VO L 141 PART 4 (pp 635 — 846) 2024 VOL 141 PART 4 2024 FOUNDED 1884 Table of ConTenTs (overleaf) noTes Re(de)fining defamation by Emile Zitzke Minister of Water and Sanitation v Msukaligwa Local Municipality: Is financial incapacity a valid excuse? by Johandri Wright Lost in the fundamental contradiction: Revisiting Beadica by Jaco Barnard-Naudé Seismic surveys in South Africa’s maritime domain: The Sustaining the Wild Coast cases by Vishal Surbun & Paul Swanepoel arTiCles Treaties apply in South African law — Now what? Analysing the courts’ interpretation of treaties over the last half-decade by Andreas Coutsoudis A statutory duty to provide financial information at an early stage in family-law matters by Madelene de Jong & Elsje Bonthuys The dignity and justice of common purpose in criminal law by Khomotso Moshikaro & Catherine Willis-Smith Non-variation clauses by Jacques du Plessis book review Ninette Kelley People Forced to Flee: History, Change and Challenge (2022) by Sky Kruger ISSN 0258-2503 © Juta and Company (Pty) Ltd, 2024 PO Box 24299, Lansdowne 7779 Exclusive distributors in North America: Gaunt Inc, Gaunt Building, 3011 Gulf Drive, Holmes Beach, Florida 34217-2199, USA Tel: 941-778-5211. Fax: 941-778-525. E-mail: info@gaunt.com TYPESET BY ELINYE ITHUBA DTP SOLUTIONS PRINTED AND BOUND BY CONTENTS NOTES Re(de)fining defamation by Emile Zitzke .................................................................. 635 Minister of Water and Sanitation v Msukaligwa Local Municipality: Is financial incapacity a valid excuse? by Johandri Wright .................................................. 652 Lost in the fundamental contradiction: Revisiting Beadica by Jaco Barnard-Naudé ..................................................................................... 666 Seismic surveys in South Africa’s maritime domain: The Sustaining the Wild Coast cases by Vishal Surbun & Paul Swanepoel .............................................. 685 ARTICLES Treaties apply in South African law — Now what? Analysing the courts’ interpretation of treaties over the last half-decade by Andreas Coutsoudis ...... 703 A statutory duty to provide financial information at an early stage in family-law matters by Madelene de Jong & Elsje Bonthuys ............................................... 748 The dignity and justice of common purpose in criminal law by Khomotso Moshikaro & Catherine Willis-Smith ................................................................... 771 Non-variation clauses by Jacques du Plessis .......................................................... 804 BOOK REVIEW Ninette Kelley People Forced to Flee: History, Change and Challenge (2022) by Sky Kruger ..................................................................................................... 839 This book is copyright under the Berne Convention. In terms of the Copyright Act, No. 98 of 1978, no part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without permission in writing from the Publisher. SALJ 2024 Issue 4 (Journal).indb 1SALJ 2024 Issue 4 (Journal).indb 1 2024/11/29 12:452024/11/29 12:45 SALJ 2024 Issue 4 (Journal).indb 2SALJ 2024 Issue 4 (Journal).indb 2 2024/11/29 12:452024/11/29 12:45 EDITORS Managing Editor Professor Graham Glover BA LLB PhD (Rhodes) Editors Emeritus Professor Hugh Corder BCom LLB (Cape Town) LLB (Cantab) DPhil (Oxon) Professor Jacolien Barnard LLB (Pretoria) LLM (UNISA) LLD (Pretoria) Professor P J Schwikkard BA (Wits) LLB LLM (Natal) LLD (Stellenbosch) Associate Professor Helen Kruuse BA LLB LLM PGDHE PhD (Rhodes) Associate Professor Carika Keulder LLB LLM LLD (Pretoria) EDITORIAL ADVISORY BOARD Professor Lawrence Baxter, Duke Law School, North Carolina, USA Professor Katharina Boele-Woelki, Bucerius Law School, Hamburg, Germany Justice Azhar Cachalia, former Justice of the Supreme Court of Appeal, South Africa Professor François du Bois, Leicester University, UK Professor John Dugard, Leiden University, The Netherlands Professor David Dyzenhaus, University of Toronto, Canada Professor Christopher Forsyth, Cambridge University, UK Justice Avinash Govindjee, High Court, Eastern Cape Division, South Africa Justice Carole Lewis, former Justice of the Supreme Court of Appeal, South Africa Justice Mbuyiseli Madlanga, Constitutional Court, South Africa Gilbert Marcus SC, Advocate, Johannesburg Bar, South Africa Justice Mandisa Maya, Chief Justice, Constitutional Court, South Africa Justice Kate O’Regan, Emeritus Justice, Constitutional Court, South Africa and Director of the Bonavero Institute of Human Rights, Oxford University, UK Professor Cheryl Saunders, Melbourne University, Australia Professor Danie Visser, former Deputy Vice Chancellor, University of Cape Town, South Africa Professor Reinhard Zimmermann, Max Planck Institute, Hamburg, Germany SALJ 2024 Issue 4 (Journal).indb 3SALJ 2024 Issue 4 (Journal).indb 3 2024/11/29 12:452024/11/29 12:45 EDITORIAL POLICY The South African Law Journal is a peer-refereed journal which publishes original contributions on all fields of law. It provides a forum for scholars and practitioners, from South Africa and elsewhere, to reflect on issues that are internationally significant and locally relevant. The SALJ aims to be essential reading for those inside and outside South Africa who wish to keep abreast of the development of the South African legal order and its relationship to legal issues internationally. The SALJ is published four times a year. SUBMISSION OF MATERIAL FOR PUBLICATION All material for publication, including articles, recent case notes, corre- spondence, notes and comments, is to be sent to the Managing Editor, Professor Graham Glover. e-mail g.glover@ru.ac.za Books for review are to be discussed with the Book Review Editor, Emeritus Professor Hugh Corder. e-mail hugh.corder@uct.ac.za Subscriptions are to be addressed to orders@juta.co.za GUIDELINES FOR THE SUBMISSION OF MANUSCRIPTS Authors are requested to consult the section ‘To contributors’ at the end of this volume, where the conditions under which manuscripts will be considered for publication, as well as the rules of style to which authors should adhere, are set out. Please note that submissions not in house style will not be considered for publication. ABSTRACTS Authors of articles are required to prepare an abstract of no longer than 200 words. The abstract should summarise rather than introduce the argu- ment of the article, and should contain appropriate key words. Authors of notes are required to prepare an abstract of no longer than 100-150 words, briefly summarising the note, and to provide appropriate key words. SALJ 2024 Issue 4 (Journal).indb 4SALJ 2024 Issue 4 (Journal).indb 4 2024/11/29 12:452024/11/29 12:45 635 VOL 141 (Part 4) 2024 THE SOUTH AFRICAN LAW JOURNAL NOTES RE(DE)FINING DEFAMATION EMILE ZITZKE† Associate Professor of Law, University of the Witwatersrand This note is about the definition of common-law defamation. The authoritative definition of common-law defamation (the wrong ful and intentional publication of a defamatory statement concerning the plaintiff) is weighed against the five general elements of the common law of delict (conduct, damage, fault, wrong fulness, and causation) to determine to what extent the general and specific elements cohere. It is argued that the time has come to alter the definition of common-law defamation (slightly) to give a more accurate account of what courts do in defamation cases, which would also ensure greater unity between general and specific elements for liability. Defamation – definition – delict INTRODUCTION In Khumalo v Holomisa 2002 (5) SA 401 (CC) para 18, the South African common-law delict of defamation was defined according to five elements: ‘(a) the wrongful and (b) intentional (c) publication of (d) a defamatory statement (e) concerning the plaintiff’. Once elements (c) to (e) are established, rebuttable presumptions of (a) and (b) arise, triggering the need for the alleged wrongdoer to raise defences targeting wrongfulness and intent. This definition was regarded as ‘obvious’ by the same court years later in Le Roux v Dey (Freedom of Expression Institute and Restorative Justice Centre as Amici Curiae) 2011 (3) SA 274 (CC) paras 84–5 and was again more recently endorsed in Reddell v Mineral Sands Resources 2023 (2) SA 404 (CC) para 39. † https://orcid.org/0000-0001-5288-8679. I would like to thank Maggie Benjamin from the Wits Law Library for research assistance. Troy Charman, C J Visser and Andrea Zitzke provided useful feedback on earlier drafts. The Oppenheimer Memorial Trust and the NRF are acknowledged for the funding of this research. I am, of course, the factual and legal cause of any errors. ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 636 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 These definitional elements also enjoy academic support in the major commentaries that have appeared since Khumalo: F D J Brand ‘Defamation’ in W A Joubert (founding ed) The Law of South Africa vol 14(2) 3 ed (2017) para 111; Max Loubser & Rob Midgley (eds) The Law of Delict in South Africa 3 ed (2017) 412; J Neethling, J M Potgieter & A Roos Neethling on Personality Rights (2019) 199; J Neethling & J M Potgieter Law of Delict 8 ed (2020) 400. (The last two publications refrain from using the term ‘defamatory effect’ and instead speak about the effect of undermining reputation, but they cite Le Roux, seemingly with approval, so at first glance it looks like mere creative paraphrasing — a point to which I will return in more detail below.) The definitional elements of the common law of defamation seem to be uncontroversial and settled. In this note, I aim to start a conversation about whether the time has come for us to reconsider these definitional elements. The problem that I identify and grapple with here is that the ‘specific elements’ of defamation do not fully cohere with the so- called ‘general elements’ of the common law of delict. To explain what I mean by this, I will first explain the origin of the general/specific elemental dichotomy in delict. Secondly, I will show that the general and the specific do not properly align in our common law of defamation. Thirdly, I will explore what a proper alignment might look like. In this process, I must necessarily determine whether such an alignment is appropriate. Even though I am no fanatical generalist, I will show that a stronger alignment between general and specific elements for liability may be defensible. This will lead me to crafting a refined definition of common-law defamation. THE FIVE GENERAL ELEMENTS OF A COMMON-LAW DELICT The modern textbooks on the common law of delict, which deal with the discipline in a holistic way, usually start by declaring that there are five general elements to a delict (see J C van der Walt and J R Midgley Principles of Delict 4 ed (2016) para 3; Loubser & Midgley op cit at 23–4; Neethling & Potgieter op cit at 4). These elements are supposed to apply across the common-law board to all types of injuries. The five general elements are: (1) damage (covering both patrimonial and non-patrimonial damage); (2) conduct (acts and omissions); (3) fault (or culpability, which has a prerequisite of capacity for fault and substantively includes intention and negligence); (4) wrongfulness; and (5) causation (which has both factual and legal sub-components). Different thinkers have arranged these elements in different orders. Some have preferred conduct as the starting point because it is the element on which all the other elements depend (see Neethling & Potgieter op cit at 27). A different starting point might be the element of damage, NOTES 637 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 which is necessarily the keyword that takes one’s mind to the law of delict (see Loubser & Midgley op cit at 75). Other debates have concentrated on the order between fault and wrongfulness (see the overview provided in Neethling & Potgieter op cit at 155–7). These intricacies of ordering aside, the existence of the five elements is almost universally accepted as true in the major holistic works on the common law of delict. The five general elements have recently been adopted holus bolus by the Constitutional Court in De Klerk v Minister of Police 2021 (4) SA 585 (CC) para 13. They were partly adopted earlier in Oppelt v Department of Health, Western Cape 2016 (1) SA 325 (CC), where conduct was left out in the set of elements that establish a delict (para 34) but later implicitly regarded as important for purposes of the causation enquiry (paras 47–9) and wrongfulness (para 51ff), where the rules are modulated, depending on whether an act or omission is involved. There are, however, at least three conceptual complications to the five- element approach to the common law of delict. I will briefly elaborate on each of these complicators in turn. The first complicator is that some elements have sub-components that might be treated as bodies of doctrine in their own right. For example, in Geldenhuys v Minister of Safety and Security 2002 (4) SA 719 (C) at 737C, Davis J said that there are six elements of liability. He split fault into two, keeping its prerequisite of capacity for fault as a separate element from the substantive forms of fault, namely intention and negligence. Anton Fagan in Aquilian Liability in the South African Law of Delict (2019) x identifies seven elements to this specific delict by splitting (what others might call) damage into both harm and loss and splitting causation into its factual and legal sub-components. Be that as it may, the essence of the five elements is still present in a case such as Geldenhuys and literature such as that of Fagan. The second complicator is that some doctrines allow for a circum- vention of a rigid adherence to the five elements. In cases of strict liability, the element of fault is said to fall away (see the foundational work of J C van der Walt ‘Strict liability in the South African law of delict’ (1968) 1 CILSA 49). Popular examples of strict liability recognised in our common law of delict include damage caused by animals (where the owners of animals may be held liable regardless of whether they acted with fault: see eg Neethling & Potgieter op cit at 435–40) and vicarious liability (where the employer is held liable even if the employer had no personal fault: see eg ibid 444–54). The easiest way to deal with the doctrine of strict liability is to label it as an exception to the general rule that five elements must be observed (ibid 435). The takeaway message in this regard is that there are some, though limited, exceptions to the general-/ five-elements approach. 638 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 The third complicator is that our law of delict has a blend of generalist and casuistic elements. In Roman law, a series of delicts (theft, robbery, damage to property, and personality injury) each had their own (very) peculiar elements for and consequences of liability (see Inst 4.1 pr). There were no clear general elements for liability in Roman law. To the contrary, a strong casuistic approach to delicts was followed (see Paul du Plessis Borkowski’s Textbook on Roman Law 6 ed (2020) 322). The requirements for liability under the lex Aquilia for property damage seem rather far removed from non-agrarian, modern democratic societies. The owning of slaves and four-footed grazing animals, as well as their slaying, burning, breaking, and fracturing, all seem like bizarre delineations for property-injury liability today. Those requirements were much more pedantic and intricate than the simple requirements of intention plus personality-right violation under the classical Roman actio iniuriarum (see Neethling, Potgieter & Roos op cit at 67). Understandably, a general theory of delictual liability may have been difficult to construct during that time. In Roman-Dutch law, attempts were made to provide an overarching definition of a delict. Hugo de Groot and Johannes Voet are two useful examples of Roman-Dutch thinkers who attempted generalising definitions of a delict. De Groot (De Iure Belli ac Pacis 2.17.1) defined a delict as a culpable act or omission in breach of a duty which naturally attracts an obligation to repair the damage. Joe Sampson, in The Historical Foundations of Grotius’ Analysis of Delict (2017), reads De Groot’s work holistically. Sampson (ibid at 230–1) explains that culpa, as De Groot used it, covers both intention and negligence — effectively, the South African concept of fault. Furthermore, Sampson shows that De Groot’s general principles of delict clearly contain an implicit element of causation linking the acts or omissions and damage (ibid at 27). Considering this, we see the first signs of our modern five elements of conduct (acts and omissions), fault (intention and negligence), wrongfulness (arguably and roughly a breach of a duty), rudimentary causation, and damage. Voet’s Commentarius (47.1.1) provided another general definition of a delict. He thought of a delict as words or deeds that injure another through the intentional or negligent breach of a right. Whispers of conduct, fault, and damage are also present in Voet’s conceptualisation. What is interesting about both De Groot and Voet is that, despite their best efforts to provide generalised definitions of delict, they continue to deal with the different historical Roman delicts separately, where different types of injuries have nuanced requirements and defences. These Roman-Dutch influences made their way to South Africa. Watermeyer J in Perlman v Zoutendyk 1934 CPD 151 at 155 said: NOTES 639 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 ‘Roman Dutch Law approaches a new problem in the continental rather than the English way, because in general all damage caused unjustifiably (injuria) is actionable, whether caused intentionally (dolo) or by negligence (culpa).’ To understand the continental/English terminology employed in Perlman, it is useful to understand that civil codes had proliferated on the European continent since the 1800s by the time the judgment was handed down. Those civil codes tended to define a delict in an overarching way with reference to general elements for liability, though, sometimes, these were followed by more concrete and/or special rules that would apply to different types of damage (for a useful overview of some modern civil codes and their provisions on delict see Mauro Bussani, Anthony Sebok & Marta Infantino Common Law and Civil Law Perspectives on Tort Law (2022) ch 2). The continental approach of generalisation stood in contrast with the English system of tort law, where different types of injuries are compartmentalised and labelled as different torts, each with unique requirements (ibid at 9). In South Africa today, we seem to have aspects of Roman casuistry and continental generality at play. We have academics and apex courts supporting the idea of five general elements in the common law of delict, canvassed above. At the same time, it is common to find remnants of the Romanist tradition in operation. For example, lawyers still speak about ‘Aquilian liability’ when they speak about damage to property, or ‘instituting the actio iniuriarum’ or ‘Germanic action’ when claiming for a personality violation. This is partly because of tradition and partly because certain types of injuries have required the development of certain peculiar rules to accommodate them. That is also why people speak about ‘claims for pure economic loss’, ‘liability for omissions’ or ‘breach of a statutory duty’ almost as if they are separate delicts, while in fact they all ultimately fall under the scope of Aquilian liability. The reason why they almost seem like different delicts is because of the varying ways by which wrongfulness is determined in each of those different constructions. So, as people who think about and work with delict, we continuously navigate the interplay between the general and the specific. Continuous alignment of the specific within the broader structure of the general is a theme that has ensured some degree of coherence within the law of Aquilian liability and claims for pain and suffering under the Germanic action. It is less clear whether the five general elements play a benchmarking role in cases of personality violations under the historical actio iniuriarum. While Neethling & Potgieter immediately come to mind as strong supporters of the generalist school of thought in their treatise on the general principles of the common law of delict, two elements seem to fall by the wayside when dealing with personality rights (op cit at 5–6). Neethling, Potgieter & Roos op cit at 71, relying 640 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 on Roman-Dutch authorities and national appeal-level court decisions such as Minister of Justice v Hofmeyr 1993 (3) SA 131 (A) at 154C, define an iniuria as ‘the wrongful and intentional infringement of an interest in personality’. Conduct and causation are either implicitly present or totally absent. Perhaps a statement of law that is closer to the five elements was provided by the same court a few years earlier in Delange v Costa 1989 (2) SA 857 (A) at 860I–861A and regarded as ‘firmly entrenched in our law’: ‘Melius de Villiers The Roman and Roman-Dutch Law of Injuries at 27 notes three essential requisites to establish an action for  injuria. They are: “I. An intention on the part of the offender to produce the effect of his act; II. An overt act which the person doing it is not legally competent to do; and which at the same time is III. An aggression upon the right of another, by which aggression the other is aggrieved and which constitutes an impairment of the person, dignity or reputation of the other.”’ A clearer reference is made to conduct as ‘an overt act’ in this extract, and perhaps one could argue that the constitution of an impairment is a rough expression of causation. However, all of this must be read against the new backdrop created by the Constitutional Court’s 2021 judgment in De Klerk (supra). There, the court endorsed the five general elements of a common-law delict (para 13) and proceeded to spell out how those elements play out in the context of the iniuria of wrongful arrest and detention (para 14): the damage is an interference with liberty; the fault requirement here is intention; wrongfulness must be established; and the conduct in question must be the factual and legal cause of the victim’s damage. I acknowledge that the court did not lay it down as a hard rule that all iniuriae now have five elements. But this decision does provide a springboard from which to consider whether the elements for common-law defamation align in any way with the general elements for common-law delictual liability. THE FIVE GENERAL ELEMENTS OF A COMMON- LAW DELICT AND THE FIVE SPECIFIC ELEMENTS OF DEFAMATION To recap briefly, a common-law delict is generally said to comprise (1) damage; (2) conduct; (3) causation; (4) fault; and (5) wrongfulness. The common-law delict of defamation, in particular, involves (a) publi- cation; (b) about the plaintiff (which we can also call ‘reference’ for short); (c) defamatory effect; (d) wrongfulness; and (e) intention. Let us consider how the academic commentaries have linked the general and specific elements. Brand’s LAWSA entry does not attempt to link the element of common- law defamation to the general elements of a delict. He does not do this when he first provides the definitional elements (para 111), and he does not do this in his more intricate discussions of the elements that follow NOTES 641 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 (para 112ff). I am not saying that he should have undertaken this type of conceptual exercise. After all, this is an encyclopaedic text about a specific delict, not the law of delict in general. Neethling, Potgieter & Roos op cit at 199 make some attempt at linking the general elements and those of defamation: conduct (or ‘the act’) manifests as publication; damage plays out as the ‘defamatory nature of the words of behaviour’ (emphasis in the original); wrongfulness is the ‘impairment of the personality right to a good name’; and fault is the animus iniuriandi. While they define defamation in a way that seems to align in substance with Khumalo and Le Roux, their systematisation of the elements is somewhat different from what the Constitutional Court has laid down. The first striking observation is that Neethling, Potgieter & Roos virtually conflate wrongfulness and damage. As their discussion on defamation unfolds, the authors do not deal with the element of damage as something distinct from wrongfulness at all. This might be because of Neethling & Potgieter’s popular approach to wrongfulness as the unreasonable infringement of a subjective right (op cit at 55–9). In my view, this is an unnecessary conflation of elements. One could ask why defamatory effect, on its own, is not the most appropriate placeholder for the general element of damage. From here, wrongfulness would be determined only through the relevant defences, which then, as a separate enquiry, evaluates the reasonableness of the infringement of the right in question. A clearer distinction between wrongfulness and damage, as I have just explained it, was also drawn by Manfred Nathan The South African Law of Torts (1921) 93 (cf his more elaborate and convoluted definition, which goes beyond mere elements for liability, in The Law of Defamation (Libel and Slander) in South Africa (1933) 17). We are also not told in Neethling on Personality Rights exactly where the element of reference fits in and whether it relates to causation in any way. When discussing the element of publication, Neethling, Potgieter & Roos op cit at 203 mention in passing that certain people involved in the publication chain will only be deemed ‘publishers’ (under this element) if they had subjective knowledge of the communicated content or if they should have reasonably foreseen the ensuing reputational injury. The reasonable foreseeability consideration in this regard, they say, is an application of legal causation. Why this is an expression of legal causation is not made clear to us. It is a difficult conceptual fit under the ambit of the element of conduct (or ‘the act’) which, as Neethling & Potgieter in their general commentary on delict (op cit at 27–8) explain, must be understood as something objective (based on human voluntariness) and separate from the state of mind of the alleged wrongdoer. Where factual causation fits in, and whether other theories about legal causation may be applied instead, is similarly left uncertain. 642 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 Loubser & Midgley do the best job at explicitly linking the general and specific elements for liability. They suggest that publication satisfies the conduct requirement (op cit at 417), defamatory effect equals damage (op cit at 421), reference does the work of causation (op cit at 430), and wrongfulness and intention are clear enough. Loubser & Midgley’s approach seems to do the trick rather neatly. However, as I am about to show, some more nuance is missed in this linking. In what follows, I will present an argument about the proper linking and labelling of the elements of defamation. Seeing that wrongfulness and intention are patently present in the general elements and the elements for defamation, I will pay closer attention to conduct, causation and damage. CONDUCT: ACTS AND OMISSIONS The commentaries of Loubser & Midgley, as well as Neethling, Potgieter & Roos, suggest that the element of publication in defamation law constitutes the general conduct requirement. In Le Roux (supra) para 85, publication was defined as ‘the communication or making known to at least one person other than the plaintiff. It may take many forms. Apart from the obvious forms of speech or print, the injurious information can also be published through photographs, sketches, cartoons or caricatures.’ According to this authoritative definition, publication involves positive conduct (‘making known’) that results in its communication to a third party. While I have no problems with how publication is defined as a communicative phenomenon, I am not sure that it is necessarily the best descriptor of the conduct requirement. Several reasons can be identified why it might be sensible to reposition publication in our definitional elements of defamation. Most often, publication occurs through positive conduct — sending an email, typing and posting a comment on a social network, writing a defamatory print media article, drawing a defamatory cartoon, and so forth. Controversially, on at least two occasions our courts have recognised, at least by implication, liability for defamation on account of omissions. In Dutch Reformed Church Vergesig v Sooknunan 2012 (6) SA 201 (GSJ), the defendant created and controlled a Facebook page on which community members could post content. Some community members posted apparently unlawful material on that page, but the controller failed to remove that content. The court held that the controller was a ‘publisher’ of the content posted by others because he was duty-bound to remove unlawful content as the one who had created the opportunity for others to post their thoughts (para 49). Effectively, the Facebook page controller was held liable for his failure to remove defamatory content authored by another — perhaps an omissio per commissionem. NOTES 643 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 The following year, in Isparta v Richter 2013 (6) SA 529 (GP), the first defendant authored a defamatory Facebook post about a victim. That wrongdoer tagged the second defendant in that post. Roughly, tagging is a practice by which one Facebook user links the name of another Facebook user to a post containing text, image and/or video. If a third-party observer saw this, that third party would be able to click on the name of the tagged party and be led to the tagged party’s Facebook profile. Furthermore, with the automatic privacy settings on Facebook, the Facebook friends of the tagged party would be able to see the post even though another person had authored it. The tagged party (the second defendant) in this case was held to be ‘as liable as’ the first defendant in the sense that, even though he did not create the defamatory content, he passively but knowingly allowed his name to be associated with the first defendant’s post (para 36). Essentially, the second defendant was held liable for failing to un-tag himself from the first defendant’s post. Reading the judgment holistically, there is no evidence that the second defendant had committed any positive act. He was held liable for his omission. There is an article-length contribution to be written about the recog- nition of defamation by omission and the boundaries of that, especially in terms of the presumptions. I will postpone that exercise for a later occasion. Let us, however, accept for a moment that our law does recognise publication by omission after Sooknunan and Isparta. This is certainly accepted by Loubser & Midgley op cit at 418 as the current legal position: ‘In circumstances where there was a duty to act positively, a person can also publish by omission, for example, where an internet service provider fails to remove defamatory material from its servers after being requested to do so. In this case, the publication element is satisfied.’ Even if this reading of Sooknunan and Isparta is wrong, it is not bizarre to suggest that defamation by omission could be actionable. There is English precedent to support the recognition of defamation by omission that Loubser & Midgley rely on to make their claim (see eg Byrne v Deane [1937] All ER 204). Recognising defamation by omission may also be desirable in this time of social network pervasiveness, where content creators may be anonymous and virtually impossible to unmask. Passive parties with the ability and authority to minimise the damage caused by such anonymous postings (by, for example, removing the content) may need to be duty-bound to prevent the ongoing damage in certain clearly defined circumstances so that the victims are not left completely remediless. The difficulty with all of this is that the term publication necessarily presupposes some kind of positive action. As a matter of plain legal language, it seems like we are stretching the word beyond its logical limits. While lawyers and lawmakers can give random definitions to words beyond ordinary meaning, there is a more logical way to express what happens in cases of defamation by omission. Consider the following 644 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 analogue derived from Minister of Safety and Security v Van Duivenboden 2002 (6) SA 431 (SCA): when a police officer negligently fails to confiscate a violent gunman’s firearm while being duty-bound to do so, and the gunman proceeds to kill and injure innocent civilians, we would say ‘the gunman has shot the victims’. We would not say ‘the police have shot the victims’. We could say that ‘the police officer wrongfully caused the shooting of innocent civilians’, but we would not, when using plain language, say that the police officer is a shooter. Taking this back to defamation by omission, we could argue that passive wrongdoers can cause the publication of defamatory content and thereby contribute to or exacerbate the damage suffered by the victim. I am not convinced that we can call them publishers in the ordinary use of the word. Two conclusions may be drawn from the observations made thus far about publication and conduct. First, defamation can arguably arise from both acts and omissions. Perhaps the general element of ‘conduct’ could be a suitable placeholder to describe what we are dealing with. That would also be useful in the sense that we would be incorporating the general principles pertaining to conduct (like voluntariness) in a defamation suit. While most cases of defamation are surely committed by people who can control their bodily movements, I can see no good reason why a defence of automatism should not be available to an alleged wrongdoer, at least in principle. Secondly, to accommodate the wide range of persons that could be involved in a reputational-damage-causing event, the element of causation could be meaningfully introduced, as I will argue in due course. This does not mean that the concept of publication becomes irrelevant. Instead, it should be repositioned under a different element, namely damage. DAMAGE: PUBLICATION OF DEFAMATORY CONTENT ABOUT THE PLAINTIFF Traditionally, the rules of common-law defamation have been understood as being designed to protect a person’s right to reputation. In more recent times, scholars have examined the gist of defamation, suggesting now that it is about protecting people from: the risk of reputational damage (see Francois Jurgens Risk as Injury: An Alternative Interpretation of the South African Law of Defamation (PhD thesis, University of Cape Town, 2019)); disrespect and truth-blurring about one’s character (see eg Anton Fagan ‘Defamation, reputation and respect’ in P J Schwikkard & S V Hoctor A Reasonable Man — Essays in Honour of Jonathan Burchell (2019) 37)); or unjust contempt (see eg Khomotso Moshikaro ‘Unjust contempt as the gist of defamation law’ (2022) 12 Constitutional Court Review 59). In this short note, I will not attempt to resolve the debate about the gist of defamation law because it does not directly affect the labelling and/or ordering of the definitional elements of the claim. While there are radically different NOTES 645 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 angles of approach to defamation, all the above theorisations are based on the common law of defamation as it stands. Significantly, for present purposes, they all conceptualise the damage as a violation of a victim’s interest that occurs in the public domain beyond the relationship between the victim and the wrongdoer. All these theorisations hold that publication (communication to a third party) distinguishes defamation from the iniuria of insult and is intimately tied up with the nature of the damage involved. In the reputation school, one’s reputation is violated in the eyes of third parties. In the disrespect/ contempt school, the way in which third parties engage in relationships with us is at stake. Therefore, publication is a constituent of the damage rather than the conduct. We could turn to an Aquilian analogue to illustrate this point. Whether a police officer actively shoots an innocent victim or fails to confiscate a firearm of a trigger-happy man, the damage to the victims is the same. Whether an act or omission is at stake may have implications for other elements, such as the rebuttable presumptions about wrongfulness, but the conduct does not really have a bearing on damage. Turning back to defamation, the communication of content to a third party (publication) is fundamental to proving the type of damage such that it is distinguished from mere insult. On my interpretation, publication seems to have much greater relevance to damage than conduct — which also sits comfortably with what I have argued about a more generic descriptor of conduct above. On my reading of the theories and the case law, the other constituent part of the element of damage is whether the published material had defamatory effect on the plaintiff. Following Le Roux (supra) para 89, by ‘defamatory effect’ we mean that a reasonable observer would interpret the content and, based on the meaning derived (either the obvious primary meaning or a special secondary meaning), would conclude that it has the effect of violating the victim’s fama. (I deliberately use the ambiguous Latin phrase fama here because it could, in theory, keep the proponents of both the reputation and respect/contempt schools happy.) By ‘reference’ we mean that the content, upon interpretation by the reasonable observer, would be understood to be about the plaintiff (see eg Isparta (supra) para 23). I would utter defamatory effect and reference in one breath because those two are inextricably linked. After all, we use the same test of the reasonable observer to test for both defamatory effect and reference. To illustrate my point, it may be useful for us to turn again to an analogue. If a police officer tortured a victim, we can ask whether the victim suffered damage. On the non-patrimonial front, we could show that the victim has endured pain and suffering, disfigurement, psychological injuries, and loss of amenities of life. On the patrimonial front, we would consider the victim’s medical bills, loss of earning capacity and so forth. We do not 646 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 ask whether these forms of damage have been suffered in the air (or by an elusive ‘someone’) and then attach it to a specific victim. The damage is adjoined to the victim. Similarly, in a defamation dispute, there is little use in determining that a cartoon or text has defamatory effect ‘of an elusive someone’ and only then considering the more concrete question whether it is defamatory of the particular victim. Defamatory effect and reference go together. We even use the same interpretative test when analysing the allegedly offending material to determine defamatory effect and reference — the reasonable observer test. For these reasons, it would make a lot more sense to group defamatory effect and reference together under the ambit of damage. In summary, I would argue that damage, in a defamation case, has three sub-components: ‘publication’ of content that is ‘about the plaintiff’ and has ‘defamatory effect’. Interestingly, C F Amerasinghe Defamation and Other Aspects of the Actio Iniuriarum in Roman-Dutch Law (in Ceylon and South Africa) (1968) 9 also grouped these three elements together as the factum of a defamation suit. At this point, one might ask why reference does not amount to causation, as Loubser & Midgley contend. Causation, as I will explain in more detail below, is not about linking the damage to the victim. It is about linking the victim’s damage to the wrongdoer’s conduct. FACTUAL AND LEGAL CAUSATION Causation does not feature as a clear element in the accepted definition of common-law defamation. My thinking about causation in defamation is prompted by Jonathan M Burchell’s definitional elements set out in The Law of Defamation in South Africa (1985) 35. He defines common- law defamation as ‘[t]he unlawful, intentional publication of defamatory matter (by words or conduct) referring to the plaintiff, which causes his reputation to be impaired’. Later in the book, he recognises that this element does not feature much in the courts, but it is, in theory, essential (ibid at 114–15). As a general element in the common law of delict, causation is made up of two sub-components — factual causation and legal causation — the former being a precondition for the latter (see Lee v Minister for Correctional Services 2013 (2) SA 144 (CC) para 38). The factual causation enquiry aims to determine whether the victim’s damage and the alleged wrongdoer’s conduct are connected. We can determine factual causation with the good old ‘but-for’ test or, where justice requires some more flexibility, the increased-risk test or material-contribution test (see Alistair Price ‘Factual causation after Lee’ (2014) 131 SALJ 491; Mashongwa v PRASA 2016 (3) SA 528 (CC) para 65). Legal causation concentrates on the factual connection between the victim’s damage and the alleged wrongdoer’s conduct. It aims to determine NOTES 647 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 whether the link is close enough for liability to ensue or whether the relationship is too remote (Lee (supra) para 38). What reasonableness, fairness and justice require in relation to closeness or remoteness is given concrete expression through various tests that must be applied flexibly (see eg Fourway Haulage SA v SA National Roads Agency 2009 (2) SA 150 (SCA) paras 33–4). The tests include proximate cause, foreseeable consequences, adequate causation, intended consequences, and the like (see Van der Walt & Midgley op cit para 181ff; Loubser & Midgley op cit at 128ff; Neethling & Potgieter op cit at 230ff). While causation is not typically expressly invoked in the defamation context, I have already suggested above that if defamation by omission is possible, an element of causation would be conceptually useful. But there is a further reason too. Some of causation’s related concepts make guest appearances in our extant defamation law, especially when determining whether an alleged wrongdoer should be labelled as a publisher in the existing discourse. After defining publication as the communication to a third party, Neethling, Potgieter & Roos op cit at 203 write that ‘if it is clear that publication has taken place, the plaintiff has to show that the defendant was responsible for the publication’ (emphasis in the original). Loubser & Midgley op cit at 417 and Brand op cit para 113 make a similar point about responsibility for publication. According to Neethling, Potgieter & Roos, responsibility is determined with reference to whether the publication was actually foreseen or should have been reasonably foreseen. Those authors argue that this is an expression of legal causation’s foreseeability test. The case law seems to support this line of reasoning, and I could find no critiques of this line of reasoning in any of the major contemporary works on defamation. The issue of foreseeability has especially arisen in cases where an unexpected third party has received information intended for another recipient. Four cases deserve mention in this regard. In Hall v Zietsman (1899) 16 SC 213, a letter containing strong words about the victim’s behaviour was mailed to the victim. However, the victim’s bookkeeper opened the letter and read it. While dealing with an exception and the elements of the delict of defamation, the court held that an essential component of the element of publication is that the alleged wrongdoer must be responsible for the third party opening the letter ‘or that he knew that it would be opened by [the bookkeeper]’ (ibid at 216). In Martin v Kemlo (1909) 26 SC 457 at 459, the court explicitly endorsed and applied this approach from Hall. There, a wrongdoer wrote a defamatory letter about the victim. It was placed in an open envelope addressed to the victim, which was then placed in another envelope addressed to the victim’s clerk and delivered to the clerk’s private residence on a Saturday. The court held that the facts indicated that 648 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 the wrongdoer intended for the clerk to read the letter, so the wrongdoer was responsible for the publication (ibid at 461). Arguably, the court developed the Hall principle even further by looking at who the wrongdoer intended as the audience. Facts similar to Hall were seen in Van Vliet’s Collection Agency v Schreuder 1939 TPD 265 at 269, where the court held: ‘I think in a case of this description it is essential for the plaintiff both to allege and to prove affirmatively either knowledge or reasonable expectation on the part of the defendant that, in the circumstances, publication would be a likely result of his conduct and of his dealing with the defamatory matter complained of.’ In Pretorius v Niehaus 1960 (3) SA 109 (O) at 112, the dictum from Van Vliet was applied in a situation where a wife opened a letter addressed to her husband. She had opened the letter because it was not marked ‘private and confidential’. The court held that there was no evidence that the wrongdoer knew or reasonably could have expected that the wife would open the letter (ibid at 116A–B). Neethling, Potgieter & Roos’s claim that actual or reasonable foreseeability fulfils the role of legal causation in cases like this makes sense. In Hall, Martin, Van Vliet, and Pretorius, there was no dispute that the alleged wrongdoers factually caused the publication of defamatory content about the victims. While the language of actual or reasonable foreseeability may lure us into thinking that we are dealing with fault, the courts in all these cases are clear that we are dealing with responsibility for publication. Responsibility is traditionally used as the jurisprudential backdrop for the element of causation, tying together the victim’s damage and the wrongdoer’s conduct. Inserting an element of causation (factual and legal) into the definition of common-law defamation will simply give effect to what is already implicit in the case law. It will not (or at least should not) complicate much litigation in this area. In many cases, causation will simply not be in dispute. But, in those cases where it is, we will have the proper legal tools to deal with it systematically. This is especially so if we accept that our courts are moving in the direction of imposing liability for defamation by omission, where multiple actors could be involved in the causal chain of events that leads to a victim’s damage. On social networks today, many people could be involved in a common-law defamation dispute. There is, for example, the social network platform (which might be covered in theory by the exemptions in chap 11 of the Electronic Communications and Transactions Act 25 of 2002, although this is not likely in practice given the requirement that they must be registered with a representative body, of which there are few currently in existence), the content creator, those who have liked or otherwise reacted to a post, commenters on posts, sharers of posts, and so forth. They could all potentially play a factual role, in some way, NOTES 649 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 in the distribution of the content. Whether each one of those persons should be held liable as a matter of reasonableness, equity and fairness is a completely different question. At the moment, it seems that we are all too quick to say that they are ‘publishers’ without considering the factual and legal causation issues more robustly (see eg Lee Swales & Sizwe Snail ka Mtuze ‘Freedom of expression and the Internet’ in S Papadopoulos & S Snail ka Mtuze (eds) Cyberlaw@SA: The Law of the Internet in South Africa 4 ed (2022) 395ff; Sanette Nel ‘Freedom of expression’ in Dana van der Merwe (ed) Information and Communications Technology Law 2 ed (2016) 492ff). Space does not allow me to take this point further here, but we should be thinking more seriously about the causation dimension of these cases in the future. One way to ensure that is to introduce causation as a distinctive element in the definition of common-law defamation. To conclude on causation: if foreseeability plays a legal-causation role in a defamation suit, in terms of general principle, it does leave one wondering why we do not approach the causation issue in a more robust and systematic way. We surely need to establish factual causation more carefully before we jump to legal causation. The but-for, material-contribution, and increased-risk tests could all be usefully employed. We should also be thinking and reasoning more clearly through the legal causation enquiry, not necessarily limited to the foreseeability test. As the court wisely cautions in Fourway Haulage (supra) para 34, none of the legal causation tests should be approached ‘exclusively and dogmatically’. In other words, there is no magic formula for establishing legal causation. If we insist on foreseeability as the beginning and end of the legal causation enquiry, I am concerned that we would be closing our minds to the possibility that justice, fairness and reasonableness might require alternative lines of reasoning in future. Saying this will not radically disrupt the law. On the contrary, it will ensure greater consistency of principle in our law. WRONGFULNESS AND INTENTION Wrongfulness and intention are mirrored in the general elements and the elements for defamation. For this reason, I will not say much about these elements. Traditionally, rebuttable presumptions of wrongfulness and intention arise once a publication about the plaintiff with defamatory effect has been proven. Practically, wrongfulness and intention play out as defences. Defences against wrongfulness (grounds of justification) traditionally include truth and public interest, protected comment, privileged occasion, and the media’s reasonable publication of an untruth. Defences against intention traditionally include jest and mistake. Because wrongfulness and intention usually play out as defences, R G McKerron The Law of Delict 7 ed (1971) 170 did not include wrong- ful ness and intention as separate elements in his definition of defamation. 650 (2024) 141 THE SOUTH AFRICAN LAW JOURNAL ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 Instead, he speaks about the absence of ‘lawful justification or excuse’ as an essential element that ultimately fulfils the same role. In my proposed redefinition, wrongfulness and intention will remain elements of common-law defamation. I say this being fully conscious of the fact that there are significant debates about the nature of wrongfulness and fault in the common law of defamation. Three major controversies are worth flagging here. First, there have been arguments about introducing liability for negligent defamation alongside the traditional animus iniuriandi. Some have given the case of National Media Ltd & others v Bogoshi 1998 (4) SA 1196 (SCA) an overbroad reading to this effect (see Neethling, Potgieter & Roos op cit 244–6). However, the court in Bogoshi at no point said that it was redefining defamation to be the ‘wrongful and negligent publication of a defamatory statement about the plaintiff’ in cases of media defendants. The court simply introduced a defence against wrongfulness that allows mass media defamation to be excused on account of the reasonable publication of an untruth and clarified that in those specific circumstances, to prevent the futility of this new defence, the media cannot get away with saying that it lacked animus iniuriandi. This narrower reading of Bogoshi is also the interpretation that the same court recently supported in EFF v Manuel 2021 (3) SA 425 (SCA) paras 41–8. While some High Court decisions point in the direction of negligence-based liability for defamation (see eg Marais v Groenewald 2001 (1) SA 634 (T); Pieterse v Clicks Group 2015 (5) SA 317 (GJ)), this is not yet the law confirmed by a national appeal-level court. If this becomes law at some point, a more generic placeholder of ‘fault’ could be used instead of intention. However, for the moment, on the law as it stands, intention should remain part of the elements of the common law of defamation. Secondly, the definition of animus iniuriandi has also been a site of contestation, with some uncertainty prevailing about whether it involves the direction of will and consciousness of wrongfulness, or only the direction of will. The issue of consciousness of wrongfulness need not be taken further here because whether we accept it as a constitutive part of intention or not, the fault threshold, for the moment, remains intent. Thirdly, Anton Fagan, in ‘Wrongfulness in the South African law of defamation’ (2023) 140 SALJ 285, has recently argued that intention is a fundamental requirement for wrongfulness. He expresses the general test for wrongfulness in common-law defamation as follows (ibid at 291): ‘A defendant who made a defamatory statement about a plaintiff to a third party acted wrongfully by doing so if and only if: (1) the statement represented the plaintiff to the third party as having a worth which is less than the worth which the plaintiff ought to be estimated to have by the third party; and (2) the defendant intended this.’ NOTES 651 ht tps://doi.org/10.4734 8/SAL J/v141/i4 a1 He clarifies (ibid at 307) that his alternative approach to wrongfulness in defamation ‘places intent “within”, rather than “without”, wrongfulness. It treats intent as a constitutive element of wrongfulness, as one of its essential ingredients, rather than as a requirement of liability additional to, and separate from, it. This flies in the face of a long-cherished scholarly dogma about the nature of, and the relationship between, wrongfulness and fault in the South African law of delict.’ This explanation helps us to make sense of wrongfulness in defamation when dealing with positive conduct (ie ‘a defendant who made a defamatory statement’) where presumptions of wrongfulness and intent arise, but it is less clear how we would go about determining wrongfulness in cases of omissions where presumptions of intent and wrongfulness might not be equally appropriate. In those cases, we should probably require the victim to establish wrongfulness and intention as we do in cases of Aquilian liability rather than requiring the wrongdoer to rebut presumptions of the same. Furthermore, while we wait for greater clarity on the precise definition of animus iniuriandi and the related questions as to whether defences like jest and mistake will continue to exist in our law, it may be useful to keep intention as a separate element that necessarily informs wrongfulness but can also stand in its own right. Thus, for the purposes of an overarching definition of common-law defamation, I would keep intention and wrongfulness as separate elements while recognising that, in certain cases, wrongfulness effectively swallows intention, as Fagan argues. TOWARDS A REFINED DEFINITION OF COMMON-LAW DEFAMATION Considering the argument developed in this note, the elements of common-law defamation should be: (i) conduct that (ii) causes (iii) the publication of defamatory content about the plaintiff that is both (iv) wrongful and (v) intentional. I would argue that this definition refines the one that the Constitutional Court has provided and endorsed over the years, in line with further precedent. This refined definition should not be read as a radical departure from the existing law, but rather one based on it. It has the potential to ensure a more logical, systematic, and lucid treatment of concepts. It also happens to provide greater coherence within the common law of delict as a holistic discipline. TO CONTRIBUTORS The editors welcome the submission of manuscripts in English for consideration for publication. A manuscript will be considered for publication: l only on the assurance that it has not in whole or in part or in substance been published or offered for publication elsewhere l on the understanding that it may be submitted in confidence to an expert referee or expert referees for evaluation l on the understanding that the editors reserve the right to make what changes they consider desirable: (a) to bring the manuscript into the house style of the South African Law Journal (b) to eliminate errors of typing, grammar, syntax, punctuation, spelling, idiom and the like (c) to eliminate ambiguity, illogicality, tautology, circumlocution and redun­ dancy (d) to produce accuracy and coherence (e) to improve the mode of expression and style of writing (f) to avoid possible criminal or civil liability. Authors are required to read their manuscripts very carefully to avoid the need for the editors to exercise these rights extensively. Note in particular that every quotation and every reference is to be carefully checked for accuracy. The editors will try to check once more, but are not always able to do so, and accuracy is to be sought at all times. In particular, authors are asked to acquaint themselves with the house style of the SALJ, and to make the effort to place any submission in the style of the SALJ. A comprehensive guide to what is required appears on the following web page: http://www.jutajournals.co.za/south-african-law-journal Authors are encouraged to examine this document carefully, and to comply with its requirements. Manuscripts which do not show any attempt to comply with the house style will, regrettably, be returned to authors with a request to do so before the substance of the manuscript will be considered. SALJ 2024 Issue 4 (Journal).indb 847SALJ 2024 Issue 4 (Journal).indb 847 2024/11/29 12:462024/11/29 12:46 VO L 141 PART 4 (pp 635 — 846) 2024 www.juta.co.za