this was so. And nor do I.

The Contributory Negligence Act was enacted to remedy the arbitrary consequences of the all-or-nothing approach which developed where the plaintiff was in part responsible for the loss which he or she suffered. It is now inappropriate to approach the application of the Act in a manner which would perpetuate arbitrary consequences, although less dramatic, of the kind which the Act was designed to remedy. It is for the Courts, in implementing the Act, to fashion a regime under the Act which is fair and efficient in apportioning responsibility for the loss to where it rightly belongs.

In the context of this case there is no merit in providing DCL with immunity from the consequences of its negligence where that negligence has clearly contributed to cause the loss simply because the Auditor-General was also negligent. One can paraphrase Rogers CJ’s question, posed in a different context, in AWA v Daniels (at p 1003) and ask why the negligent auditor should be exposed to the payment of the whole of the loss when much of the damage lies at the door of the company? The answer is clear. It would be wrong in principle if the Auditor-General could be sued for failing to report the unauthorised investments and detect the frauds which occurred and yet not be able to rely upon DCL’s own acts which permitted the unauthorised investments and frauds to occur in the first place.’

(c) In the result apportionment was applied and DCL’s damages were reduced by 40% (see 83 line 31).

[9] Comparative law : Australia

(a) The opposite approach prevails in Australia. In Astley and Others v Austrust Limited [1999] HCA 6 (197 CLR 1), the High Court of Australia had to deal with the following problem. Austrust, the plaintiff, was a trustee company. It had sought advice from the defendant, Astley, a firm of solicitors, as to whether it should assume the position of trustee of an existing trading trust. The attorneys advised that it could so and Austrust took office. Shortly thereafter the trust failed and Austrust was held personally liable for its debts in an amount exceeding the value of the trust property. It alleged that the solicitors were at fault in failing to advise it not to accept the office of trustee unless its personal liability for losses arising in the course of carrying out the trust was excluded. The defendant denied liability but pleaded, in the alternative, contributory negligence on the part of Austrust. The trial judge found negligence on the part of both parties and, pursuant to the provisions of s 27 A of the Wrongs Act 1936 of South Australia, which Act was applicable to the dispute, apportioned the damages payable by the solicitors. On appeal to the Full Court it was held that the finding of contributory negligence on the part of Austrust was wrong, and Astley was held liable for the full extent of the damage proved by Austrust.

(b) On a further appeal to the High Court it was held that the Full Court erred in finding that Austrust was not guilty of contributory negligence. However, notwithstanding this finding and the fact that Austrust had sued in contract as well as in tort, it was held that Austrust was entitled to recover the whole of the damage that it suffered because damages awarded pursuant to a claim in contract cannot be reduced by reason of conduct that would constitute contributory negligence for the purposes of the Wrongs Act. It was held that the history, text and purpose of the Wrongs Act made it clear that the Act was not intended to apply to claims for breach of contract (per Gleeson CJ McHugh, Gummow and Hayne JJ). Callinan J, in a minority judgment, was of the opposite opinion and expressed the view that the trial court was correct.

(c) The majority first defined the concept of contributory negligence. It held that at common law contributory negligence consisted in the failure of a plaintiff to take reasonable care for the protection of his or her person or property. Proof of contributory negligence defeated the plaintiff’s cause of action in negligence. Furthermore, although conduct amounting to contributory negligence may also constitute the breach of a duty which the plaintiff owes to the defendant, a plaintiff can be guilty of contributory negligence in the absence of such a duty : a pedestrian owes no duty to a speeding driver to avoid being run down but is guilty of contributory negligence if he or she fails to take reasonable care to keep a proper lookout for speeding vehicles.

(d) The majority then dealt with the proposition raised in Australia and the USA in a number of decisions, especially in cases where auditors were sued, that contributory negligence cannot arise where the very purpose of the duty owed by the defendant was to protect the plaintiffs interests. Following the decision of the New South Wales Court of Appeal in Daniels v Anderson (1995) 37 NSWLR 438, the majority held that no such rule should apply :

‘Thus, a plaintiff who carelessly leaves valuables lying about may be guilty of contributory negligence, calling for apportionment of loss, even if the defendant was employed to protect the plaintiffs valuables.’ (at para 29).

This is also the case where there was a statutory duty to protect the plaintiff (paras 31 - 32).

(e) Having come to the conclusion that Austrust was contributorily negligent in the present case, and that contributory negligence requires an apportionment of damages in an action in tort, the majority held that the question is whether such negligence also requires apportionment where the plaintiff has sued in contract in circumstances where he or she has, or could have, sued in tort. (para 37).

(f) S 1 of the Wrongs Act 1936 (South Australia) defines ‘fault’ as follows :

‘ “fault” means negligence, breach of statutory duty or other act or omission which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence’ (My emphasis)

S 27 A (3) reads:

‘Where any person suffers damage as the result partly of his own fault and partly of the fault of any other person or persons, a claim in respect of that damage shall not be defeated by reason of the fault of the person suffering the damage, but the damages recoverable in respect thereof shall be reduced to such extent as the court thinks just and equitable having regard to the claimant’s share in the responsibility for the damage ...’

(g) The majority held that the natural and ordinary meaning of s 27 A (3), read in the light of the definitions contained in the section, leads to the conclusion that the section was concerned with claims in tort rather than claims in contract :

‘The sub-section was designed to remedy the evil that the negligence of a plaintiff, no matter how small, which contributed to the suffering of damage, defeated any action in tort in respect of that damage.’ (para 41)

And :

‘Nothing in s 27 A (3) suggests that “fault” - in either of its uses in s 27 A (3) - includes rights and obligations arising from a breach of contract. Nor is there anything in the ordinary and natural meaning in the section that can be said to assume or by necessary implication authorise the apportionment of damages in claims for breach of contract. On its face, s 27 A deals only with actions in tort.’ (para 42).

(h) The question the majority posed is whether s 27 A of the Wrongs Act was not applicable where the defendant’s obligation under the contract coincides with the duty imposed by the general law of negligence i e concurrent delictual and contractual liability were present. (para 43)

(i) The majority then examined the case law on this subject, as well as the influential views of Professor Glanville Williams in his book Joint Torts and Contributory Negligence, 1951, in which he advanced the proposition that apportionment legislation always applied to claims based on breach of contract because contributory negligence always constituted a possible defence to an action for damages for breach of contract. This last mentioned point of view was rejected by the majority : the plaintiffs negligence never gave rise to a defence of contributory negligence in an action for breach of contract according to the common law.

‘No doubt a plaintiff’s conduct, which could be equated with contributory negligence” in an action in tort, could defeat an action in contract. It might, for example, show that there was no causal connection between the plaintiff’s damage and the breach of contract. But “contributory negligence”, that is, negligence which contributed to the damage was not as such a defence to an action for breach of contract.’ (para 53)

(j) Reference was also made to the English case of Vesta v Butcher and a judgment of Pritchard J in New Zealand in Rowe v Turner Hopkins & Partners [1980] 2 NZLR 550, where apportionment was applied to contractual claims because there was or could have been concurrent tortious claims. These views were followed in a number of Australian cases but were unambiguously rejected by the majority in Astley v Austrust.

(k) The majority argued as follows:

2. (i) The tripartite division adopted and applied in Vesta v Butcher is unacceptable. The legislation does not hint at such a distinction. The reasoning in cases applying apportionment to contractual claims 'is generally sparse to the point of non-existence'. (para 69) (ii) The decisions in the UK which have applied apportionment legislation to breaches of contract are wrong and should not be followed in Australia. The interpretation of the legislation is strained. It relies principally, if not exclusively, on the use of the term 'negligence' in the definition of 'fault'. But it ignores the context of the said words in the legislation and ignores the mischief which the legislation was intended to remedy. (para 70)

3. (iii) A breach of contract does not come within the meaning of fault; the word 'negligence' is furthermore limited by the words 'which gives rise to a liability in tort'; if this qualification is taken away, it would mean that all breaches of statutory duty would fall under 'fault', and not only those that constituted a tort. (para 72)

4. (iv) Ss 27 A (3) and (4) support the view that an award for damages under a contractual claim is treated differently from an award in tort, and should not be subject to apportionment. (paras 73 - 75)

5. (v) The state of the pre-existing law and the purpose of the legislation made it clear that the legislation does not affect actions for breach of contract. At common law contributory negligence was a complete defence to an action in tort for negligence. But no case could be found where contributory negligence, as such, was ever held to be a defence to an action for breach of contract. No trace of such a defence could be found in the great works on pleading written in the 19^th century. This pointed 'irresistibly' to the conclusion that the apportionment legislation is concerned only with actions in tort and did not affect awards of damages based on breach of contract:

'To what, other than a common law action in tort, can s 27 A (3) be referring when it says that a claim in respect of damage shall not be defeated by reason of the fault of the person suffering the damage”? It makes no sense now, and it made even less sense when the legislation was passed, to speak of an action in contract being defeated by "negligence, breach of statutory duty or other act or omission which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence!".' (para 80)

In view of the history of the legislation in England, in Admiralty and in Canada, the majority said :

‘It would be strange if a rule introduced to do away with an absolute defence to a claim in negligence, diminished the rights of a plaintiff who sued in contract.’

(Para 81; see also paras 82 and 84)

It was also said :

‘The section was designed to increase the rights of plaintiffs, not reduce them.’ (para 59; see also para 83)

2. (vi) The majority also found support for their views in policy considerations. The view that it would be anomalous or unfair or both not to apply apportionment legislation to contractual claims was rejected. In an action based on contract, the defendant could have limited its liability; in a claim based on tort, the respective duties of the parties are imposed on them by law. In the former case there seems to be no reason to apply the apportionment principle. (paras 84 - 87)

(l) In the result it was held that, although Austrust was correctly held to have been 'contributorily negligent', no apportionment could be applied.

(m) The decision in Astley v Austrust was subjected to criticism by Geoff Masel and David Kelly see Contributory Negligence and the Provision of Services : A Critique of Astley, in 74 Australian Law Journal 306 et seq. In particular, I draw attention to the remarks at 324 :

'In principle, the rules of our legal system should be consistent with one another. At least presumptively, there should not be a different answer in tort from the one given in contract on precisely the same issue - liability for negligent advice in performing a contract. If a plea of contributory negligence is available in one action, why not also in the other? If the plea can lead to apportionment in one action, why not also in the other? If we expect our legal system to be efficient and to be respected, we cannot tolerate overlaps and inconsistencies which have no rational foundation, but which are explicable only in terms of procedural history.'

[10] Comparative Law : Canada

In Canada the position seems to be that the wording of the various provincial statutes dealing with the problem now under consideration differs - and thus also the decisions of the provincial courts (see G H L Fridman The Law of Contract in Canada 3^rd ed., 756). In Giffels Associates Ltd v Eastern Construction Co ((1978) 84 D.L.R. (3d) 344) the Supreme Court of Canada left the matter undecided. In two Ontario decisions the courts reached a result similar to apportionment by developing a doctrine of 'anticipatory mitigation'. (Fridman op cit 756 - 757.

Fridman (loc cit) concludes :

'It would seem that the issue is unresolved by Canadian courts. On principle there should be no barrier to the application of the idea of apportionment in breach of contract cases. Even if the legislation that applies to tort cannot be interpreted to extend to contract situations, the need to establish a casual connection between the plaintiff's loss and the defendant's breach of contract, to which reference has been made should logically lead to the conclusion that where the plaintiff has been partially responsible for his loss he should bear that proportion by a reduction in his damages. Whether Canadian courts will finally arrive at this conclusion remains to be seen.'

[11] South Africa : Quo vadis?

(a) The approach followed in Vesta v Butcher in England and Dairy Containers in New Zealand i e that if the defendant is concurrently liable in delict and contract, apportionment can be applied, is not open to us. This is so by virtue of the decision of this Court in Lillicrap, Wassenaar and Partners v Pilkington Brother (SA) (Pty) Ltd 1985 (1) SA 475 (A) ('Lillicrap'). The plaintiff, a glass manufacturer, wished to build a glass plant on a particular property. It concluded a contract with the defendant, a firm of consulting structural engineers, to investigate the site and to determine its suitability, and if suitable, to design the plant. The defendant did the work, but, according to the plaintiff, negligently. Having paid a part of the defendant's fees, the plaintiff sued the defendant for payment of damages in the sum of  R3,6 million. The plaintiff sued in delict only, more particularly basing its claim on the actio le gis Aquiliae in its modern and extended form. To this claim the defendant raised an exception, particularly on two bases : (a) that, in the light of the contractual relationship between the parties, the defendant did not owe the plaintiff a delictual duty of care; and (b) that, in the light of the contractual relationship between the parties, the alleged facts did not give rise to any claim for damages in respect of pecuniary or financial loss only. (at 495 D - E)

(b) The majority, per Grosskopf AJA, held that in principle our law recognises concurrent contractual and delictual claims, and allows the plaintiff to choose the remedy which he wishes to pursue. Thus, the facts giving rise to a claim for damages under the lex Aquilia could in Roman and Roman-Dutch law overlap with those founding an action under certain types of contract, e g deposit, commodatum, lease, partnership or pledge. The learned judge concluded at 496 H - I :

'The mere fact that the respondent might have framed his action in contract therefore does not per se debar him from claiming in delict. All that he need show is that the facts pleaded establish a cause of action in delict. That the relevant facts may have been pleaded in a different manner so as to raise a claim for contractual damages is, in principle, irrelevant.'

(c) The learned judge then held that in the case under consideration there was no physical damage to the plaintiff or his property but that, per se, would not be fatal to a claim based on the lex Aquilia - our law already recognises Aquilian liability for negligent misstatements which cause pure financial loss. (Administrateur, Natal v Trust Bank van Afrika Bpk 1979 (3) SA 824 (A)). In casu, the learned judge held (at 499 A et seq) that it was not alleged by the plaintiff that the defendant would have owed it a duty to exercise diligence if no contract had been concluded between them requiring it to perform professional services. In this respect the case differed from those in which a physician operates on a person found unconscious in the street - if the doctor was negligent in performing the operation, he would only be liable ex delicto. But if there was a contract between the physician and his patient, there would be, in the case of negligence, concurrent actions in contract and delict against the physician, because even in the absence of a contract, there would have been a violation of an existing right i e that of physical integrity.

(d) But, Grosskopf AJA held, the only infringement of which the present plaintiff complained was that of the defendant's contractual duty to perform specific professional work with due diligence (at 499 D - E). Is the infringement of this duty a wrongful act for purposes of Aquilian liability? (at 499 E - F). Grosskopf AJA held that those cases in our common law where concurrent liability was recognised, occurred when the conduct of the defendant constituted both an infringement of the plaintiff's rights ex contractu and a right which he had independently of the contract. (My emphasis, see at 499 H - I and also Van Wyk v Lewis 1924 AD 443.)

(e) Is an extension of Aquilian liability justified in cases where the right allegedly infringed is one created by a contract for delivery of professional services? Is there a need therefor? Grosskopf AJA answered in the negative. (at 500 F)

'While the contract persisted, each party had adequate and satisfactory remedies if the other were to have committed a breach. Indeed the very relief claimed by the respondent [plaintiff] could have been granted in an action based on breach of contract.'

(f) Grosskopf AJA further held (at 500 G et seq) that the Aquilian action does not fit comfortably in a contractual setting like the present. Contracting parties contemplate, generally speaking, that the relationship between them should be regulated by their agreement. If one superimposes Aquilian liability on claims for breach of contract, a party's performance would, so Grosskopf AJA said (at 500 I), presumably have to be tested not only against his contractual duties but also by applying the standard of the bonus paterfamilias. But how is the latter standard to be determined? Are there two standards? There are no policy considerations for invoking the law of delict to reinforce the law of contract. (at 501 A - B) In the result, it was held that

'To sum up, I do not consider that policy considerations, require that delictual liability be imposed for the negligent breach of a contract of professional employment of the sort with which we are here concerned.' (at 501 G - H)

(g) Finally, Grosskopf AJA, after reviewing English cases such as Anns v Merton London Borough Council 1978 AC 728; Donoghue v Stevenson 1932 AC 562; Hedley Byrne & Co Ltd v Heller and Partners Ltd, supra; Home Office v Dorset Yacht Co. Ltd (1970) AC 1004 and Junior Books Ltd v Veitchi & Co Ltd 1983 AC 520, noted the difference on the matter now under discussion between English and South African law : English law adopts a liberal approach to the extension of a duty of care; South African law approaches the matter in a more cautious way (at 504 A - G). Also, the policy considerations underlying the two approaches may be different (at 504 H - I).

[12] Thus, in a nutshell : in the light of the judgment in Lillicrap the 'concurrent liability' solution is not available to us to apply in the present case, i e in the category 2 type of situation referred to in Vesta v Butcher. In this respect it must be noted, perhaps en passant, that our law now lags far behind the English law on this point - see Henderson and Others v Merrett Syndicates Ltd and Others [1994] 3 All ER 506 (HL).

[13] This is not the end of the matter, because we must now examine the position where, although it cannot be said that the defendant is not liable in delict, and only committed a breach of contract, he or she did so negligently, i e was 'at fault'. As I have said, the present case falls into the category 2 class of case developed in Vesta v Butcher, supra. That negligence in such a case becomes highly relevant for success for the plaintiff is so because the contract imports that standard. What is more, in our law s 20 (9) (a) of the Public Accountants' and Auditors' Act 80 of 1991 provides that an auditor shall not incur any liability in respect of a statement, account or document certified by him unless it was certified maliciously or pursuant to a negligent performance of his duties.

[14] Although the full ramifications and implications of the matter might require a full dissertation, expedience requires that we set out as succinctly as possible the arguments pro and contra the proposition that the Act is applicable to contractual claims.

A The pro arguments and counter submissions

1 The plain meaning of s 1 of the Act.

1.1 The first argument is that if the golden rule of statutory interpretation is applied, the provisions of s 1 of the Act do seem to be applicable to contractual claims. The rule has been stated in Adampol (Pty) Ltd v Administrator, Transvaal 1989 (3) SA 800 (A) at 804 B - C as follows:

'The plain meaning of the language in a statute is the safest guide to follow in construing the statute. According to the golden or general rule of construction the words of a statute must be given their ordinary, literal and grammatical meaning and if by so doing it is ascertained that the words are clear and unambiguous, then effect should be given to their ordinary meaning unless it is apparent that such a literal construction falls within one of those exceptional cases in which it would be permissible for a court of law to depart from such a literal construction, eg where it leads to a manifest absurdity, inconsistency, hardship or a result contrary to the legislative intent."

1.2 For the sake of convenience, s 1 of the Act will be repeated :

'1. Apportionment of liability in case of contributory negligence. - (1) (a) Where any person suffers damage which is caused partly by his own fault and partly by the fault of any other person, a claim in respect of that damage shall not be defeated by reason of the fault of the claimant but the damages recoverable in respect thereof shall be reduced by the court to such extent as the court may deem just and equitable having regard to the degree in which the claimant was at fault in relation to the damage.

(b) Damage shall for the purpose of paragraph (a) be regarded as having been caused by a person’s fault notwithstanding the fact that another person had an opportunity of avoiding the consequences thereof and negligently failed to do so.

…

(3) For the purposes of this section "fault" includes any act or omission which would, but for the provisions of this section, have given rise to the defence of contributory negligence.

'

1.3 The argument is that s 1 of the Act is clear and unambiguous. The only threshold requirement is causative fault on both sides. Fault, at the very least, includes negligence. S 1 of the Act does not specify the categories of obligations in which the 'fault' can occur, i e delict, contract, statute or ex variis causarum figuris. There is nothing in the language of s 1 (1) (a) that limits its operation to claims in delict; by its plain wording it is equally apposite to claims for damages flowing from a negligent breach of a statutory duty or a breach of a contractual duty to exercise reasonable care.

1.4 The counter-submission on this aspect was that the wording is not all that clear. It was argued that there are two provisions indicating that s 1 was not intended to apply to claims in contract :

(a) the long title of the Act, and

(b) Chapter 2 of the Act

1.5 But this counter-submission, so it was argued, is flawed. It is true that the long title of the Act, and other provisions of the Act can and should be taken into account when s 1 is interpreted (see Stellenbosch Farmers' Winery Ltd v Distillers Corporation (SA) Ltd and Another 1962 (1) SA 458 (A) at 476 E - F; Commissioner of Taxes v First Merchant Bank of Zimbabwe Ltd 1998 (1) SA 27 (ZSC) at 30 I - 32 E). But neither the long title nor the provisions of Chapter 2 of the Act support the view that s 1 of the Act is not applicable to contractual claims, so it was submitted.

1.6 Counsel for PW for ease of reference numbered the components of the long title as follows:

'1 To amend

(a) the law relating to contributory negligence and

(b) the law relating to the liability of persons jointly or severally liable in delict for the same damage, and

2 to provide for matters incidental thereto.'

S 1 of the Act gives effect to part 1 (a). Ss 2 and 3 (Chapter 2) give effect to part 1 (b). Ss 4 to 6 give effect to part 2. Counsel for PW submitted that this analysis of the long title into its components clearly shows that it was not the legislature's intention that s 1 (1) (a) or part 1(a) above should apply only to claims in delict. Part 1 (a) is at least equally consistent with the very opposite interpretation, i e that whereas part 1 (b) specifically refers to claims in delict, part 1 (a) does not do so. If it had been the intention of the legislature to limit the application of both parts to claims in delict, it would, and should, have said so in plain, conjunctive terms. That it has not done so, PW's counsel argued, supports its side of the argument.

1.7 What is more, it was submitted by PW, there was a good reason why the legislature made part 1 (b) applicable to delictual joint wrongdoers, but not to contractual co-debtors. While delictual joint wrongdoers were, in our common law, always jointly and severally liable, contractual co-debtors may according to the intention of the parties, be liable jointly or severally or jointly and severally (see Christie, The Law of Contract in South Africa, 3^rd ed 1996 at 279, et seq). It was thus both unnecessary and impossible to make the philosophy underpinning Chapter 2 of the Act applicable to contractual co-debtors.

1.8 This analysis indicates, so it was argued, that there was a sound reason for distinguishing between Chapter 1 and Chapter 2 of the Act in so far as it concerns contractual vis-à-vis delictual claims. The analysis elucidates the wording of the preamble and the interpretation of s 1 (1) (a) the Act. It supports the pro-argument, so it was submitted.

2 Purposive and teleological interpretation of the Act

But the pro-argument does not rest solely on the golden rule of interpretation. On behalf of PW it was argued that weighty considerations favour the purposive or teleological approaches. The last-mentioned approach, in particular, not only 'encapsulates in a synthesis the meritorious aspects of other theories and excludes their limitations' (Devenish, Interpretation of Statutes, 1992 at 53) but also gives expression to the fundamental principles and ethos of the legal system as a whole: it is a value-coherent approach which best accords with the values of our Constitution (see, in general, Devenish, op cit, 39 - 55).

The pro-argument, so it was submitted, is supported by both a purposive or teleological approach to the Act. Both approaches recognise the importance of the genesis and legislative history of the Act. We have discussed the common law background above. We now turn to the specific legislative history of the Act.

2.1 S 1 was modelled on the English Law Reform (Contributory Negligence) Act of 1945. It closely followed the language of s 1 (1) of the English Act which we repeat for the sake of convenience:

'Where any person suffers damage as the result partly of his own fault and partly of the fault of any other person or persons, a claim in respect of that damage shall not be defeated by reason of the fault of the person suffering the damage, but the damages recoverable in respect thereof shall be reduced to such extent as the court thinks just and equitable having regard to the claimant's share in the responsibility for the damage: Provided that -

• this subsection shall not operate to defeat any defence arising under a contract;

• where any contract or enactment providing for the limitation of liability is applicable to the claim, the amount of damages recoverable by the claimant by virtue of this subsection shall not exceed the maximum limit so applicable.'

2.2 As appears from a comparison of the two texts, the main provision of the English s was substantially reproduced in s 1 (1) (a) of our Act, and

• the provisos of the English section were substantially reproduced in our ss 4 (1) (b) and (c).

2.3 The only significant departure from the English text is in the definition of 'fault'. S 4 of the English Act defines it as follows:

' ... "fault" means negligence, breach of statute or duty or other act or omission which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence.' (My emphasis)

2.4 When this definition of fault is read into s 1 (1) of the English Act, its effect clearly appears to be to confine s 1 (1) to claims in tort.

(a) The definition purports to be an exhaustive definition of

'fault' (' ... fault means ...')

(b) The first part of the definition refers to the fault of the defendant (' ... negligence, breach of statutory duty or other act or omission which gives rise to a liability in tort ... ') and the second part to the fault of the plaintiff (' ... or other act or omission which ... would, apart from this Act, give rise to the defence of contributory negligence ...').

(c) The description of the defendant's 'fault' in the first part of the definition seems to suggest that it was limited to claims in delict (' ... which gives rise to liability in tort ...').

(d) This inference is fortified by the description of the plaintiff's fault in the second part of the definition because it is confined to conduct which would previously have given rise to a defence of contributory negligence, which was in English law a defence in delict but not in contract.

2.5 However, in 1951 Professor Glanville Williams published his Joint Torts and Contributory Negligence in which he argued forcefully and persuasively that the Act also applied to claims in contract. His argument was directed at the interpretation of the definition of 'fault' because that was considered to be the only obstacle to the application of the Act to claims in contract:

'Whether the Act applies in contract depends largely upon the wording of the definition of 'fault'. At first sight the definition ... may appear to be limited to actions in tort, but it is submitted that where a breach of contract occurs through the negligence of the defendant, the Act will apply whether the action is framed in contract or in tort.' (329)

He proposed various arguments in support of this interpretation (328 - 332). The second of those arguments eventually prevailed in the English courts :

'There is another line of argument. Even if the interpretation just advanced is thought to be too fine-spun it is submitted that where the same act or omission constitutes both a tort and a breach of contract, so that in its tort aspect the case is subject to the provisions of the Act, then the case is subject to the provisions of the Act even in its contract aspect.' (330) (See also Cheshire, Fifoot & Furmston's Law of Contract 12^th ed 619; Jackson and Powell on Professional Negligence 4^th ed 60 to 62; and Treitel The Law of Contract 10^th ed 915 to 919.)

2.6 The point is that our departure from the English text strongly suggests that our s 1 (1) (a) was intended to apply to claims in contract as well.

2.7 Our Act was only enacted in 1956, well after the publication of the treatise of Professor Glanville Williams which triggered the controversy abroad.

2.8 The differences between the definition of 'fault' in the English Act and our Act are significant :

(a) The English definition exhaustively defines 'fault' (' ... fault means ...'). Our section is open-ended (' ... fault includes ...'). The effect of our section is in other words merely to extend the ordinary meaning of 'fault' and not to limit it in any way.

(b) The only real obstacle to the application of the English section to claims in contract, lay in the first part of their definition of 'fault' which defined the defendant's fault, and therefore the plaintiff's cause of action, as conduct giving rise to liability in tort. That part of the definition has been wholly omitted from our section.

(c) The combined effect of these two changes is that the only real obstacle to the application of the English Act to claims in contract, has been removed from our Act, and apparently deliberately so.

2.9 It is also significant, so it was submitted on behalf of PW, that there was a simultaneous and comparable but opposite amendment of s 2 of our Act which deals with joint wrongdoers and which expressly limits its application to persons liable in delict.

2.10 It accordingly seems clear, so it was argued, that our Act was deliberately drafted so as to differ from the English text to remove from s 1 the limitation of its operation to claims in delict, and

• to introduce in s 2 an express limitation of its operation to joint wrongdoers in delict. On behalf of PW it was accordingly submitted that the legislative history of our Act, when compared to its English parentage, overwhelmingly suggests that it was deliberately tailored so as to make the principle of apportionment also applicable to claims in contract when there is a similar co-incidence of negligence.

2.11 But both approaches to the interpretation of statutory interpretation mentioned above also attach importance to the purpose of the legislation, i e the mischief aimed at and the societal and legal ends desired. The phenomenon of causative negligence of the part on both a plaintiff and a defendant is not limited to delictual claims. It is obvious that in many instances of contractual claims for damages there can and will be a co-incidence of both contractual and delictual liability (i e if there was damage of the kind giving rise to Aquilian liability, e g in the case of a physician's negligence, as in Van Wyk v Lewis 1924 AD 438 or Mukheiber v Raath 1999 (3) SA 1065 (A)). If the plaintiff sues in delict, the Act would apply and the plaintiff would be liable only in part; if the action is brought in contract, the plaintiff would succeed totally if one follows the approach of our courts at present, OK Bazaars, supra. Why should there be a difference, it was rhetorically asked by PW, depending not on the acts or the respective degrees of fault or blameworthiness of the parties, which are the same in both actions, but on the form of action chosen by one of the parties viz the plaintiff? Commenting on the position in the English law, Buckley (The Modern Law of Negligence, 3^rd ed 1999 at 84) says of the approach in Vesta v Butcher that it is sensible, and proceeds:

'A rigid demarcation between tort and contract would seem mechanistic and outdated today, not least in the expanding field of professional negligence where allegations, amounting in substance to claims that defendants failed to take reasonable care, are often advanced in a contractual context.'

2.12 South African legal writers, it was pointed out by counsel for PW, have also remarked on the indefensibility of the distinction between contractual and delictual claims as far as the applicability of s 1 (1) (a) of the Act is concerned. Christie (The Law of Contract in South Africa,. 3^rd ed 1996 at 613 - 614) is of the view that

' … it is undesirable to leave the law in a state where the employment of purely technical skill in pleading may lead to a result fundamentally different from that which would be reached if a lesser degree of technical skill were employed. When a contract contains an express or implied term imposing an obligation not to be negligent (which very frequently happens) a breach of this term may equally well be described as a breach of contract or a delict giving rise to Aquilian liability. Under our law as it presently stands a skilful pleader, by pleading such a case in contract, could avoid the danger of a reduction of damages by apportionment under the Act, whereas a less skilful pleader, pleading the same facts in delict, would lay his client open to a reduction in damages … This degree of knife-edge technicality should be eliminated from the law where possible … '

(In similar vein, see also D J Lötz Vermindering van kontraktuele skadevergoeding in 1996 TSAR 172; P H Havenga Contractual Claims and Contributory Negligence (2001) 64 THRHR 125 - 126.) Commenting on the judgment of the Court a quo in the present case, Havenga, who supports the judgment, states at 128 :

'Goldstein J's interpretation of the Act is to be commended. It is not absurd, inconsistent or anomalous. Quite the contrary : it is absurd to non-suit a plaintiff merely because he or she has suffered damage caused partly by his or her own fault. In this case, it would be inconsistent and anomalous to have different rules for claims based on breach of contract and for claims founded in delict.

B Thus far the pro-arguments. We now turn to the contra arguments and the counter-submissions thereto.

1 The Act, so it was argued by TBA's counsel , is not applicable to contractual claims because contributory negligence was never by our common law and pre-1956, considered as a factor to be taken into account where a plaintiff sues in contract. This is borne out, so it was argued, by s 1 (3) of the Act. It was argued that the section says 'fault' 'includes any act or omission which would, but for the provisions of this section, have given rise to the defence of contributory negligence'. TBA argued that such a defence was not in the common law applicable to a plaintiff's claim in contract, therefore it must have been the legislature's intention that s 1 of the Act would apply to delictual actions only - where contributory negligence was a defence. This is the strongest argument presented on behalf of TBA, and the one on which the other judgments in the present case are based. The argument submitted on behalf of TBA is, perhaps, best formulated by Masel and Kelly (op cit at 313 - 314) and also best refuted by them.

'The 1945 English reform and subsequent apportionment legislation allows for apportionment where a person suffers damage as a result partly of his or her own fault and partly of the fault of another person. It is the definition of "fault" which has given rise to the problem, as it was defined ambiguously to mean "negligence, breach of statutory duty or other act or omission which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence".

The drafters presumably intended this provision to apply to actions in tort. They probably assumed that contributory negligence was not a defence to an action in contract. If it had been, it would have been a complete bar to a plaintiff's action and it would have been necessary to protect plaintiffs by extending the legislation to that situation. As the majority judgment states: "The section was designed to increase the rights of plaintiffs, not reduce them." However, the legislation was passed at a time when it was thought that contract and tort were completely discrete in their coverage, and there was no suggestion of concurrent liability. The relatively recent emergence of the principle of concurrent liability, which was reaffirmed by the High Court in Astley, has given rise to a new question. The question is whether the emergence of concurrent liability requires an adjustment to be made of the application of the legislation to avoid the anomaly that arises if the legislation is restricted to actions brought in tort.

The majority judgment analyses those decisions which have applied the apportionment legislation in cases where the defendant's obligation under contract was commensurate with his or her duty in tort. It recognises that the concluding phrase of the definition of "fault" - "which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence" - might be interpreted as only governing "other act or omission" and not as governing the opening words "negligence, breach of statutory duty". But it fails to follow up the possibility. It also notes, but does not adopt, the suggestion made by Pritchard J in Rowe v Turner Hopkins & Partners that :

"The second limb of the definition means simply and logically that no act or omission of the plaintiff will entitle the defendant to a reduction of damages unless it amounts to the sort of conduct which, prior to the enactment of the Contributory Negligence Act, would have afforded a defence of contributory negligence."

The majority interprets the apportionment legislation as being restricted to an action in tort. But there is nothing in the legislation itself that requires that result. The legislation defines fault in terms of conduct which would give rise to a defence of contributory negligence. The plaintiff's conduct would have given rise to that defence in an action in tort. Consequently, the legislation applies. The fact that the particular action is one in contract, not tort, is irrelevant. The basis for the action - the defendant's breach of the duty of care - is precisely the same in each case. Why not apply the legislation?'

But there is a further argument, once again based on a clear difference between our Act and the English statute. The English Act specifically states in s 4 that fault

'means negligence, breach of a statutory duty or other act or omission which gives rise to a liability in tort or would, apart from this Act, give rise to the defence of contributory negligence.'

Our s 3 reads :

'For the purposes of this section "fault' includes any act or omission which would, but for the provisions of this section, have given rise to the defence of contributory negligence.'

The wording of the English section is much stronger than ours. 'Fault' is defined in definite terms - it 'means' only what is described in the section. In our section, fault is not so defined. Fault only 'includes' the acts or omissions described, but does not exclude other acts or omissions which in law would also amount to 'fault'. In our Act, this definition could simply signify an attempt to indicate that the well-known rule of contributory negligence in the law of delict is to be retained, i e to leave no doubt that contributory negligence is a species of 'fault'.

At best for TBA, s 1 (3) is ambiguous. We thus have this position : s 1 (1) (a), on its wording, unambiguously would allow apportionment in contractual claims. S 1 (3) is ambiguous and possibly excludes apportionment in such claims. In such a case - and even if one views section1 (3) as unambiguous - the situation is that two conflicting interpretations can be given to s 1. In such a case the correct approach, in my view, is the one laid down by this Court as long ago as 1931 in Principal Immigration Officer v Bhula 1931 AD 323 at 336 per Wessels JA as follows :

'There is another principle which ought to be invoked in this case, and which would lead to the same interpretation. It has been repeatedly laid down by this Court that where a statute is clear, the Court must give effect to the intention of the Legislature, however harsh its operation may be to individuals affected thereby. Where, however, two meanings may be given to a section, and the one meaning leads to harshness and injustice, whilst the other does not, the Court will hold that the Legislature rather intended the milder than the harsher meaning. This principle is thus stated by Maxwell (3^rd ed., p. 299): "A sense of the possible injustice of an interpretation ought not to induce judges to do violence to well-settled rules of construction, but it may properly lead to the selection of one rather than the other of two possible interpretations per Lord Herschel in Arrow Shipping Co v Tyne Commissioners (1894) A.C. 516. Whenever the language of the Legislature admits of two constructions, and if construed in one way would lead to obvious injustice, the Courts act upon the view that such a result could not have been intended unless the intention had been manifested in express words"; and cases cited.'

There can be no doubt that, for the reasons discussed above, fairness and justice favour the approach that s 1 of the Act should apply also to contractual claims.

2 The second contra-argument was based on the precedents created by the decisions in Shaw and OK Bazaars, supra.

2.1 The two said decisions were, to put it mildly, not well received by our academic writers (see inter alia Boberg 1965 Annual Survey of South African Law 179 - 180; Jean Davids, Altered cheques : Apportionment of loss (1965) (82) SA Law Journal 289; Jean Davids, Apportionment and contractual damages (1966) 83 SA Law Journal 226; Boberg, The Law of Delict, vol 1, 1984, 710 - 713; Louise Tager, 1976 Annual Survey of South African Law 87 - 88; DJ Lötz Vermindering van kontraktuele skadevergoeding, in (1996) TSAR 170 - 174; P H Havenga Contractual Claims and Contributory Negligence in 2001 THRHR 124 - 130; Christie, op cit 613 et seq. In a submission to the SA Law Commission Professor A J Kerr opposes the view that the Act is applicable to contractual claims but favours apportionment on the basis of causation.)

2.2 The judgment in OK Bazaars was and can be criticised on the following bases:

(a) The main argument in OK Bazaars in favour of applying the Act was that the word 'fault' in s 1 of the Act was wide enough to include a breach of contract. Watermeyer J dismissed this argument : some forms of breach of contract do not require fault on the part of the debtor. It follows that if s 1 of the Act is applicable, it would apply to some breaches of contract (where fault is required) but not to others. This was the first and seemingly most important reason why Watermeyer J rejected the argument.

With respect, this argument is not convincing. The object of the Act was to regulate those cases where both parties acted negligently. It excluded from its operation cases of strict liability, statutory liability and contractual liability which do not depend on proof of negligence. S 1 (1) (a) of the Act specifically refers to cases where both parties are at 'fault'. How can the argument that the section cannot be applied, even if this particular defendant is at fault, because other defendants in contracts may be liable without any fault, be sound?

(b) It was also held in OK Bazaars that the history of the Act shows that it was intended to apply to delictual actions only. But, as was argued by counsel for PW, this argument loses sight of the crucial difference between our Act and the English Act. If our Legislature intended s 1 to apply to delictual actions only, why did it not simply follow the English Act?

(c) It was also held in OK Bazaars that the words in s 1 '... a claim in respect of that damage shall not be defeated by reason of the fault of the claimant' support the view that the section was intended to apply to delictual claims only: it would not be apposite to say that fault on the part of the plaintiff who sued on contract would 'defeat' his or her claim; the plaintiff's claim would fail because he would not be able to show the causal connection between the defendant's act and the damage.

The argument, so it was argued, is untenable. The very same can be said of a claim in delict : by 1956 a plaintiff's claim in delict was not 'defeated' by reason of his or her fault - it would fail, sometimes, because the plaintiff's conduct, and not that of the defendant, was the proximate cause of the loss (i e he or she had the last opportunity to avoid the loss).

(d) It was then said in OK Bazaars that the rule in s 1 (1) (b) that damage, for the purpose of s 1 (1) (a) shall be regarded as having been caused by a person's fault notwithstanding the fact that another person had the last opportunity of avoiding the consequences thereof and negligently failed to do so, showed that the apportionment was intended to apply to delictual claims only.

But the argument seems to be a non sequitur. Even if it is assumed that s 1(1)(b) was enacted to overcome the line of decisions based on the 'abracadabra' criterion of causation and the last opportunity rule, non constat that the new apportionment principle cannot and should not be equally applicable to contractual claims. S 1 (1) (a) is the dominant clause, not s 1 (1) (b).

(e) It was also held by Watermeyer J that contributory negligence is not normally one of the recognised defences to a claim based upon a breach of contract. On behalf of PW it was submitted that this argument overlooks the wording of the whole of the Act and the principle laid down in Principal Immigration Officer v Bhula, supra, viz that in the case of conflicting or ambiguous provisions, the fair and equitable interpretation should be followed, rather than a harsh and uncompromising one or, one can add, rather than an approach which leads to unjustifiable discrimination

between classes of defendants.

(f) It was also held that the application of various canons of constructions, inter alia relating to the long title of the Act which refers to contributory negligence, indicates that the Act was intended to apply to delictual claims only.

PW submitted that there is no canon of construction which militates against the view that the Act applies to contractual claims. On the contrary, it was argued, an analysis of the wording of the long title (as done above) and of s 1 (1) (a) shows the opposite.

(g) It was also held in OK Bazaars that, inasmuch as contributory negligence was, before the passing of the Act, not a valid defence to a claim based upon a breach of contract, then, if it was the legislature's intention to change the legal position, it would have done so explicitly. Based on its previous submissions, PW argued that it was clear what the legislature intended to do; furthermore, if it was the legislature's intention to apply the Act only to delictual claims, why did it not simply follow the English Act?

(h) Finally, the argument was put before the court in OK

Bazaars that even if the Act did not apply to all contractual claims, it should at least be construed as covering claims for breach of contracts which import a duty not to be negligent i e at fault, in committing the breach of contract. Watermeyer J was not convinced by the English authorities and distinguished them.

But, PW argued, the substance of the argument was not addressed : if the breach of contract by the defendant requires proof of fault to found a claim for damages against the defendant, and the plaintiff is also at fault, why should s 1 (1) (a) of the Act, according to its clear terms, and as a matter of logic, legal policy, fairness and justice not be applicable?

3 The status of OK Bazaars

The next argument submitted by TBA was that the decision in OK Bazaars has now stood for more than two decades and should not be overturned even if wrong, except by a clear legislative intervention.

This Court has on several occasions rejected this approach. In Dukes v Marthinusen 1937 AD 12 Stratford ACJ said at 23 :

'If the decisions had disregarded fundamental principles of our law, we might have to reassert those principles even at the cost of reversing judgments of long standing.'

(See also S v Bernardus 1965 (3) SA 287 (A) at 297 in fine - 299 A; Peri-Urban Areas Health Board v Munarin 1965 (3) SA 367 (A) at 376 E - G.)

[15] Conclusion

(a) What is plain from the above discussion is that the feasibility of a plea of contributory negligence in the case of a claim for breach of contract on the defendant's failure to exercise due care depends upon an exercise of statutory interpretation. Behind this, however, lies important policy considerations. That being so, there are two interrelated considerations which cause me to lean in favour of the applicability of s 1 to claims of a contractual nature. These are :

(i) the need for its applicability. This is not simply an academic exercise: there is a definite lacuna in the law if such a defence is to be denied in the narrow circumstance which apply in this case.

(ii) the glaring inequity of denying the existence of such a defence in circumstances such as those prevailing in this case.

(b) The facts of the current case provide a perfect illustration of both propositions. It would be patently unfair if PW should have to bear the full brunt of the entire loss when TBA was itself partly to blame for its occurrence. The greater the comparative degree of a plaintiff's lack of precaution in relation to the harm of which he complains, the more apparent will be the inequity of the denial of a plea of contributory negligence.

(c) One can readily conjure up other comparable instances. One that was much discussed during the course of argument is this: a contract is entered into between a building owner and a contractor. The specifications furnished by the building owner to the contractor negligently stipulate an incorrect mix for the concrete he is to use. The contractor, on the other hand, is negligent in that, contrary to proper specifications about reinforcing, he provides inadequate reinforcing. As a result a wall collapses. According to expert evidence both factors contributed thereto. The building owner sues the contractor for the damage it sustained as a result thereof. Is it fair that the plaintiff should succeed in full or not at all? Another telling example is furnished in the recent report of the Scottish Law Commission entitled Report on Remedies for Breach of Contract” (1999) which, in part 4 thereof, deals with this very issue. In para 4.10 it is stated:

'On principle it would seem to be desirable to take into account the conduct of the aggrieved party in contributing to the loss or harm. This is just an extension of the policy underlying the well-established rules on mitigation of loss. In cases where loss of damage is sustained as a result of breach of contract it will often be the case that the aggrieved party is partly to blame for the loss or harm. To force courts into an all or nothing choice is likely to produce unreasonable results.

Example. A contractor contracts with an electricity supply company for a continuous supply of electricity. The company, in breach of the contract, allows an interruption in the supply. This is one of the causes of a loss to the contractor who has to re-lay a large column of concrete. Another causal factor was that the contractor failed to take reasonable steps to see that a back-up system was available before beginning a task for which a continuous supply of concrete was indispensable.

In a case like this, awarding the contractor full damages or no damages may be equally unattractive. The reasonable course may be to apportion the liability, taking the conduct of both parties into account. Other, more commonplace, examples could easily be imagined. For example, a party to a contract for the carriage of goods gives the carrier a wrong address and then, when the carrier fails to take all reasonable steps to ascertain the correct address in time, claims damages for late delivery. Or a person who has bought sophisticated electronic equipment which is not in all respects conform to contract causes damage to it by ignoring the clear instructions supplied with it and taking foolish and unreasonable steps to remedy the small defect. Or a woman injures herself in foolishly and unreasonably attempting to climb over a high gate which ought, in terms of a contract, to have been left open. In some such cases the effect of the existing law may be that the aggrieved party recovers nothing. A court, faced with arguments that there is no room for apportioning liability, may feel obliged to hold that the aggrieved party’s conduct was the sole cause, or the sole effective cause, of the loss.'

(d) Other examples of how unfairly the denial of such a plea would operate are suggested by the facts in O K Bazaars, quoted above or British South African Co v Lennon Bros Ltd 1913 SR 94. Contrast, with these cases the judgment of the Queen's Bench Division (per Brabin J) in De Meza and Stuart v Apple, Van Straten, Shena and Stone [1974] 1 Lloyds Rep 508 (Q.B.). There the defendant auditors were found to have been negligent, but also the instructing plaintiff, a firm of solicitors. Relying on a number of previous cases and well-known text-books, the learned judge found that the contract had imported a duty on the part of the auditors not to be negligent, and held that the English Law Reform (Contributory Negligence) Act of 1945 did apply. The fairness of this result speaks for itself.

(e) What is clear is that the corrective which the English law invented as a counter to the denial of a claim to a plaintiff who was himself negligent, namely, the last opportunity rule, formed no part of the Roman-Dutch law and was imported into this country from England. One of the motivations for the Act was undoubtedly to regularise the position in this country as far as the last opportunity rule was concerned. That could of course only be done if the principle of apportionment was introduced as a balancing factor. This was clearly a peculiarity of the law of delict. It may be that the Act was primarily concerned to rectify the kind of problem which occurred consistently in the law of delict and less so in the law of contract. But the ultimate question is not whether the Act was designed to rectify a problem of the law of contract but whether, in its terms and as a corollary of treating a problem surfacing more directly in the sphere of delict, it managed also to provide a satisfactory answer to a problem which, although it may have occurred less often in the law of contract, was nevertheless a real one.

(f) In the result I would apportion the loss suffered by TBA according to the standard laid down by s 1 (1) (a) of the Act. Having regard to the analysis of the facts by my colleague Nienaber JA, I am of the view that both parties were equally at fault, with the result that the claim of TBA

should be reduced by 50%.

(g) The effect of my approach is that the appeal succeeds but not to its full extent. In the court a quo TBA's claim was reduced by 80%. In my view it should only be reduced by 50%. This means that although its appeal succeeds only partly, it is successful in getting R694 909,00 instead of R249 279,69 awarded to it a quo. This represents substantial success, and TBA is therefore entitled to its costs of appeal.

As far as the remaining orders made by my colleague Nienaber JA are concerned, I would not suggest any changes.

P J J OLIVIER