Standard Bank of South Africa Ltd v Sibanda

• Jurisdiction: http://justis.com/jurisdiction/166,South Africa
• Judge: Meyer J
• Judgment Date: 28 November 2019
• Court: Gauteng Local Division, Johannesburg
• Hearing Date: 28 November 2019
• Citation: 2021 (5) SA 276 (GJ)
• Counsel: JC Viljoen for the plaintiff. I Oschman for the defendant.
• Docket Number: 38883/2011

Meyer J:

[1] The defendant, Mr Nkulumo Hopewell Sibanda (Mr Sibanda), conducted a current account at the plaintiff bank, Standard Bank of South Africa Ltd (Standard Bank). Mr Sibanda drew the amount of R448 179,71 by means of a cheque-card purchase before certain effects deposited to the account were cleared. Standard Bank honoured the withdrawal. Subsequently the effects were not paid. Standard Bank is suing Mr Sibanda, in contract or based on enrichment, for payment of the sum of R472 996,66, being the debit balance on the account that includes the said sum of R448 179,71.

[2] The hearing of evidence has been completed, but the presiding judge became indisposed and unable to deliver judgment. The parties agreed that a transcript of the proceedings, which include the evidence and Standard Bank's closing argument, together with the documentary exhibits and the parties' written heads of argument that were furnished to the trial judge, be placed before me for the delivery of judgment. In his letter of consent to this procedure being followed, Mr Sibanda stated: 'I have no financial muscle to restart these proceedings afresh.' I agreed to the procedure for the reasons that follow and invited the parties to address further oral argument to me, if they wished.

[3] Standard Bank instituted this action against Mr Sibanda on 12 October 2011. The trial commenced on 14 October and was concluded on 16 October 2015. Standard Bank called Mr Mark John van der Walt, a senior manager in its legal collections department, as its only witness. In defence, only Mr Sibanda testified. After the leading of evidence only Standard Bank's counsel presented an oral closing argument. Mr Sibanda elected only to furnish the trial judge with written heads of argument by an agreed date. It was further agreed that Standard Bank,

Meyer J

in addition to its oral closing argument, would also furnish the trial judge with written heads of argument by an agreed date. Due to the trial judge having become indisposed, the Judge President of this division allocated the matter to continue before me.

[4] Mondi Shanduka Newsprint (Pty) Ltd v Murphy2018 (6) SA 230 (KZD) is a matter that went to trial, all the evidence was led, argument was heard and judgment reserved. Before giving judgment the trial judge died. When the matter went back to court to continue before another judge (Lopes J), the parties agreed and requested the court to finalise the matter by the presiding judge reading all of the documents that would have been available to the deceased judge, hearing argument from the parties and then deciding the matter, rather than the trial beginning de novo. In refusing the proposal and directing that the trial begin de novo if the parties wished to continue with the matter, Lopes J said the following:

'[21] In my view none of the arguments advanced before me, nor the cases cited in favour of the matter being heard as sought by the parties, have provided a solution to the problem that matters of credibility cannot be dealt with in the manner suggested by the parties. There are numerous disputes of fact and expert opinion in the record of the proceedings, and a determination of those would be crucial to any decision.

. . .

[23] Whilst the parties may well place whatever evidence they wish before a civil court, the court still has to decide the matter on the applicable principles of law. Parties may, for example, agree that a certain fact can be accepted by the court as being true, when there is no documentary or viva voce evidence to support the finding of fact. In this way parties to civil actions may agree to limit, to some extent, the role of a judicial officer in determining matters. That is a very different thing, however, to parties being able to dictate to a judge how to exercise his oath of office by restricting the judge's adherence to legal principles, statutes, and precedent. Given the number of conflicts of fact and expert opinion in this case, I am of the view that a judge would not be able properly to determine the matter upon a mere reading of the record.

[24] It is also no answer to the above to suggest that one can simply apply the tests set out in Stellenbosch Farmers' Winery [Stellenbosch Farmers' Winery Group Ltd and Another v Martell et Cie and Others 2013 (1) SA 11 (SCA)] for the resolution of disputes. That is because the first two aspects referred to by the learned judge of appeal are the credibility of the factual witnesses and their reliability. The very fact that they cannot be decided merely on paper is recognised in Plascon-Evans and provides a limitation on the ability of judges to make such decisions, except in special circumstances. This matter is distinguishable from the situation where a case is part heard, and the judge may recall one or more witnesses (who have recently testified) in order to clarify any particular uncertainty.

[25] Were I merely to override those considerations, albeit with the consent of the parties, I have serious doubts as to whether I would be

Meyer J

fulfilling my oath of office by allowing the parties to a civil action to restrict the ordinary performance of my duties.'

[5] The premise of the finding in Mondi appears to be that the opportunity of a judge presiding at a trial to observe the demeanour of a witness is of great value in deciding whether or not to believe the witness's testimony. But that premise does not seem to be supported by relevant social science. On the contrary, as WH Gravett 'Spotting the Liar in the Witness Box — How Valuable is Demeanour Evidence Really?(1)' 2018 (81) THRHR 437 convincingly argues, that premise is contradicted by extensive empirical social-science data.

[6] Dr WH Gravett, who is a senior lecturer, Department of Procedural Law at the University of Pretoria, examined the value of a witness's demeanour as a guide to the truth of testimony in the light of a well-developed body of behavioural-science research spanning some seven decades. The extensive empirical evidence, as demonstrated by him, 'shows that demeanour — as a means of accurate reliable credibility assessment and decision-making in litigation — essentially is worthless'. Human lie detection, according to the learned author, 'is fraught with difficulty. It is predicated upon a multitude of misconceptions about how liars behave, including specific verbal and nonverbal cues commonly believed to indicate dishonesty.' The empirical research, according to the learned author, 'overwhelmingly demonstrates that ordinary people, including fact-finders, have no particular talent for spotting lies' and that this 'inability of most ordinary people to detect deception accurately has even greater implications in a heterogeneous society, such as ours, in which fact-finders often have to overcome racial and cultural differences in determining witness credibility'.

[7] The empirical research, according to the learned author, demonstrates that —

'(t)he traditional legal perspective on the evaluation of demeanour evidence is premised on four fallacies regarding how liars and lie-detectors behave: (i) that detecting deception in another is a matter of common sense; (ii) that liars betray themselves through certain telltale signs in their physical demeanour; (iii) that observers know which behavioural cues to look for in evaluating speakers' truthfulness; and (iv) that observers thus have a substantially better-than-average chance of catching liars'. [At 443 et seq.]

In conclusion he states (at 450):

'In short, there exists cogent evidence from social science studies that demonstrates that the concept of demeanour evidence as accepted by the law is invalid as it stands. In attempting to use a witness's conduct, manner, bearing ("demeanour") to assess that witness's credibility, most fact-finders will in fact rely on highly manipulative cues that mislead them, and will conclude that a witness is perjurious more often than they should.'

[8] The learned author also considered, inter alia, the reasons for the poor lie-detecting ability of people and why lie-detecting in court is even more difficult than in the laboratory, in his second instructive article on the topic: WH Gravett 'Spotting the Liar in the Witness Box — How

Meyer J

Valuable is Demeanour Evidence Really?(2)' 2018 (81) THRHR 563. According to the research referred to by him (at 563 – 4):

'(C)ertain signs of perceived deception, especially those involving the face, are also simply signs of nervousness and distress. It is almost impossible to distinguish between a person who experiences stress because she is guilty and on the verge of being exposed, and someone who experiences stress because she is innocent and stands falsely accused. . . . Yet, researchers have consistently found that observers attach meaning and significance to these behavioural cues of nervousness or anxiety even when the message is truthful. The mistaken interpretation of interrogation stress as deceit is so prevalent in the psychological literature that the phenomenon has come to be called "Othello's error" because it is excellently illustrated by Othello's mistaken interpretation of Desdemona's distress and despair in response to his accusation of infidelity.'

[9] I respectfully agree entirely with the learned author that 'by far the best determinant of the truth of testimony is not a witness's demeanour (visual or auditory behavioural cues) at all, but the actual content of the testimony', and that factors 'such as self-contradiction, inherent plausibility or the lack thereof, omissions and imprecisions, verification of facts testified to by other witnesses and exhibits, bias or motive on the part of the witness, and limitations of recall are among the most important indications of witness credibility', all of which would be readily discernible by a reading of a transcript of the evidence (at 566).

[10] Our highest courts have...