Last week in commercial law: arbitration reviews, negotiating damages, and what Kenny Rogers can teach you about the SCA

[Fans of Kenny Rogers know that you’ve got to know when to fold ‘em, know when to walk away. Litigants seldom do. When they lose in the High Court, they ask the High Court for leave to appeal. When they lose leave to appeal in the High Court, they ask two judges of the SCA for leave to appeal. When they (again) lose leave to appeal, they ask the President of the SCA to refer the decision to the SCA for reconsideration under section 17(2)(f) of the Superior Courts Act. The President may refer the decision to the SCA for reconsideration “in circumstances where a grave failure of justice may otherwise result or the administration of justice may be brought into disrepute”. Let’s call that the statutory trigger for reconsideration. Who decides the statutory trigger for reconsideration? Does the President decide, or does the SCA decide if the President makes the referral? On the first interpretation, the President decides the statutory trigger and, if she decides the trigger is met, she refers the decision on the application for leave to appeal to the SCA (and so all that’s left for the SCA to decide is whether to grant leave to appeal; the SCA does not (re-)decide the statutory trigger before deciding the application for leave to appeal). On the second interpretation, the President decides the trigger and, if she decides that the trigger is met, she refers (presumably) the reconsideration application to the SCA (and the SCA must then (re-)decide whether the statutory trigger is met and only then decide the application for leave to appeal). Your summarist thinks the first interpretation better fits the text of section 17(2)(f)—after all, the text says the President refers “the decision” to the SCA, which must mean the decision on leave to appeal. But in 2024, the SCA unanimously adopted the second interpretation (that is, the SCA must (re-)decide if the statutory trigger is met before deciding the application for leave to appeal). The SCA unanimously affirmed the second interpretation in 2025. Defenders of the first interpretation dissented twice in 2025. The second dissent was unusually strident, describing the second interpretation as “a manifest absurdity” and a “defence of judicial formalism” (ouch!). But then plot twist! At the beginning of this year, the dissenting voices were vindicated, with the SCA unanimously holding that its previous unanimous judgments adopting the second interpretation were clearly wrong, and now adopting the first interpretation (that is, the President alone decides the statutory trigger, and the SCA then just decides the application for leave to appeal). The same judges then unanimously twice affirmed the first interpretation. Oddly, though, all three judgments upholding the first interpretation seemed to go on to ask whether there were “compelling and substantial factors … indicative of the existence of exceptional circumstances”, which sounds a lot like the second interpretation. But then a month later, another plot twist! And you’ll need to pay attention now: the SCA unanimously held that the three unanimous judgments of the SCA that departed from the previous unanimous judgments of the SCA upholding the second interpretation and instead adopted the first interpretation were themselves clearly wrong because those judgments were panels of 3 judges, and, apparently, a panel of 3 judges cannot overrule a panel of 5 judges. Your summarist thinks that’s wrong because a 3-judge panel of the SCA is still the SCA (and, apart from the difference between a single High Court judge and a full court, our law has never really distinguished between appellate panel sizes, unlike en banc courts in federal appellate courts in America, or larger panels in the UK Supreme Court). Last month, the SCA delivered its thirteenth judgment on reconsideration for the year (meaning that reconsideration has taken up a full 10% of its judgments this year). Splitting 4-1, the SCA reaffirmed the second interpretation and reaffirmed that a 3-judge panel can’t overrule a 5-judge panel. The dissent stuck to the first interpretation and disagreed with the majority on the 3-judge panel vs. 5-judge panel debate. Oh and the SCA is also divided on what the order should be if the statutory trigger is not met: a majority of the SCA says the proper order is to confirm the order dismissing the application for leave to appeal, another majority says the proper order is to strike the application from the roll, which was previously what a dissent said, but which differs from another dissent that would dismiss the application for reconsideration of the order refusing leave to appeal. As it happens, one of your summarist’s exciting hobbies is keeping detailed statistics of what’s happening at the SCA (more on this quarter’s statistics in next week’s edition). So far, the SCA’s unanimity rate this year is 82%. On that metric, reconsideration is, by some margin, the most divisive issue at the SCA—though, to be clear, your summarist has yet to see a reconsideration application that has actually succeeded, despite all this debate about the test]. With that background in mind, the SCA last week: “The jurisprudence of this Court on the approach to matters referred for reconsideration … is well established.” [Your summarist hopes his show-don’t-tell strategy has left you, dedicated reader, thinking that “well established” is maybe not the phrase you’d reach for if one day asked to comment on the SCA’s jurisprudence on reconsideration. But sometimes even the most enthusiastic keyboard commentator got to know when to fold ‘em, know when to walk away].

Neil Harvey and Associates developed software called Medware, which it licensed to Medshield, a medical aid administrator. Their relationship breaks down. Neil Harvey sues Medshield in arbitration, alleging that, amongst other things, Medshield stole its trade secrets and breached the licence agreement to copy some functions of Medware for its own software. Neil Harvey catalogues the Medware functions that it alleges Medshield copied in a document called C6. C6 amounts to one big inference: it shows that Medware has a feature or function, it shows that Medshield’s software has the same feature or function, and it infers that because Medshield has access to Medware, Medshield must have copied its homework. Taking a broad view of C6, the arbitrator finds that it’s not good enough evidence to support Neil Harvey’s case. Neil Harvey appeals to an arbitration appeal tribunal, arguing that the arbitrator was wrong to consider C6 in such a broadbrush way. The appeal fails. Neil Harvey reviews the appeal award, arguing that the tribunal failed to consider its ground of appeal about the arbitrator’s broadbrush approach to C6. Gauteng Division, Johannesburg: review dismissed because Neil Harvey never argued in its appeal to the tribunal that the arbitrator failed to properly consider C6 (and the tribunal decided all the grounds of appeal that were pleaded).

Atlantic iBuild appoints PDS Africa as its sub-contractor to drill pre-collar holes, whatever that might mean. PDS Africa drilled some pre-collar holes but Atlantic iBuild did not pay, so PDS Africa downs its drills. Atlantic iBuild claims it overpaid PDS Africa and, growing concerned by the worrying lack of holes being drilled, pre- or post-collar, threatens PDS Africa with contractual fire and brimstone if the drills do not start drilling again. PDS Africa applies to liquidate Atlantic iBuild. But then PDS Africa urgently applies for the appointment of an arbitrator (their contract has an arbitration clause), arguing that its claim against Atlantic iBuild will have prescribed by the time an arbitrator is appointed in the ordinary course. Western Cape Division: PDS Africa cannot have its cake and arbitrate it too; it’s an abuse of process for PDS Africa to want to arbitrate despite having chosen to litigate for Atlantic iBuild’s liquidation, so its application for the appointment of an arbitrator is struck [your summarist is sceptical of PDS Africa’s prescription argument for urgency, but this is as good a time as any to remind everyone that instituting arbitration proceedings does not interrupt prescription, unlike serving a summons: if a debt is “the object of a dispute subjected to arbitration”, then the completion of prescription is delayed until one year after the arbitration award (meaning that the clock still runs while you arbitrate, and you effectively have one year after the award to apply for its enforcement)].

Landlord terminates a lease and sues to evict its tenant. Tenant argues that they orally agreed to cancel the existing written lease and replace it with another lease for a few more years. KwaZulu-Natal Division, Durban: the non-variation clause says it covers variations, amendments, or cancellations, so Shifren stands in the way of the alleged oral cancellation.

And finally, exciting news from England on damages. Samsung makes smart watches. Those who buy smart watches want their smart watches to look cooler. So they go onto the Samsung app store and download apps that make their smart watch displays look like fancy real watches. Swatch owns a few of those fancy watch brands. It sues Samsung, and wins on the liability leg. At the quantum trial, Swatch wants $170 million; Samsung argues that $300 is more like it. The offending apps were downloaded about 160,000 times between 2015 and 2018. Most of the downloads were free: there were only about 600 paid downloads, generating about $1,000 for the app developers (and Samsung generally takes a third of app revenue, so roughly $300). Swatch bases its damages claim on something called negotiating damages (or Wrotham Park damages, following the favourite pastime of English lawyers: naming doctrine after cases). Under this type of damages, damages are assessed by reference to a hypothetical negotiation between the plaintiff and the defendant. The point of the hypothetical negotiation is for a court to guess determine the amount that the plaintiff could have received from the defendant in return for the plaintiff agreeing to release the defendant from the obligation in question. The English are still figuring out when negotiation damages apply, but they definitely apply to claims based on infringement of a property right or a closely analogous right, so infringement of IP rights is a core case (and in this context, negotiation damages are sometimes also called licence damages). More recently, in One Step, the Supreme Court clarified that negotiating damages are compensatory, not restitutionary [restitutionary meaning profit-stripping or disgorgement, not restitutionary in the way we usually use that word to mean reversing a transfer]. The thought experiment goes something like this: how much would Swatch have required from Samsung if Samsung had asked nicely to flog its IP on its app store? It’s all hypothetical, so it doesn’t matter if Swatch would never have agreed, and it doesn’t matter if Samsung would’ve made a loss. Chancery DivisionBusiness and Property Division: Swatch probably would’ve agreed to co-brand Samsung watches for a percentage royalty for each co-branded Samsung watch, giving a ballpark figure of $11 million.

Support Saflii.