A bi-annual update complementing the Commentary on … · A bi-annual update complementing the Commentary on the Criminal Procedure Act ... Is it time to jettison the rules in R v

A bi-annual update complementing theCommentary on the Criminal Procedure Act

NO 1 OF 2017

ANDREW PAIZES, Author (Editor and Author)STEPH VAN DER MERWE, Author

Contents

Editorial Note.............................................................................................................................................. 3

(A) FEATURE ARTICLES ....................................................................................................................... 4Circumstantial evidence and inferential reasoning: Is it time to jettison the rules in R v Blom? ....... 4Perspectives on parole: Two recent cases ............................................................................................ 8What is the meaning of ‘hearsay evidence’ in s 3(4) of the Law of Evidence Amendment Act 45 of

1988?.............................................................................................................................................. 12

(B) LEGISLATION .................................................................................................................................... 17The Justice Administered Fund Act 2 of 2017..................................................................................... 17Implementation of the Rome Statute of the International Criminal Court Act Repeal Bill

[B23—2016]: Withdrawal by Minister.......................................................................................... 17Children’s Amendment Act 17 of 2016 ............................................................................................... 17Judicial Matters Amendment Bill [B14—2016] .................................................................................. 18Criminal Procedure Amendment Bill [B2—2017] .............................................................................. 18

Proposed amendments to s 77(6)(a)(i) .......................................................................................... 18Proposed amendments to s 77(6)(a)(ii) ......................................................................................... 18Proposed amendment to s 78(6) .................................................................................................... 19Proposed amendments to s 79........................................................................................................ 19Sections 77, 78 and 79: ‘mental defect’ to be replaced by ‘intellectual disability’ ...................... 20

(C) CASE LAW .......................................................................................................................................... 21(a) Criminal Law.................................................................................................................................. 21

Is a person who assists another to commit suicide guilty of murder? Physician administeredeuthanasia and physician assisted suicide considered ............................................................ 21

Constitutional validity of statutory offences prohibiting use of cannabis for personal consump-tion and related matters ........................................................................................................... 23

(b) Criminal Procedure and Evidence.................................................................................................. 25(i) Pre-sentence .......................................................................................................................... 25

ss 7 and 8: Private prosecution and s 6(2)(e) of the Societies for the Prevention of Crueltyto Animals Act 169 of 1993........................................................................................... 25

s 60: Bail: Right to a fair bail hearing................................................................................... 27ss 151 and 166: Onus of proof—implications of. Judicial officer descending into the

arena—how to deal with inexperienced prosecution..................................................... 28s 152: Reconstruction of the record: Can the accused waive his or her rights to participate

in the process?................................................................................................................ 30s 208: Identification evidence guidelines and pitfalls ........................................................... 31ss 217, 219, 166: Admissibility of confessions under the Criminal Procedure Act and the

Constitution.................................................................................................................... 33(ii) Sentencing ............................................................................................................................. 34

Sentencing and trial fairness: No reference in charge-sheet to minimum sentencelegislation ....................................................................................................................... 34

s 276B: Procedural fairness where court considers a non-parole period .............................. 36s 280(2): Sentencing court’s discretion to order concurrent running of sentences............... 37

(iii) Forfeiture and confiscation ................................................................................................. 38s 20: Sections 48(1) and 50(1) of the Prevention of Organised Crime Act 121 of 1998:

When is residential property an ‘instrumentality of an offence’ for the purpose of aforfeiture order? ............................................................................................................. 38

(iv) Appeal and review................................................................................................................ 40Appeal: Requirement that court of appeal gives notice of possible increase of sentence .... 40Appeal: Disposal of an appeal in terms of s 19(a) of the Superior Courts Act 10 of 2013 . 42

TABLE OF CASES .................................................................................................................................... 43

2

Editorial Note

There are three feature articles in this edition. Thefirst examines critically the so-called ‘cardinal rulesof logic’ set out in the locus classicus in the area ofcircumstantial evidence, R v Blom 1939 AD 188,which were, once again, relied upon by the SupremeCourt of Appeal in a recent decision. The question isasked whether these rules, in spite of their firm graspon the judicial imagination, are now more a hin-drance than a help, and whether the time has nowcome to abandon them in favour of a more sophisti-cated and nuanced approach.

The second considers two recent cases which exam-ine the legal duty of the executive to address mattersrelating to parole in a manner that is consistent withlegal principles and requirements: the first caseconcerns the duty of the executive to ensure that itsdecision to release a sentenced offender on parole isbased on adequate information; the second the dutyto follow fair administrative action in any decisionby the executive to refuse parole.

The third feature article concerns the meaning of‘hearsay evidence’ in the Law of Evidence Amend-ment Act 45 of 1988 and challenges the view,expressed in a recent case, that this definition ismuch the same as that which applied at common law.

The section on legislation describes various statutoryprovisions but the focus is mostly on the CriminalProcedure Amendment Bill, which seeks to makeimportant amendments to Chapter 13 of the CriminalProcedure Act. This chapter deals with mattersrelating to an accused’s mental illness, criminalresponsibility and the capacity of an accused tounderstand proceedings. The discussion includes aconsideration of the extent to which the legislationsuccessfully addresses constitutional objectionsraised by the courts in respect of the existingprovisions. The section covers legislation up to 12June 2017.

The case notes, as usual, cover a wide range ofissues. Two very interesting questions relating to thesubstantive criminal law are considered. First, is aperson who assists another to commit suicide guilty

of murder, and is there a difference in this contextbetween ‘physician administered euthanasia’ and‘physician assisted suicide’? Second, are the statu-tory offences prohibiting the use of cannabis in smallquantities for purely personal use valid in the light ofthe constitutionally protected right to privacy?

Among the procedural issues canvassed, the follow-ing may be mentioned:

• When is residential property an ‘instrumentalityof an offence’ for the purpose of a forfeitureorder in terms of the Prevention of OrganisedCrime Act (POCA)?

• When can a statutory juristic body institute aprivate prosecution?

• In what circumstances will bail applicationproceedings be rendered unfair by remarksmade by a judicial officer that compromise hisor her impartiality?

• Can an accused waive his or her rights toparticipate in the process of reconstructing alost record of the trial proceedings?

• What practical guidelines should a court followin avoiding the pitfalls of (a) identificationevidence; and (b) evidence of a confession?

• Does the failure to warn an accused prior toconviction that he or she faces a possiblesentence of life imprisonment render the trialunfair?

• What considerations should a court take intoaccount in ensuring trial fairness when it con-siders imposing a non-parole period as part ofthe sentence?

• When should a court exercise its discretion toorder a concurrent running of sentences in orderto avoid inhumane or unfair sentences?

• What happens if a court of appeal fails in itsduty to give an appellant notice of a possibleincrease of sentence?

• When may a court of appeal dispose of anappeal without the hearing of oral argument?

Andrew Paizes

3

(A) FEATURE ARTICLES

Circumstantial evidence andinferential reasoning: Is it time tojettison the rules in R v Blom?The recent decision of the Supreme Court of Appealin S v Mahlalela [2016] ZASCA 181 (unreported,SCA case no 396/16, 28 November 2016) raisessome important and interesting issues concerning theappropriate mode of reasoning when inferences aresought to be drawn from primary or intermediatefacts. The locus classicus in this regard is themuch-cited decision in R v Blom 1939 AD 188, inwhich Watermeyer JA set out what he described (at202–203) as ‘two cardinal rules of logic’ whichcould not be ignored when it came to reasoning byinference:

‘(1) The inference sought to be drawn must beconsistent with all the proved facts. If it isnot, then the inference cannot be drawn.

(2) The proved facts should be such that theyexclude every reasonable inference fromthem save the one sought to be drawn. Ifthey do not exclude other reasonable infer-ences, then there must be a doubt whetherthe inference sought to be drawn is cor-rect.’

These cardinal rules have become so entrenched inour law that there is hardly a case on circumstantialevidence that does not cite them or purport to rely onthem for guidance. They have achieved a status notfar short of canonical imprimatur, to the extent thatone hesitates to criticise them for fear of beingaccused of heresy.

But it cannot be denied that the rules in Blom are farfrom perfect or, even, satisfactory. They beg morequestions than they answer; they are unclear inimportant respects; they cannot be rules of ‘logic’ inany meaningful sense; they seem, in so far as theyare rules of ‘law’, to find application (at least thesecond rule) in the criminal law alone; and yet theyseem to be imperfect servants of the rules governingthe onus of proof. They are, in my view, misleadingand dangerous and they lead to wrong answers tocomplex questions.

Such sharp criticism will surely be viewed by someas heretical or, at least, misconceived. But I remainunrepentant. I have subjected the rules in Blom tocareful scrutiny in The South African Law of Evi-dence (Zeffertt and Paizes, 2 ed (2009) at 100–13),and have levelled a number of accusations at them.

At the very least, it must be accepted that the rulesraise a number of questions that are very difficult toanswer with any precision or certainty and whichgive the lie to the notion that they at least provide asimple or helpful direction for untying an intricateknot.

These are some of those questions: (1) What consti-tutes a ‘proved fact’? (2) What standard of proof is tobe deployed in determining whether a ‘fact’ hasindeed been ‘proved’? (3) Is there only one suchstandard, to be used as a universal solvent, or arethere different standards to be used in differentsituations? (4) Does the standard—if, in fact, it doesvary from one situation to another—depend onwhether the proceedings are criminal or civil? (5)Does it depend, further, on whether the ‘fact’ reliedupon as a ‘proved fact’ is a ‘primary’ or an ‘interme-diate’ fact? (6) Does the consideration that the ‘fact’rests on inferential reasoning drawn from a largenumber of prior (primary or intermediate) facts, asopposed to merely one or two, signify at all? (7) Isthe second rule in Blom a restatement or an expres-sion of the onus of proof in criminal proceedings (ieproof beyond a reasonable doubt)? (8) If so, how canthe rule be one of logic and not of law? (9) And,further, how can it be made to apply to civilproceedings? (10) And, further still, how does itsapplication in criminal cases vary, if at all, in respectof issues directly related to guilt (such as theessential elements of the offence and other closelyrelated issues) and those much less directly related?

A consideration of all these questions leaves therules very vulnerable to attack. In The South AfricanLaw of Evidence I have not hesitated to shine a lighton their frailties and shortcomings in an attempt toweaken their grip on the imagination of our judgesand commentators. But their ubiquity in judicialpronouncements, age, succinct formulation, apparent(albeit deceptive) simplicity and kinship with thecriminal onus of proof, have given them an aura thatwill be very difficult to dispel. Who will be boldenough to shake the firmament or, even, give it agentle prod? What messy, complex, tangled web ofinter-related ideas, notions and questions will beunearthed if we do so? Can we bear to engage withthe complexity that will emerge if we expose therules for what they really are: an anodyne thatprovides some comfort but fails to cure or treat thereal disorder?

I will not repeat all the criticisms I have raisedelsewhere but will limit myself to one crucial aspectof those criticisms, which has been given fresh

4

impetus by what has been said in a recent decision ofthe Supreme Court of Appeal.

In S v Mahlalela [2016] ZASCA 181 (unreported,SCA case no 396/16, 28 November 2016), theappellant had been convicted of murder and robberywith aggravating circumstances. The case againsthim (and his two co-accused) had rested almostentirely on circumstantial evidence. A vital item ofcircumstantial evidence, received by the trial courtand relied upon by it in convicting the appellant, wasthat the appellant and one of the co-accused had,shortly after the robbery, each made deposits ofR2000 into their respective bank accounts on thesame day, at the same bank, and with the same teller.An amount of approximately R8000 had been takenin the robbery. The appellant did not testify at theclose of the State’s case.

Dlodlo AJA (with whom Shongwe, Van der Merweand Mocumie JJA and Potterill AJA agreed) was‘persuaded that the State did not prove beyondreasonable doubt that there was a factual basisentitling the trial court to draw inferences eitherintermediate or otherwise’ (at [19]). The discovery ofthe bank slips evidencing the two deposits did ‘notgo far enough to assist the State’ since the evidence‘hardly satisfie[d] the first rule of logic in Blom’:there was ‘a host of reasonable inferences to bedrawn from the discovery of these deposit slips’, andit did not follow, said Dlodlo AJA, that the depositedmoney belonged to the deceased who had beenrobbed of R8000 in cash.

If it were to be ‘accepted that these deposit slipsconstitute[d] a fact from which an inference of guilt[could] be drawn, that would’, said Dlodlo AJA,‘stretch inferential reasoning too far’ (at [20];emphasis added). The ‘proved facts’ envisaged inBlom were, according to Dlodlo AJA, ‘facts provedbeyond reasonable doubt’, and ‘intermediate infer-ences’ too, had to be based on such ‘proved facts’ (at[15]).

Moreover, said the judge (at [20]), for an inference tobe permissible, it ‘not only had to be based onproved facts, but also had to be the only reasonableinference from those facts to the exclusion of allother reasonable inferences’. The ‘ultimate question[was] whether, in the light of all the evidenceadduced at trial, the guilt of the appellant wasestablished beyond reasonable doubt’.

This judgment touches on several of the concerns Ihave expressed in The South African Law of Evi-dence and raises fresh anxieties as well. One of the

concerns expressed above related to the uncertainmeaning of a ‘proved fact’ and the standard of proofrequired to render a fact a ‘proved fact’ for thepurpose of the rules in Blom.

The court in Mahlalela, however, seemed to have nouncertainty at all. It spoke unequivocally about suchfacts having to be proved beyond a reasonable doubt,and made it quite clear that intermediate inferenceshad to be based on facts so proved. It is submitted,however, that such certainty was entirely misplacedand that an insistence on ‘facts’ being proved beyonda reasonable doubt in all cases before being able tosustain legitimate inferences, is wrong in principleand in law. In R v Mtembu 1950 (1) SA 670 (A) at679–80, Schreiner JA emphatically opposed anapproach that would require all intermediate facts ina criminal trial to be proved beyond a reasonabledoubt. He set out a broad test for determining whenproof on this standard would be necessary, but madeit plain that it would not always be required. He said:

‘I am not satisfied that a trier of fact is obligedto isolate each piece of evidence in a criminalcase and test it by the test of reasonable doubt.If the conclusion of guilt can only be reached ifcertain evidence is accepted or if certain evi-dence is rejected then a verdict of guilty meansthat such evidence must have been accepted orrejected, as the case may be, beyond reasonabledoubt. Otherwise the verdict could not properlybe arrived at. But that does not necessarilymean that every factor bearing on the questionof guilt must be treated as if it were a separateissue to which the test of reasonable doubt mustbe distinctly applied. I am not satisfied that thepossibilities as to the existence of facts fromwhich inferences may be drawn are not fitmaterial for consideration in a criminal case onthe general issue whether guilt has been estab-lished beyond reasonable doubt, even though, ifthe existence of each such fact were to betreated by the test of reasonable doubt, mereprobabilities in the Crown’s favour would haveto be excluded from consideration and mereprobabilities in favour of the accused wouldhave to be assumed to be certainties. Circum-stantial evidence, of course, rests ultimately ondirect evidence and there must be a foundationof proved or probable fact from which to work.But the border-line between proof and probabil-ity is largely a matter of degree, as is the linebetween proof by a balance of probabilities andproof beyond reasonable doubt. Just as a num-

5Issue 1, 2017

ber of lines of inference, none of them in itselfdecisive, may in their total effect lead to a moralcertainty (Rex v de Villiers (1944, A.D. 493 at p.508)) so, it may fairly be reasoned, a number ofprobabilities as to the existence of the factsfrom which inferences are to be drawn maysuffice, provided in the result there is no reason-able doubt as to the accused’s guilt.’

This dictum has been referred to and applied in anumber of cases (see, for instance, R v Manda 1951(3) SA 158 (A) at 166B–D, R v Sibanda & others1965 (4) SA 241 (SRA) at 246–7, S v Shepard &others 1967 (4) SA 170 (W) at 172–3 and S v Smithen andere 1978 (3) SA 749 (A) at 754–5; see, too, S vNtsele 1998 (2) SACR 178 (SCA) at 189c–d and S vMorgan & others 1993 (2) SACR 134 (A) at 172–3).However, it is, perhaps, not as well known, under-stood or even appreciated as it should be. That it isnot is probably due to the unfortunate ‘lines ofinterference’ emanating from the rules in Blom,which regrettably entangle the rules for reasoning byinference with those governing the criminal onus.That there are certainly instances where inferencesmay properly be drawn from ‘facts’ that are prob-able—albeit not proved at the level excluding rea-sonable doubt—is demonstrated by the decision inR v Manda (supra), where the appellant had beenconvicted of murder and evidence had been adducedat his trial that (a) the appellant had receivedscratches on his face at approximately the time thatthe deceased, a nine-year-old boy, had been killed;and (b) material had been scraped from under thefingernails of the deceased which was probablyhuman skin. The latter probability, said SchreinerJA, was something that could be used in arriving at aconclusion.

This perspective is one that the High Court ofAustralia reached only after a false start. In Cham-berlain v The Queen [No 2] (1984) 153 CLR 521,that court, in the famous ‘dingo trial’, insisted (asDlodlo AJA seems now to have) that all primaryfacts in a criminal case had to be proved beyond areasonable doubt, since ‘[n]o greater cogency can beattributed to an inference based upon particular factsthan the cogency that can be attributed to each ofthose facts’ (at 599; emphasis added). In a strongdissent, however, Deane J took a very different view,one that squares with the pronouncement ofSchreiner JA in Mtembu. In considering what stan-dard should be imposed for proving primary facts, hesaid (at 626–7):

‘In the view I take, it is impossible to give ageneral theoretical answer to that question.There is certainly no requirement of the law thatthe members of a jury must examine separatelyeach item of evidence adduced by the prosecu-tion and reject it unless they are satisfiedbeyond reasonable doubt that it is correct. Noris it the law that a jury is in all circumstancesprecluded from drawing an inference from aprimary fact unless that fact is proved beyondreasonable doubt. If a primary fact constitutesan essential element of the crime charged, ajuror must be persuaded that that fact has beenproved beyond reasonable doubt before he orshe can properly join in a verdict of guilty.Whether or not a juror must be satisfied that aparticular fact has been proved beyond reason-able doubt will, however, otherwise depend notonly on the nature of the fact but on the processby which an individual juryman sees fit to reachhis conclusion on the ultimate question of guiltor innocence. If, for example, the case againstan accused is contingent upon each of fourmatters being proved against him, it is obviousthat each of those matters must be provedbeyond reasonable doubt. Indeed, it would beappropriate for the presiding judge to empha-sise to the jury in such a case that even aminimal doubt about the existence of each ofthese matters would be greatly magnified in thecombination of all. On the other hand, if theguilt of an accused would be established by, or aparticular inference against an accused could bedrawn from, the existence of any one of twohundred different matters, each of which hadbeen proved on a balance of probabilities, itwould be absurd to require that a jury shoulddisregard each of them unless satisfied, either inisolation or in the context of all of the facts, thatany particular one of those matters had beenproved beyond reasonable doubt.’

Deane J’s exposure of the flaws in the majority’sreasoning prompted a quick retraction and about-face. In Shepherd v The Queen (1990) 170 CLR 573the High Court did what appellate courts often dowhen they know they have been wrong: maintainthat what they were understood to have said was notwhat they actually said. The majority in Chamber-lain, said the court in Shepherd, intended theirremarks to apply only to those intermediate factswhich were ‘a necessary basis for the ultimateinference’ (at 581). The court distinguished between

Criminal Justice Review6

situations where intermediate facts constituted‘indispensable links in a chain of reasoning towardan inference of guilt’, where proof of those factswould have to be on the criminal standard and wouldhave to exclude reasonable doubt, and situationswhere the evidence could better be described asconsisting of ‘strands in a cable rather than links in achain’ (at 579), where that standard would not apply.

The crucial question, then, in cases of inferentialreasoning, is this: Which form of reasoning betterdescribes the process one is dealing with—‘chain’reasoning or ‘cable’ reasoning? In answering thisquestion, these incidental questions may assist: Howimportant is the inferred ‘fact’ in the broader schemeof things and, in particular, in establishing theultimate issue of the accused’s guilt? How manyfacts are relied on to establish the inferred ‘fact’?The more facts that are relied upon, the less likely itis that each or any primary fact is important enoughto be decisive of the issue by itself, and the morelikely it is that guilt itself (or an element of it or anintermediate fact itself crucial to establishing guilt orsuch element) may be proved beyond a reasonabledoubt even though a particular primary fact or, even,each of them, is shown to be no more than probable.

Consider a murder trial where the court has taken theview that it has been shown beyond a reasonabledoubt that the accused had a motive for the murderand that the murder weapon belonged to him, butthat these facts still left the matter of his guilt shortof being proved beyond a reasonable doubt. Assume,further, that the court accepted it as no more than‘probable’ that the accused had been seen near thescene of the crime shortly before the time of themurder and had been inexplicably absent from hisplace of work during the crucial time period. Wouldnot the inference that the accused had, impelled byhis motive, used his weapon to kill the deceased, notbe significantly strengthened by this evidence—perhaps to the point of proving his guilt beyond areasonable doubt? McHugh J in Shepherd believed,correctly in my view, that it would.

The lesson to be learned is that it is surely naïve tobelieve that the facile rules expressed in Blom’s casecan effectively and comprehensively engage with allthe complexities and nuances presented by cases ofthis kind. In fact, far from helping, they get in theway of clear thinking and may obstruct in answeringthe most important question of all—whether theaccused’s guilt has indeed been proved beyond areasonable doubt.

To ensure that the correct answer is achieved inrespect of what Dlodlo AJA correctly described asthe ‘ultimate question’, there is, in my view, only onemethod that will work in all cases: to begin with theultimate question, work backwards, and ensure thatthe mode of reasoning by inference is, in eachinstance and at every level of the inferential process,sufficiently sound to secure proof of the ultimateissue (guilt) beyond a reasonable doubt.

Consider this example: X is charged with the murderof his wife, Y, and a man, Z. The State’s case is thatX killed both of them out of anger and jealousybecause Y and Z were having an affair. No othermotive is put forward and the State’s case restsheavily upon the intermediate fact that Y and Z were,in fact, having an affair. X’s guilt, in such a case,cannot be established beyond a reasonable doubt ifthis intermediate fact is not also proved beyond areasonable doubt. That fact, then, may be likened toa link in the probative chain.

Assume, further, that the State, in order to prove theexistence of this affair, calls as a witness the recep-tionist at a hotel who testifies that, a year before thetrial, she checked in at the hotel for a night a couplewho, to the best of her recollection, were Y and Z,even though they checked in under false names.Cross-examination reveals weaknesses in the identi-fication, and the court is prepared to accept no morethan that it is probably accurate. If, then, it isreasonably possible that it is not, can the inferencethat Y and Z were having an affair be drawn? If thisis the only evidence to support that inference, itcannot be drawn, as the existence of the affair cannotbe established beyond a reasonable doubt if there isreasonable doubt attaching to the only evidencetendered to prove it. However, the position would bequite different if the State could call several differentreceptionists, working at various hotels, who couldtestify to the same facts. Even if the accuracy of eachof these accounts was accepted as being no morethan probable, it could legitimately sustain an infer-ence that Y and Z were having an affair. If so, theintermediate fact (the affair) would be provedbeyond a reasonable doubt and the probative edificewould stand. The evidence of each of these witnessescould appropriately be likened to a strand in thecable, which, given the support of a sufficient num-ber of other similar strands, would be strong enoughto secure the necessary evidentiary connection toallow, at the end of the process, for proof of guiltbeyond a reasonable doubt.

7Issue 1, 2017

It is unprofitable, in such cases, to work ‘forward’from primary to intermediate to ultimate ‘facts’. Thatis one of the biggest flaws in the rules in Blom. Andthat is why the rules seek—in an unscientific andmuddled fashion—to tie the rules of inferentialthinking with the standard that encapsulates thecriminal onus. If we, instead, work ‘backwards’ fromthat standard, it has these benefits: First, it guaran-tees that we can never fail to meet that standard incriminal cases. Secondly, we are able to develop afar more nuanced approach to inferential reasoningthat depends on the amount of probative work eachintermediate fact is required to do as well as themode of reasoning (‘chain’ or ‘cable’) we areemploying to connect it with the fact (ultimate orother) we are seeking to establish.

The rules in Blom provide little guidance for chartingthe murky waters of inferential reasoning. Worse,still, they have led to or allowed for many miscon-ceptions. Some of these are: (i) All primary orintermediate facts have to be proved beyond areasonable doubt (Chamberlain and Mahlalela). (ii)Inferences may not be drawn from other inferences(Mahlalela at [15]). There is no reason why a wholeseries of cascading inferences cannot be upheldprovided the ultimate issue (guilt) is proved beyond areasonable doubt. The model described ensures thatit cannot fail to be, and care will be given to ensurethat the mode of reasoning (cable or chain) does notallow for a violation of what is a constitutionalentitlement—the right to be presumed innocent untilproven guilty. (iii) A court is ‘entitled to drawinferences but not to speculate’ (see, for instance, S vEssack & another 1974 (1) SA 1 (A) at 16D, S vCooper 1976 (2) SA 875 (T) and S v Nkuna 2013 (2)SACR 541 (GNP) at [11]). Why should a court notbe allowed to come up with an innocent explanation,one that was, for some reason, not furnished by theaccused (he or she may wish to protect another, ormay wrongly believe the true explanation to be

unwise because it is incriminatory in respect of alesser offence) if that ‘speculative’ explanation has areasonably possible chance of being true? To disal-low such ‘speculation’ would not only violate theaccused’s constitutional right (the presumption ofinnocence) but would disregard the authority of themajority judgment in R v Mlambo 1957 (4) SA 727(A). It is a strange irony that that case is usually citedmore for the attack on ‘speculation’ in the minorityjudgment of Malan JA than for the proposition,correctly accepted by the majority, that the appel-lant’s conviction for murder could not stand becauseof the reasonable possibility—not raised by theappellant—that he had unintentionally killed thedeceased in the course of a sexual assault (see TheSouth African Law of Evidence at 121–6). Labels, inmy view, are at best unhelpful and at worst danger-ous in this context. If there is a reasonable doubt thatthe accused is innocent, the reason for that doubt—whether it be speculation, conjecture or ‘solid evi-dence’—should not signify. The mere fact that itexists, does. This view is strengthened by the recentdecision of the Supreme Court of Canada in R vVillaroman 2016 SCC 33 at [35].

I do not believe that the court in Mahlalela reachedthe wrong result. The evidence in that case wasclearly insufficient to sustain the convictions. But theprocess of reaching that result was not facilitated bythe rules in Blom. Instead, those rules muddied thewaters even further by leading to unnecessary com-plications regarding the meaning of ‘proved facts’,the relation between the rules and the criminal onus,and a consideration of whether inference couldfollow inference.

It is time to see the rules in Blom for what they wereprobably intended to be: an early, tentative step on achallenging journey. We have been on this journeyfor long enough now to see more of the terrain andto understand more of its dangers. It is time to letthem go.

Andrew Paizes

Perspectives on parole: Two recentcasesMoses Parole in South Africa (2012) 10 describesparole as ‘a form of conditional release of prisonersinto the community, before the expiry of theirsentences, subject to, and on the basis of, an under-taking by the parolee that he will. . .comply withsuch conditions as. . .imposed and prescribed. . .’.

Our courts have over many years consistently main-tained that, in line with the principle of separation ofpowers, the decision to grant parole to a sentencedoffender is ‘the exclusive field’ of the executive. SeeS v Jimmale 2016 (2) SACR 691 (CC) at [11]; S vMhlakaza & another 1997 (1) SACR 515 (SCA) at521f–i. In S v Botha 2006 (2) SACR 110 (SCA)Ponnan JA said that courts do not have the power toprescribe to the executive branch of government how


long a sentenced offender should be in jail, ‘therebyusurping the function of the executive’ (at [25]). Acourt’s limited power to determine a non-paroleperiod as provided for in s 276B of the CriminalProcedure Act 51 of 1977 is an exception. Thissection came into operation only on 1 October 2004.See further the discussion of s 276B in Commentary,sv The judiciary and the executive, as well as thebrief discussion of S v Ndlovu [2017] ZASCA 26(unreported, SCA case no 925/2016, 27 March 2017)in Section C of this issue of CJR, sv ‘s 276B:Procedural fairness where court considers a non-parole period’. In all cases falling outside the ambitof s 276B, the executive enjoys the prerogative ofgranting parole. See generally S v Van Loggenberg2012 (1) SACR 462 (GSJ) at [13].

It is sometimes argued that parole and the system ofparole lead to unacceptable interference with ajudicial decision, that is, the sentence imposed by thecourt. See generally Van Heerden 2011 24(2) ActaCriminologica: Southern African Journal of Crimi-nology 17 at 26–7. However, the fact of the matter isthat parole serves as an incentive to prisoners torehabilitate themselves so that they can be releasedwithout having served their full prison sentence. Acourt is required to sentence an accused on the basisof all the relevant facts and without regard to thepossibility of parole. See S v Matlala 2003 (1) SACR80 (SCA) at [7]; S v Motloung 2016 (2) SACR 243(SCA) at [18]. The executive is better placed than thecourts to determine parole because release on parole‘requires the assessment of facts relevant to the. . .[sentenced offenders’]. . .conduct after the impo-sition of sentence’ (S v Madolwana (unreported,ECG case no CA&R 436/12, 19 June 2013) at [7],emphasis in the original).

The two cases discussed below provide insightfulperspectives on parole and the executive’s legal dutyto address parole matters in a manner consistent withlegal principles and requirements. The first casedeals with the duty of the executive to ensure that itsdecision to release a sentenced offender on parole isbased on adequate information. The second casedeals with a situation where the executive’s refusalof parole was set aside because of unfair administra-tive action.

Naidu v Minister of Correctional Services [2017] 2All SA 651 (WCC)

In the above matter the plaintiff claimed damagesarising from an attack on her by one MM who wason parole at the time of the attack. The attack, sheclaimed, was a direct result of the negligent release

of MM on parole—a matter for which the defendant,the Minister of Correctional Services, was liable. Atthe trial, counsel for the defendant made someconcessions: first, the plaintiff was attacked by MMas alleged in the particulars of claim; second, if itwere to be found that the defendant was negligent inreleasing MM on parole, such negligence was caus-ally linked to the harm suffered by the plaintiff; and,third, there was a legal duty on the defendant toensure the safety of members of the public, such asthe plaintiff, and this duty was sufficient to foundliability. The defendant’s legal duty, noted Meer J at[4], was akin to that relied upon in Minister of Safetyand Security & another v Carmichele 2004 (3) SA305 (SCA), a case which concerned liability for thenegligent failure to oppose bail for a dangerousoffender. At [47] Meer J also made brief reference tos 131 of the Correctional Services Act 111 of 1998(hereafter the ‘CS Act’). This section refers topatrimonial loss and provides as follows: ‘In theevent of a person serving community correctionsbeing liable in delict for an act or omission in thecourse of such service, the damages sustained maybe recovered from the State.’

Given the concessions as outlined above and the factthat the parties had agreed that the merits andquantum be separated, the only question waswhether the defendant—acting through the Brand-vlei Prison Correctional Supervision and ParoleBoard (hereafter ‘the Board’)—was negligent inreleasing MM on parole. According to the plaintiffthe Board was negligent in that it had failed to actwith reasonable care and diligence in determiningwhether MM could be released on parole: inadequateattention was given to the previous convictions ofMM and the fact that he had previously violated hisparole conditions (at [3]). The plaintiff also assertedthat in its assessment of MM as a candidate forparole, the Board had relied on inadequate informa-tion and had decided the matter in the absence ofcertain reports which the Case Management Com-mittee (hereafter ‘the CMC’) should in terms of theCS Act have submitted to the Board. One of the corefunctions of a CMC is to report to the relevant Boardon the possibility of placing a sentenced offender onparole.

But in Naidu it turned out that the CMC had failed toperform its statutory duty to place certain crucialreports and information before the Board hearingMM’s parole application. There was no report on themental state of MM (s 42(2)(d)(iii)). Vital informa-tion as required by s 42(2)(d)(iv) was also not

9Issue 1, 2017

furnished. This section requires a report on ‘thelikelihood of a relapse into crime, the risk posed tothe community and the manner in which this risk canbe reduced. . .’. There was also no report or informa-tion regarding ‘the assessment results and progressreport’ in respect of MM’s correctional sentence planas contemplated in s 38 and required bys 42(2)(d)(v).

The Board, in turn, also failed to comply with itslegal duty because it proceeded to assess MM’sparole application in the absence of the aboveprescribed information (at [48]). Clearly, in thesecircumstances no proper assessment could be madeof all the issues and risks involved in releasing aprisoner on parole.

Meer J had good reason to find that there was‘simply. . .no clear evidence that. . .[MM]. . .hadbeen rehabilitated and could be granted parole’ (at[51]). In this regard the court also relied on theevidence of one P, an expert witness called by theplaintiff. P qualified as an expert on account of hisexperience over two decades as a social workerspecialising in the rehabilitation of offenders (at[18]). The defendant did not call any expert.

In P’s opinion ‘the decision to release or not torelease on parole. . .had become a logistical consid-eration rather than an enquiry into rehabilitation andreadiness to be released into society’ (at [22]).P expressed the view that in his experience MMshould, on the information that was available to theBoard, not have been released on parole and thatfurther intervention by the Board was required for itto have determined MM’s parole readiness (at [23]).On the evidence before the Board, MM had in prisonattended two programmes—of only three days’ dura-tion each—on ‘Aggression and Life Skills’. Accord-ing to P, however, it was impossible to rehabilitate aperson with MM’s background after a few days ofgroup sessions. The defendant, testified P, usedattendance of a programme as the ‘yardstick’,whereas the true criterion was ‘how much impact theprogramme had on an inmate’ (at [23]). Meer Jconcluded that P’s ‘extreme scepticism about theprospects of a person with. . .[MM’s]. . .criminalprofile being rehabilitated by the mere attendance ofgroup sessions over a few days,. . .[was]. . .morethan warranted’ (at [51]).

MM’s ‘criminal profile’ consisted of ‘a long list ofprevious convictions dating back some 26 years to1980’ (at [7]). For present purposes only some ofthese convictions need to be identified as the perti-nent highlights in an impressive criminal career. In

November 1987 MM was sentenced to 12 years’imprisonment for murder. This sentence ran concur-rently with an earlier sentence of nine months’ directimprisonment imposed for assault with intent to dogrievous bodily harm. In December 1996 MM wasreleased on parole in respect of the murder sentenceimposed in 1987. Eleven months later and whilst stillon parole, he committed theft for which he wassentenced to four and a half years’ direct imprison-ment (at [7.12]). In 2004 MM was again convictedof, amongst others, two thefts, assault with intent todo grievous bodily harm and assault (at [8.1]–[8.2]).The result was that by the time he was placed atBrandvlei Correctional Centre he had a sentence ofnine and a half years’ imprisonment to serve, withMarch 2013 as the date on which his sentence was toexpire (at [9]–[10]). In 2007 MM was considered butnot recommended for parole (at [11]). On threedifferent occasions during the course of 2008, MM’sconduct as a prisoner was such that he was foundguilty of internal ‘disciplinary infringements’ asprovided for in s 23 of the CS Act (at [11]–[12]). InMarch 2010 MM was granted parole—some threeyears prior to his normal release date. And it waswithin four months of having been released onparole that MM assaulted the plaintiff in her home.

At [50] Meer J stated that the reasonable person inthe position of the Board and appraised of thefollowing three facts would, in the absence of clearevidence of rehabilitation, have foreseen the reason-able possibility that MM, if released on parole,‘would cause harm of the kind he ultimately causedto the plaintiff’ (at [50]): first, MM’s record indicatedthat he was ‘a habitual violent criminal who washistorically not rehabilitated by time spent in prison,nor by early release on parole’ (at [50.1]); second,MM had previously violated parole conditions whilstserving a sentence for murder (at [50.2]); third, after2004 MM had committed offences in prison and ‘hisaggression was flagged on more than one occasion’by prison officials (at [50.3]). At [53] Meer Jconcluded:

‘Given the absence of evidence that. . .[MM-]. . .had been rehabilitated, the Board ought inthe circumstances to have taken reasonablesteps to guard against the foreseeable harmof. . .[MM’s]. . .release on parole, by refusinghis parole application. Failure to do so was anact of negligence. A further act of negligencewas the failure to comply with the mandatorystatutory requirements of. . .[s 42(2) of the CSAct].’


In Naidu the defendant’s negligence ‘was causallyconnected to the harm suffered by the plaintiff’ (at[53]). The plaintiff, it can be argued, was the primaryvictim of the defendant’s negligence. However, seenfrom a different and more subtle angle, there wasalso an indirect or secondary ‘victim’, namely MM,the hapless and irresponsible parolee whose parolereadiness was never properly assessed before beingreleased on parole. MM was indeed convicted of thecrimes he had committed against the plaintiff whilehe was on parole but should have been in prison.This, in turn, resulted in a further but well-deservedhefty and effective sentence of 15 years’ directimprisonment. It would appear that in Naidu theparole system failed not only society and the plain-tiff, but also MM, who was negligently released onparole in circumstances where his parole was, in theabsence of rehabilitation, simply an opportunity tocommit further crimes. To put the matter in properperspective: it is also in the best interests of aconvicted offender that parole readiness be assessedproperly and, where required, parole be refused.

Kelly v Minister of Correctional Services & others2016 (2) SACR 351 (GJ)

In the above matter Satchwell J set aside the Minis-ter’s refusal of parole and the recommendationsmade by the National Council for CorrectionalServices (the ‘NCCS’). The NCCS is a councilestablished in terms of s 83 of the CS Act. Itsfunctions and duties are set out in s 84. One of itsfunctions is to advise the Minister.

The applicant in Kelly was serving a sentence of lifeimprisonment. He sought a review of the Minister’srefusal of his parole, which was based on recommen-dations of the NCCS and required him as a sentencedoffender to comply with certain conditions beforebeing released.

One condition of the NCCS was that the applicantwas ‘to participate further in the Gang ManagementStrategy’ programme (the ‘GMS programme’).However, it turned out that no GMS programme‘was ever availed to the applicant prior to the NCCSrecommendation’ and that there ‘had been no initialparticipation to be taken further’ (at [21]). Theapplicant was also not offered the opportunity toattend such a GMS programme at some other prisonat any time before and after the recommendation anddecision. There was no fair administrative action(at [25]). Satchwell J explained as follows (at[27]–[28]):

‘If the prisoner had been told that his paroledepended on the condition that he fly to themoon by waving his arms, there would be nodoubt in finding that such condition is animpossibility and therefore the recommendationand decision predicated upon such flight to themoon were unreasonable and must be setaside[.]

In the present circumstances, attendance of acourse on ‘‘Gang Management Strategy’’ hasproven equally impossible for respondents toprocure and applicant to attend prior to makingthe recommendation and decision (as well asthereafter). It cannot be considered to have beenanything other than an inadmissible or incom-petent consideration.’

The second condition set out in the NCCS recom-mendation, and adopted by the Minister in refusingparole, was that the applicant be encouraged andassisted to develop academic and/or practical skills‘to compete in the labour market once released onparole’ (at [7]). This condition was, according toSatchwell J, too vague and lacked details such ashow or when or where, or in what manner, theapplicant should develop skills which will assist himin the ‘labour market’ (at [37]). It appeared, further-more, that the NCCS or the Minister had paidinsufficient attention to the many courses success-fully completed by the applicant in prison, such as‘basic computer skills’ and ‘information technology’(at [30], [33] and [41]). Satchwell J also referred to amatter which she preferred to call a ‘general con-cern’, namely that sentenced offenders with anadvantaged background and academic qualificationsmight somehow be released on parole earlier thanthose of disadvantaged background and less educa-tion.

Satchwell J also found it unacceptable that theMinister should have set a condition that the offender‘should participate in Restorative Justice Processesinvolving the family of the victims as well as thecommunity’ (at [42]). She observed as follows (at[46]–[47]):

‘If any restorative justice is desired by victimsor their families, or the community, it is hardlypossible for the applicant to go and visit thevictims or hold a town-hall meeting, or devise aprogramme. Respondents are absolutely silenton what could or should be done, and whoshould do it, and when or how. . .. It may be, ofcourse, that none of the victims or their families

11Issue 1, 2017

wish to participate in any restorative-justiceprogramme. That is their right. They are notobliged to interact with, meet with, communi-cate or engage with—and certainly not for-give—the offender. And no burden should everbe placed on such a victim, that he or she isresponsible for the continued incarceration ofthe offender.’

In setting aside the Minister’s refusal of parole aswell as the NCCS recommendations, Satchwell Jfound it necessary to emphasise that whilst theapplicant was not as of right entitled to parole, hehad the right to a fair hearing which involves a fairprocess (at [56]). The process is unfair where theauthorities set up ‘hurdles which he cannot over-come’ (at [57]).

Having set aside the refusal of parole and therecommendations, the court ordered that the pre-scribed processes be carried out ‘in a manner whichrequires expedition’ (at [59]). For this purpose, thecourt ordered the respondents to reconsider theapplicant’s parole in terms of a strict and tight timeframe which was duly set out by the court in [60](b)to [60](g) of its judgment.

Remarks in conclusionIn Democratic Alliance v President of the Republicof South Africa & others 2013 (1) SA 248 (CC) at[13] the Constitutional Court observed that ‘aneffective criminal justice system. . .is vital to ourdemocracy’. It is also true that parole and the systemof parole form an important component of thecriminal justice system (Joubert (ed) Criminal Pro-cedure Handbook 12 ed (2016) 8).

In Naidu and Kelly—the two cases discussedabove—the High Court was required to deal withsituations where it was the poor and imperfect

implementation of the rules governing parole—andnot parole itself—that yielded results like delictualliability (in Naidu) and judicial intervention in theparole process (in Kelly).

The risk of a parolee violating his parole conditionsby reoffending is a risk inherently present in thesystem of parole. What is required, though, is thatparole readiness be properly assessed on the basis ofreliable information concerning rehabilitation.Clearly, in Naidu the Board’s decision to release MMon parole was unacceptable—not because it turnedout that MM had committed crimes whilst on parole,but because MM’s release on parole was approved inthe absence of evidence of rehabilitation and withoutproper consideration of his criminal profile andhistory of violent crime. Indeed, in Naidu theBoard’s negligence defeated one of the main objec-tives of parole, namely the successful reintegrationof the sentenced offender into society.

Moses Parole in South Africa (2012) 14 states: ‘Thetension in the relationship between the State and itsimprisoned citizens is acutely manifested in theassessment and consideration of prisoners for paroleand the ultimate decision whether or not to releasesuch prisoners on parole.’ It is against this particularbackground that Kelly (supra) must be understood.The ‘tension’ that exists is a healthy and necessaryone. However, it gets disturbed when parole pro-cesses and decisions lack reasonableness and fair-ness. This much happened in Kelly where a robustjudgment by Satchwell J was necessary to enforcefair administrative action (at [56]).

To be sure, the question whether parole should begranted or not is at the best of times not an easy one.However, those who have the responsibility to assessparole readiness and decide on parole would benefitgreatly from the principles and guidelines identifiedand applied in Naidu and Kelly.

Steph van der Merwe

What is the meaning of ‘hearsayevidence’ in s 3(4) of the Law ofEvidence Amendment Act 45 of 1988?This was one of the questions that Eksteen J had toconsider in Pentree Limited v Nelson Mandela BayMunicipality [2017] 2 All SA 260 (ECP). He wasrequested, in that case, to ‘deal generally with theissue of whether a party in expropriation proceedingsmay, through a valuer called as an expert witness,adduce evidence of statements made to the valuer by

other persons in respect of matters which haveinfluenced her valuation of the land, in circum-stances where such other persons are not called aswitnesses’ (at [9]).

Eksteen J looked at the position before the enactmentof the Law of Evidence Amendment Act 45 of 1988and considered what effect the Act had to change thatposition. In Estate De Wet v De Wet 1924 CPD 341Watermeyer J described hearsay evidence as ‘evi-dence of statements made by persons not called as


witnesses which are tendered for the purpose ofproving the truth of what is contained in the state-ment’. In R v Miller 1939 AD 106, the same judge,now elevated to the Appellate Division, went further(at 119):

‘Statements made by non-witnesses are notalways hearsay. Whether or not they are hearsaydepends upon the purpose for which they aretendered as evidence. If they are tendered fortheir testimonial value (ie as evidence of thetruth of what they assert), they are hearsay andare excluded because their truth depends uponthe credit of the asserter which can be testedonly by his appearance in the witness box. If onthe other hand they are tendered for theircircumstantial value, to prove something otherthan the truth of what is asserted, then they areadmissible if what they are tendered to prove isrelevant to the enquiry.’

These pronouncements, said Eksteen J, led Botha Jin Lornadawn Investments (Pty) Ltd v Minister vanLandbou 1977 (3) SA 618 (T), where the court ‘gaveconsideration to a similar issue’ to that beforeEksteen J, to ‘conclude that evidence which isapparently of a hearsay nature tendered for a purposeother than establishing the truth of the contentthereof is not struck by the hearsay rule and that suchevidence is admissible provided that such otherpurpose for which it is tendered is relevant to theissues in dispute’ (at [12]).

It was argued in Pentree that the validity of thesepronouncements ‘had been impugned by the promul-gation of the Evidence Act and that the common-lawrules relating to hearsay [found] no application after1988’ (at [12]). Counsel relied, for this submission,on this passage in The South African Law of Evi-dence 2 ed (2009) by Zeffertt and Paizes at 389:

‘No longer is hearsay defined along assertion-oriented lines, with the result that the hearsaystatus of the evidence no longer depends uponwhether or not it is tendered to prove the truthof what it (expressly or impliedly) asserts. In itsplace is a declarant-oriented definition that, ineffect, identifies hearsay according to whetheror not the traditional ‘‘hearsay dangers’’ arepresent.’

The definition in s 3(4) of the 1988 Act provides that‘‘‘hearsay evidence’’ means evidence, whether oralor in writing, the probative value of which dependsupon the credibility of any person other than theperson giving such evidence’. Eksteen J, while he

conceded that ‘the common law rules in respect ofhearsay no longer find application’, was, however,not persuaded that the statement of Zeffertt andPaizes could be accepted ‘without qualification’ (at[13]). His reasons were these:

‘At common law, if the evidence of a statementby a non-witness is tendered for a purpose otherthan its testimonial value then the truth orotherwise is irrelevant and it is not dependentfor its reception on the credibility of theasserter. It is not hearsay. . ..The position underthe Evidence Act appears to me to be much thesame. If it is not tendered for its testimonialvalue its probative value is not dependent on thecredibility of any person other than the persongiving such evidence. It is therefore not hearsayas defined in the Evidence Act.’

He drew further on what was said in Mdani v AllianzInsurance Ltd 1991 (1) SA 184 (A), where VanHeerden JA (at 189–90) made these remarks inconsidering the effect of s 3:

‘If A testified that B made such an admission,A’s evidence in itself is clearly not hearsay.Whether B in fact made the admission dependsupon A’s credibility and can be tested by cross-examination. What is hearsay is the content ofthe admission if it is to be used to establish thetruth of what was said. Whether the content istrue or not, depends entirely on B’s credibility.’(Emphasis added by Eksteen J).

In the light of these observations and pronounce-ments, reinforced by the approach of Botha J inLornadawn (at 622F–H), Eksteen J was convinced‘that the purpose for which the evidence is tenderedmay still determine whether it is hearsay or notwithin the definition in the Evidence Act’ (at [15]).

So was the evidence in question hearsay? In Lor-nadawn Botha J described the function of the courtin expropriation matters, since there was no lisbetween the parties and no onus on the plaintiff, asbeing that of a ‘super valuer’: it had to place itself inthe shoes of the notional informed seller and buyerand have regard to everything which a seller andbuyer would have experienced in the open market aswell as all the information that would have been attheir disposal in order to determine market value.The court, of course, could not go into the market togather information, but it was part of the function ofan expert valuer to do so and to found his opinion onthat information. In determining compensation, acourt was, according to Botha J, involved in ‘an

13Issue 1, 2017

enquiry into the probable conduct of two fictitiouspersons to whom we should attribute full knowledgeof all circumstances which would play a role in themanner in which they would have acted in theirnegotiations with one another in determining apurchase price for the subject property’ (per EksteenJ at [21]). And those fictitious persons would beinfluenced in their actions and conduct by relevantcommunications received by other people. Thesecommunications are, said Botha J (at 627A–D),‘direk ter sake in die ondersoek warmee die Hofbesig is, en dit is so afgesien van die vraag of dieinhoud van sulke medelings objektief gesproke diewaarheid is of nie’ (cited by Eksteen J at [21]).

In essence, Eksteen J concluded (at [22]), ‘evidenceof such communications are not dependent for theirreception on the credibility of the asserter and theprobative value thereof in expropriation proceedingsdepends on the weight which the notional willingbuyer and willing seller would have attached theretoin the light of all the known facts’. It was therefore‘not hearsay at all in the context of the expropriationproceedings for the probative value thereof is notdependent on the credibility of any other person, butrather on the weight which a notional willing buyeror seller would have attached to the statement in hisnegotiation’ (at [27]).

It is submitted that Eksteen J was correct in labellingthe evidence non-hearsay. It was similar, in nature, tothe situation that arose in International Tobacco Co(SA) Ltd v United Tobacco Cos (South) Ltd 1953 (3)SA 343 (W): if W were to testify that he is acommercial traveller and that other people had toldhim that rumours were circulating that a particularbrand of cigarette caused a particular disease, andthis evidence were tendered to prove that suchrumours were, in fact, circulating, would it behearsay? It would, in my submission, not be hearsaybecause what the other people told W would be noless proof of the circulation of those rumours thanwhat others told them, so that the probative value ofW’s evidence that rumours were circulating woulddepend entirely on his own credibility.

I have, therefore, no quarrel with Eksteen J’s conclu-sion that the evidence was not hearsay. What I do,with respect, take issue with is his reasoning con-cerning the impact of s 3(4) on the meaning anddefinition of hearsay, and, in particular, his assertion(at [15]) that ‘the purpose for which the evidence istendered may still determine whether it is hearsay or

not within the definition in the Evidence Act’(emphasis added).

It is important, in my view, to understand why theold approach to hearsay, which hinged aroundwhether the evidence was tendered to prove the truthof its contents, was dysfunctional and unsatisfactory.Consider these situations:

Situation 1: The issue in a particular matter iswhether X, following a traumatic event, was able tospeak. W, who testifies on the issue, tells the courtthat he came upon the traumatised X and asked himif he was too shocked to speak, whereupon X toldhim: ‘I can speak’. This is clearly adduced to provethe truth of what X said to W, that he could, indeed,speak. But it is not hearsay; nor should it be solabelled. The probative value of what was said by Xto W clearly depended entirely on the credibility ofW, who heard the words used by X, irrespective ofwhat those words actually were. The same wouldapply to similar expressions, such as ‘I am alive’ or‘I am here’, if used to prove those very facts. The‘assertion-oriented definition’, then, fails and the‘declarant-oriented’ definition (as embodied ins 3(4)) provides the correct solution. It can, ofcourse, still be said that the evidence is not beingtendered to prove the truth of what is asserted, andthat we are concerned here only with the circumstan-tial use of language by X and not with the content ofthe words uttered by him. But my point is that weshould not even be engaging in such subtle distinc-tions since the plain wording of s 3(4) simply ren-ders this unnecessary.

Situation 2: The issue is the sanity of the testator, T,whose will is challenged by those who would benefitfrom its invalidity. In order to show that T was saneat the time of making the will, evidence is adducedof letters addressed to and received by him. They aretendered to prove, not anything expressly said by thewriters of those letters, but that those writers, giventhat the contents were on matters and in languageappropriate to communications with a person ofreasonable understanding and intelligence, believedT to have possessed an adequate comprehension ofthe affairs described. This type of situation—thefacts of which are those of an old English case,Wright v Doe d Tatham (1837) 7 Ad & El 313—hasgiven rise to the notorious problem of so-called‘implied hearsay assertions’: is evidence hearsaywhere it is tendered to prove, not what it expresslyasserts, but something that it asserts by implication?The assertion-oriented definition of old did notprovide an answer to this question, and it tended to


lead into a semantic swamp where fine and oftenunhelpful distinctions would be made between thecircumstantial use of the evidence and its tendencyto assert something indirectly by implication. Thedeclarant-oriented definition in s 3(4) avoids suchdilemmas by going straight to the heart of the matter:if the probative value of the evidence depends on thecredibility of the absent declarant, the evidence ishearsay. In the example taken from Wright’s case, theevidence would be hearsay, since the assessment ofT’s sanity by the writers of the letters dependsstrongly on their credibility: were they sincere inwriting these letters or were they part of an elaboratecharade; did they know T well enough to make anysuch assessment; how much time had elapsed sincelast they saw him? These are questions for cross-examination. The prejudice caused to the opposingparty by the denial of the opportunity to cross-examine the writers of the letters on these matters isthe strongest—but not the only—ground for classify-ing the evidence as hearsay.

It is significant that it took the Canadian courts aslong as it did to identify that no distinction should bedrawn between express and implied assertions forthis purpose. For, in the absence of a declarant-oriented definition of the kind employed in s 3(4),there was less to alert the courts in Canada to whatFish J, in the landmark decision of the SupremeCourt in R v Baldree 2013 SCC 35, [2013] 2 SCR520 (discussed in Commentary in s 216 sv Definitionof hearsay) described as the absurdity ‘that anythingshould turn on the grammatical form of thedeclarant’s assertions’ (at [42]). The key to encour-aging this breakthrough is that in Canada, as inSouth Africa since the enactment of s 3, there is aflexible and nuanced regime for dealing with theadmissibility of hearsay. Many implied assertionswill be highly reliable, especially since the danger ofinsincerity tends to be reduced given the indirect,implicit use of what was said. Where the hearsaydangers (discussed in Commentary in s 216) aresatisfactorily accounted for, these will be admitted inSouth Africa under s 3(1)(c). Without this liberalapproach to admissibility, courts would be much lessinclined to bring implied assertions within the hear-say fold and to risk, by doing so, that they might belost as evidence despite their reliability in a givencase.

Situation 3: A ship has sunk and the question raisedby the insurer is whether it was seaworthy when itsailed. Evidence is tendered that the captain of theship, after conducting an inspection, embarked on

the voyage, taking with him his entire family. Is thisevidence hearsay? The assertion-orientated defini-tion does not even get to grips with the problem, as itis meaningless to talk about the truth of the contentsof the communication—express or implied—wherethe communication is non-verbal and not evenintended to be communicative in the first place.

The declarant-oriented definition does. It asks if theevidence—tendered in oral form by a witness whosaw all these events (and so satisfying the require-ment in s 3(4) that the evidence be ‘oral or inwriting’)—depends for its probative value on thecaptain’s credibility. That it clearly does, is illumi-nated by the kinds of question the opposing party,denied the opportunity to cross-examine him, mighthave liked to ask: how rigorous was the examination;how experienced was the captain in assessing suchmatters; was he sober at the time; was he a risk-taker; how good was the lighting when he inspectedthe ship?

It would take some boldness for courts employingthe common-law definition to identify such evidenceas hearsay. Interestingly, even the Canadian SupremeCourt in Baldree (at [63]) thought the issue ‘best leftfor another day’, adding merely this comment by PRRice (1992) 65 Temple Law Review 529 at 536:‘[O]ne can engage in conduct without ever intendingto communicate anything to anyone [but] the same isnot true of speech or a combination of speech andconduct (for example, placing a bet) because the solepurpose of speech is communication.’ It is submittedthat, whereas the danger of insincerity will usuallybe significantly reduced in such cases, those oferroneous memory and perception will frequentlyremain. Whether they are strong enough to signifyshould be something a court will take into accountin deciding admissibility under s 3(1)(c), not cate-gorisation. The mere potential for such danger—andthe prejudice it could cause—is reason enough tolabel the evidence hearsay. Admissibility is anothermatter.

In short, the triumph of s 3 is that it liberates us froma number of straitjackets: the tyranny of rule andfixed exceptions; the forcing of evidence into ill-fitting and synthetic categories; the service of for-malism at the expense of substantive justice andeffective function; the unquestioning deference tostale precedent; the dead hand of linguistic andgrammatical artifice; and the muddle of hair-finedistinctions.

It is in this spirit that I made the comments in TheSouth African Law of Evidence that are referred to by

15Issue 1, 2017

Eksteen J. To say, as Eksteen J did in Pentree, thatthe position under the Evidence Act ‘appears to bemuch the same’ as that at common law, and that the‘purpose for which evidence is tendered may stilldetermine whether it is hearsay or not within thedefinition in the Evidence Act’ (at [14] and [15];emphasis added), is to deny or negate some of thesegains. It runs, moreover, counter to the very wordingof s 3(4), which makes no mention of whether or notthe evidence is tendered to prove the truth of itscontents.

It must be acknowledged that the old definition willproduce the correct results in the majority of cases.

But it is a flawed theory that fails entirely to accountfor some cases. To borrow an analogy from the fieldof theoretical physics: Newton’s laws served us wellfor centuries and were good enough to get us to themoon. But Einstein’s theory showed Newton’s phys-ics to be flawed and incomplete: it could not accountfor the motion of Mercury, the bending of light fromthe stars by the sun, and many other phenomena. Wewill probably never reach the stars. But Einstein’srevelations at least cleared away a fatal impediment.

Andrew Paizes


(B) LEGISLATION

During the period under review (1 February 2017 to12 June 2017) there were no amendments to theCriminal Procedure Act 51 of 1977. Bills which willaffect provisions in this Act are referred to below.Other legislative activity pertinent to our criminaljustice system is also identified below.

(i) The Justice Administered Fund Act2 of 2017

The above Act was signed by the President on 4April 2017 and published for general information on6 April 2017. See GN 344 in GG 40773 of 6 April2017. The Act will come into operation on a date tobe fixed by the President by proclamation in theGazette.

The Act establishes a Justice Administered Fund (the‘Fund’) and provides for the management, control,investment and utilisation of money in this Fund.

Section 3(b) of the Act provides that money receivedas bail, payable in terms of the Criminal ProcedureAct 51 of 1977 or any other Act, must be adminis-tered through the Fund. It should be noted that thissection regulates the administration of bail moneythrough the Fund and has no effect on payment andrepayment of bail money as provided for in Chapter9 ‘Bail’ (ss 58–71) of Act 51 of 1977.

(ii) Implementation of the Rome Statuteof the International Criminal CourtAct Repeal Bill [B23—2016]:Withdrawal by Minister

The Minister of Justice and Correctional Serviceshas withdrawn the above Bill in accordance withRule 334 of the Rules of the National Assembly. SeeAnnouncements, Tablings and Committee Reports ofthe Parliament of South Africa (13 March 2017) at 2.

(iii) Children’s Amendment Act 17 of2016

The above Act was signed by the President on 18January 2017 and was published for general infor-mation on 19 January 2017. See GN 42 in GG 40773of 19 January 2017. The Act (the ‘Amendment Act’)will come into operation on a date to be fixed by thePresident by proclamation in the Gazette.

In terms of s 120(1)(b) of the Children’s Act 38 of2005, a finding that a person is unsuitable to workwith children may be made by ‘any. . .court in any

criminal or civil proceedings in which that person isinvolved. . .’. Section 2(b) of the Amendment Actinserts a new s 120(4) which sets out the circum-stances and offences when a person, in criminalproceedings, ‘must be deemed unsuitable to workwith children. . .’. However, in terms ofs 120(4A)—as inserted by s 2(c) of the AmendmentAct—a court may not make an order of unsuitabilitywhere the person concerned was a child at the timeof the commission of the offence unless:

‘(a) the prosecutor has made an application tothe court for such an order;

(b) the court has considered a report by theprobation officer referred to in section 71for the Child Justice Act, 2008, whichdeals with the probability of committingan offence contemplated in subsection (4),against a child;

(c) the person concerned has been given theopportunity to address the court as to whyhis or her particulars should not beincluded in the Register; and

(d) the court is satisfied that substantial com-pelling circumstances exist based uponsuch report and any other evidence, whichjustify the making of such an order.’

The ‘Register’ referred to in s 120(4A)(c) above isthe National Child Protection Register (the ‘NCPR’)as created and governed by ss 111–128 of theChildren’s Act 38 of 2005.

The new s 120(4A) of Act 38 of 2005 is similar tos 50(2)(c) of the Criminal Law (Sexual Offences andRelated Matters) Amendment Act 32 of 2007. Sec-tion 50(2)(c) was inserted by s 7(b) of Act 5 of 2015which came into operation on 7 July 2015. See GG38977 of 7 July 2015. Section 50(2)(c) was intro-duced as a result of the Constitutional Court decisionin J v National Director of Public Prosecutions &others 2014 (2) SACR 1 (CC), a decision thatfollowed S v IJ 2013 (2) SACR 599 (WCC). Seefurther the notes on these two cases in the discussionof s 276 in Commentary, sv The National Registerfor Sex Offenders (‘NRSO’): The convicted sexoffender who was a child at the time of the commis-sion of the offence. The new s 120(4A) of Act 38 of2005 will ensure that placement of a child’s name onthe NCPR would be consistent with the decision ofthe Constitutional Court in J (supra).

17

(iv) Judicial Matters Amendment Bill[B14—2016]

In para 1 of the memorandum on the objects of theabove Bill published in GG 40274 of 14 September2016, it is stated that the Bill seeks to amendprovisions of several Acts in order to address practi-cal and technical issues of a non-contentious nature.The important clauses seeking to amend the Crimi-nal Procedure Act 51 of 1977 were discussed in CJR2016 (2) at 9–10.

The invitation by Parliament’s Portfolio Committeeon Justice and Correctional Services to make writtensubmissions on the Bill, closed on 15 March 2017. Itwould therefore appear that the Bill might becomelaw before the end of this year.

(v) Criminal Procedure AmendmentBill [B2—2017]

The above Bill was published in GG 40487 of 9December 2016. It seeks to make various amend-ments to Chapter 13 ‘Accused: Capacity to Under-stand Proceedings: Mental Illness and CriminalResponsibility’. This chapter consists of three sec-tions: s 77 (capacity of accused to understand pro-ceedings); s 78 (mental illness or mental defect andan accused’s criminal responsibility) and s 79 (panelfor purposes of an inquiry and report under ss 77 and78).

Proposed amendments to s 77(6)(a)(i)

The above section provides that where a court findsthat an accused is incapable of understanding theproceedings so as to make a proper defence, thecourt must—in the case of certain serious crimes andcircumstances as identified in s 77(6)(a)(i)—directthe detention of the accused in a psychiatric hospitalor a prison pending the decision of a judge inchambers in terms of s 47 of the Mental Health CareAct 17 of 2002. However, in a judgment delivered on26 June 2015 the Constitutional Court declareds 77(6)(a)(i) inconsistent with the Constitution to theextent that it provides for (a) the compulsory impris-onment of an adult and (b) compulsory hospitalisa-tion or imprisonment of children. The declaration ofinvalidity was suspended for two years from the dateof the judgment in order to allow Parliament toaddress the matter having regard to the defects raisedin the judgment. See De Vos NO & others v Ministerof Justice and Constitutional Development & others2015 (2) SACR 217 (CC) at [69]. This case isdiscussed in the notes on s 77 in Commentary, sv

Section 77(6): The unconstitutionality of subpara-graphs (i) and (ii).

Clause 1(b) of the Bill aims to amend s 77(6)(a)(i) inaccordance with constitutional requirements identi-fied by the Constitutional Court in De Vos (supra). Interms of this clause a court will have a discretion—indealing with a s 77(6)(a)(i) situation—to order thatthe accused be detained in a psychiatric hospital (ortemporarily in a correctional health facility of aprison where no bed is available in a psychiatrichospital) ‘if the court is of the opinion that it isnecessary to do so on the grounds that the accusedposes a serious danger or threat to. . .[himself]. . .orto members of the public or to any property belong-ing to him. . .or any other person’. In terms of clause1(b) the court will also, depending on all the circum-stances, be able to order the detention of the accusedin a designated health establishment as if he were aninvoluntary mental health care user contemplated ins 37 of the Mental Health Care Act 17 of 2002. Afurther alternative for the court would be to order therelease of the accused subject to such conditions asthe court considers appropriate.

It is submitted that clause 1(b) is consistent withconstitutional requirements and meets all the prob-lems identified in De Vos (supra) at [46], [48], [51]and [52]. Given the time limit set by the court, it is tobe expected that clause 1(b) will become law on orbefore 25 June 2017.

Proposed amendments to s 77(6)(a)(ii)

Section 77(6)(a)(ii) deals with the situation where acourt finds that the accused has committed anoffence other than the serious offences stipulated ins 77(6)(a)(i) or that the accused has not committedany offence. Before the decision of De Vos (supra) on26 June 2015, the court was simply required to makean order that an accused in the s 77(6)(a)(ii) situationbe admitted to and detained in an institution as if heor she were an involuntary mental health care usercontemplated in s 37 of the Mental Health Care Act17 of 2002. In De Vos it was found that s 77(6)(a)(ii)breached the right to freedom and security of theperson as guaranteed in s 12 of the Constitution (at[57]) and amounted ‘to an arbitrary deprivation offreedom under s 12’ (at [58]). ‘The mere fact’, it wassaid at [66], ‘that an accused person brushes shoul-ders with the criminal justice system is not a justcause for institutionalisation and renders the provi-sion constitutionally invalid in respect of such aperson.’ However, at [67] the court concluded that a‘reading-in’ was appropriate and that s 77(6)(a)(ii)could be remedied by extending the options avail-


able to the court. The Constitutional Court accord-ingly ordered that as from 26 June 2015 two furtheroptions had to be read into s 77(6)(a)(ii), namely anorder by the court that the accused be ‘releasedsubject to such conditions as the court considersappropriate’ or an order by the court that the accused‘be released unconditionally’ (at [69]).

In proposing a new s 77(6)(a)(ii), clause 1 of the Billincorporates the Constitutional Court’s ‘reading-in’as set out above. Indeed, in De Vos (supra) at [67] theConstitutional Court noted that the legislature couldamend s 77(6)(a)(ii) ‘as it deems fit. . .provided suchamendment complies with this judgment and theConstitution’.

Proposed amendment to s 78(6)

An accused must be found not guilty if the courtfinds that the accused was at the time of thecommission of the offence ‘by reason of mentalillness or intellectual disability not criminallyresponsible’ for his act. See s 78(6)(a) as read withs 78(1). Once an accused has been acquitted on thisbasis, the court is required to make one of severalorders as set out in s 78(6). Clause 2 inserts a furtherpossible order that the court can make, namely thatthe acquitted person be

‘temporarily detained in a correctional healthfacility of a prison where a bed is not immedi-ately available in a psychiatric hospital and betransferred where a bed becomes available, ifthe court is of the opinion that it is necessary todo so on the grounds that the accused poses aserious danger or threat to himself or herself orto members of the public or to any propertybelonging to him or her or any other person. . .’.

The above proposed amendment is an eminentlysensible one. Clause 2 also seeks to amend s 78(6) tothe extent that detention in a prison could no longerbe ordered unless it is in the circumstances asdescribed in the proposed amendment quoted above.

Proposed amendments to s 79

Section 79(1) governs the constitution of the panel ofexperts who should report to the court for purposesof ss 77 and 78. The size of the panel can varyaccording to the nature or classification of theoffence(s) involved. In S v Pedro 2015 (1) SACR 42(WCC) attention was drawn to inconsistencies in theinterpretation and application of s 79(1)(b) as far asthe composition of the panel was concerned. See inthis regard the discussion of Pedro in the notes on

s 79 in Commentary, sv Section 79(13): Directivesissued by the NDPP.

Clause 3 of the Bill aims—in line with the findingsand views in Pedro (supra)—to clarify the composi-tion of the panels and secure a consistent applicationof s 79(1) in all courts across the country. Theproposed new s 79(1) will read as follows:

‘(1) Where a court issues a direction undersection 77(1) or 78(2), the relevant enquiryshall be conducted and be reported on—

(a) where the accused is charged with anoffence other than one referred to inparagraph (b), by the head of thedesignated health establishment des-ignated by the court, or by anotherpsychiatrist delegated by the headconcerned; or

(b) where the accused is charged withmurder or culpable homicide or rapeor compelled rape as provided for insection 3 or 4 of the Criminal Law(Sexual Offences and Related Mat-ters) Amendment Act, 2007, respec-tively, or another charge involvingserious violence, or if the court con-siders it to be necessary in the publicinterest, or where the court in anyparticular case so directs—

(i) by the head of the designatedhealth establishment, or byanother psychiatrist delegated bythe head concerned;

(ii) by a psychiatrist appointed bythe court;

(iii) by a psychiatrist appointed bythe court, upon application andon good cause shown by theaccused for such appointment;and

(iv) by a clinical psychologist wherethe court so directs.’

Clause 3 of the Bill also proposes the repeal ofs 79(13). This section empowers the National Direc-tor of Public Prosecutions to issue directives con-cerning the prosecutor’s application for the appoint-ment of a psychiatrist in certain specifiedcircumstances. See also Pedro (supra) at [25] and[48]–[69]. Repeal of s 79(13) is a consequence of thefact that the proposed new s 79(1)(b) does not

19Issue 1, 2017

include a prosecutorial right to apply for an extrapsychiatrist to be appointed in certain circumstances.

Sections 77, 78 and 79: ‘mental defect’ to bereplaced by ‘intellectual disability’

The term ‘mental defect’ appears in ss 77, 78 and 79.The Bill proposes the replacement of ‘mental defect’with ‘intellectual disability’. In para 2.1.1 of theMemorandum which accompanied the Bill, it isexplained that ‘intellectual disability’ is the ‘moreacceptable term’ and that ‘mental defect’ is an

‘outdated term’. On the use of the term ‘intellectualdisability’, see the discussion of s 78 in Commen-tary, sv Section 78(6): ‘mental illness or mentaldisability’. The term ‘intellectual disability’ isalready used in s 78(6). It is submitted that theintroduction of this term in all other relevant provi-sions of ss 77, 78 and 79 would also promoteconsistency without disturbing any substantivecriminal law principles governing capacity to under-stand proceedings and criminal responsibility.


(C) CASE LAW

(a) Criminal Law

Is a person who assists another tocommit suicide guilty of murder?Physician administered euthanasia andphysician assisted suicide consideredMinister of Justice & others v Estate Stransham-Ford 2017 (3) SA 152 (SCA)

This case arose out of the wishes of a terminally illperson, SF, who had approached the High Courtclaiming an order that a medical practitioner couldeither end his life by administering a lethal sub-stance, or provide him with the lethal substance toenable him to administer it himself. He sought anorder that the common law in relation to the crimesof murder and culpable homicide should, in terms ofs 39(2) of the Constitution, be developed to allow forsuch an order to be made. Two hours after SF died,the High Court, unaware of that fact, granted anorder that the medical doctor who acceded to SF’srequest would not be acting unlawfully, and hencewould not be subject to prosecution or to disciplinaryproceedings for such conduct.

The appeal to the Supreme Court of Appeal wasagainst this order. The appeal was successful, and theorder was set aside for ‘three inter-related reasons’(at [101]), any one of which would have beendecisive. First, the death of SF two hours before theorder was made meant that the cause of action hadceased to exist. Once the judge who had made it wastold of SF’s death the following week, before deliv-ering his reasons, he refused to recall his order on theground that his judgment had broader societal impli-cations. It was held that he was wrong not to rescindhis order since it had been granted on the erroneousbasis that SF was still alive.

Secondly, there was no full and proper examinationof the present state of our law on the subject. And,thirdly, the order had been made on an incorrect andrestricted factual basis, without complying with theUniform Rules of Court and without affording allinterested parties a proper opportunity to be heard.

Since, however, there was the ‘dilemma’ that thedecision of Fabricius J in the High Court (see 2015(4) SA 50 (GP)) stood as a ‘reasoned and reportedjudgment by the High Court’which might, unless theSupreme Court of Appeal, at least to some extent,addressed the merits, ‘be taken as having someprecedential effect’ (at [27]), the court felt compelledto deal with the merits.

Wallis JA, who gave the judgment of the court,considered it necessary to undertake a ‘brief exposi-tion of the current state of our law in this area’ (at[29]). He found the succinct assertion of the HighCourt that ‘[t]he current legal position is that assistedsuicide or active voluntary euthanasia is unlawful’(at [10]), to be, ‘not only not supported by theauthorities relied upon’, but also ‘a wholly inad-equate analysis of the relevant law in this area’ (at[29]).

The judgment of Wallis JA covered a wide expanseand included a close examination of the decidedSouth African cases, the provisions of the Constitu-tion and comparative foreign law. The following aresome of the propositions expounded and issuesconsidered:

(1) Neither suicide nor attempted suicide is acrime in South Africa: see R v Peverett1940 AD 213.

(2) A person may refuse medical treatment thatwould otherwise prolong life. This is anaspect of personal autonomy that is consti-tutionally protected and it would not ordi-narily be regarded as suicide: see Re Conroy486A 2d 1209 (NJSC 1985) at 1224. Theconstitutional provisions involved are ss 10(right to dignity) and 12(2)(b) (right tobodily integrity).

(3) The refusal of treatment can be made by apatient only if he or she has the mental andlegal capacity to make such a decision. Insuch cases doctors do not commit a crimi-nal offence ‘by ceasing treatment or otherforms of medical intervention that serveneither a therapeutic nor a palliative pur-pose’ (at [33]). Where patients lack capac-ity, decisions are usually made by thedoctors in conjunction with family mem-bers and others having responsibility forthe patient. In cases of uncertainty, it mightbe desirable to seek a declaratory courtorder (see Clarke v Hurst NO & others1992 (4) SA 630 (D)).

(4) A medical practitioner ‘commits nooffence by prescribing drugs by way ofpalliative treatment for pain that the doctorknows will have the effect of hastening thepatient’s death’. This is referred to as ‘the‘‘double effect’’, where the drugs serve thepurpose for which they were prescribed,but have potentially detrimental side

21

effects’ (see Clarke v Hurst supra). Therewere, said Wallis JA (at [35]), many stepsavailable to avoid interminable and pur-poseless treatment or the preservation oflife ‘as a purely mechanical process arti-ficially maintained’, and considerableadvances had been made in palliative careto alleviate suffering.

(5) On the other hand, a ‘mercy killing’‘undoubtedly constitutes the crime of mur-der’ at [36]): see S v Hartmann 1975 (3)SA 532 (C); S v De Bellocq 1975 (3) SA538 (T) and S v Marengo 1991 (2) SACR43 (W). These cases, however, had nothingto do with either physician assisted suicide(PAS) or active physician administeredeuthanasia (PAE), since in none of themdid the deceased ask for his life to beended. They were relevant in confirming,however, that consent is not a defence tomurder.

(6) In S v Robinson & others 1968 (1) SA 666(A), the deceased did arrange for others,including family members, to kill him forfinancial (insurance-related) reasons. Thecase confirmed that PAE did constitutemurder, so that a ‘medical practitioner whoadministers a lethal agent to a patient at thelatter’s request commits the crime of mur-der’ (at [40]). Wallis JA could see noreason for distinguishing between theactions of a medical practitioner fromthose of a family member or friend in suchcircumstances.

(7) The question in this case was, therefore,whether the law should be changed. Thecase before the court was, in effect, chal-lenging both Peverett and Robinson. How-ever, neither the principles nor the cases inquestion were addressed by the HighCourt. If the law were to be developed, this‘needed to be confronted squarely and thescope and ambit of the requisite exceptionto, or departure from, existing principlehad to be defined’. An ‘order making sucha profound change to our law of murder,without any consideration of applicableprinciples, should not have been made’ (at[41]). Even if the deceased had not diedwhen he did, since no doctors had comeforward to say they were willing to do asrequested by SF, the matter was purely

‘academic’ and the court should not haveentered into such an academic inquiry.

(8) Wallis JA considered next the issue ofassisted suicide (PAS), which was soughtas an alternative order. In this context thedecision in Ex parte Die Minister vanJustisie: In re S v Grotjohn 1970 (2) SA355 (A) was relevant. In that case thedeceased was a bipolar, partially paralysedwoman, whose husband was having anaffair with another woman. In the course ofa heated argument, she said that she wouldshoot herself. The accused (her husband)obtained a bullet from elsewhere in theirapartment, loaded the gun in her presence,and said: ‘Skiet jouself dan as jy wil wantjy is a las.’ She did so. The AppellateDivision was asked to consider whetherencouraging, providing the means for orhelping another to commit suicide was acrime, and, if so, what crime. There was,said the court, no simple ‘yes’ or ‘no’answer to this question. The ordinary prin-ciples of the criminal law—includingthose relating to causation—had to beinvoked in the context of the particularfacts of each case. The court, however,recognised the possibility that murder maybe committed in such circumstances if therequisite intent were present and there wasno break in the chain of causation betweenthe accused’s actions and the death of thedeceased (as to which, compare the ‘sui-cide pact’ cases of Peverett (supra) and S vGordon 1962 (4) SA 727 (N) discussed byWallis JA). If the accused lacked intent buthad mens rea in the form of negligence,the crime of culpable homicide would becommitted if all the requirements (includ-ing causation) were present. But it couldnot be said ‘that in the current state of ourlaw PAS is in all circumstances unlawful’(at [54]); emphasis added); the HighCourt’s statement to that effect ‘went toofar’.

(9) A court confronted with a case of PASwould thus have to consider how theprinciples articulated in Grotjohn ‘shouldbe applied and adapted to the present day’(at [55]), with particular attention to (i) thefacts of the particular case; (ii) the changesin medical science in the 50 years since


that decision; and (iii) the requirement ins 39(2) of the Constitution that, in thedevelopment of the common law, thecourts must strive to give effect to thenature, purport and objects of the Bill ofRights, in respect of which assistancecould profitably be sought by looking atother jurisdictions. In particular, a court, ifit was of the view that the common lawshould be developed, would have to con-sider the area of the criminal law thatshould be so developed: whether causa-tion, intention or unlawfulness. Such con-siderations were absent in the High Courtdecision, and this rendered it impossible toassess whether any limitation of a constitu-tional right was reasonable or justifiable interms of s 36 of the Constitution.

(10) Wallis JA turned next to the foreign law,which was examined and compared. Oneaspect of this treatment stands out: the factthat the High Court’s ‘too ready adoption’of the reasoning of the Supreme Court ofCanada in Carter v Canada (AttorneyGeneral) 2015 SCC 5; [2015] 1 SCR 331‘ignored the very different context inwhich that case was decided’ (at [72]). Itwas, said Wallis JA, important to note thatit was ‘only on the question of ‘over-breadth’ that the Supreme Court of Canadaheld in Carter that the criminalisation ofaiding and abetting suicide unjustifiablyinfringed a protected right’. A South Afri-can court faced with the same issue wouldneed to examine the ‘very different cir-cumstances in this country; the availabilityof medical care and especially palliativecare; the wide diversity of our society in itscultures and belief systems; our sense ofthe need to protect the poor, the weak andthe vulnerable and the value attached toproviding such protection’. ‘The notion ofa dignified death’, said Wallis JA (at[100]), had to be ‘informed by a roundedview of society, not confined to a restrictedsection of it’. This was not done in thiscase and could not have been done‘because of the inadequacies of the evi-dence and the haste with which it wasdecided’. When an appropriate case comesbefore our courts, the common law will, heconcluded, ‘no doubt evolve in the light of

the considerations outlined. . . and thedevelopments in other countries’.

Constitutional validity of statutoryoffences prohibiting use of cannabis forpersonal consumption and relatedmattersPrince v Minister of Justice and ConstitutionalDevelopment & others; Acton & others v NationalDirector of Public Prosecutions & others (unre-ported, WCC case no 8760/2013, 31 March 2017)

How, in a constitutional democracy such as SouthAfrica, do the courts ‘demarcate the border betweenthe terrain in which disputes lie in the province of thecourts and those controversies which are betterlocated on a terrain belonging to the legislatureand/or the executive’? This broad question faced thecourt in Prince, in which Davis J (in whose judg-ment Saldanha and Boqwana JJ concurred) had toconsider this particular question: to what extent arethe statutory laws that prohibit the use of cannabisand the possession, purchase or cultivation thereofexclusively for personal consumption, valid?Another aspect of the inquiry was ‘whether theinvalidity of the relevant legislature framework interms of which these prohibitions [were] couchedshould be determined by courts or left for furtherlegislative and/or executive consideration’ (at [2]).

Among the legislative provisions that wereimpugned were: ss 4(b) and 5(b) of the Drugs andDrug Trafficking Act 140 of 1992 (the ‘Drugs Act’),read with part 3 of Schedule 2 to the Drugs Act in sofar as it related to ‘simple possession, cultivation,transportation and distribution of cannabis for per-sonal communal consumption’; s 22A(10) of theMedicines and Related Substances Control Act 101of 1965 (the ‘Medicines Act’), read with Schedule 7of that Act; and s 40(1)(h) of the Criminal ProcedureAct, which empowers a peace officer without awarrant to arrest any person who is reasonablysuspected of having committed an offence under anylaw governing the making, supply, possession orconveyance of, inter alia, dependence-producingdrugs.

Section 4(b) of the Drugs Act prohibits, except inspecific cases, the use or possession of cannabis(among other substances), and s 5(b) prohibits, againexcept in specific cases, any dealing in such sub-stances. Section 22A(9)(a)(i) of the Medicines Actprohibits the use, possession, manufacture or supplyof cannabis. The applicants invoked what Davis J (at[11]) called a ‘veritable constitutional laundry list’ to

23Issue 1, 2017

argue that the criminal prohibition of the use andpossession of cannabis in their own homes and in‘properly designated places’ was unconstitutional.The list included the rights to equality, dignity andfreedom of religion, but the thrust of the argumenthinged around the right to privacy.

The court considered that the doctrine of res judicatawas no impediment in this case since the decision inPrince v The President, Cape Law Society & others2002 (2) SA 794 (CC), relied on by the respondents,turned on the narrow question involving limitedexemptions, for religious reasons, from the criminalprovisions. The majority in that case found thelimitation of the right to religious freedom to bejustifiable on the ground that it would be impracticalto administer a religious exemption without funda-mentally undermining the general prohibitionagainst possession of cannabis.

The core of the matter, said Davis J (at [20]), waswhether the infringement of the right to privacycaused by the impugned legislation could be justifiedin terms of s 36 of the Constitution.

The court examined the nature and extent of the rightto privacy, and made the following remarks: First,the right extends only to those aspects in regard towhich a legitimate expectation of privacy can beharboured (see Bernstein & others v Bester & othersNNO 1996 (2) SA 751 (CC) at [75]). Secondly, thereis a narrowly construed, invisible core in respect ofwhich a very high level of protection is given to anindividual’s intimate personal sphere of life, which is‘beyond interference from any public authority’, andwhere no justifiable limitation can take place (Bern-stein at [77]). This core is left behind once anindividual enters into relationships with personsoutside this sphere, so that the right then becomessubject to limitation as the activities acquire a socialdimension. Thirdly, it is established law that the rightto privacy ‘becomes more powerful and deserving ofgreater protection the more intimate the personalsphere of the life of a human being which comes intolegal play’ (at [22] of Prince). Fourthly, there is aconnection between the right of privacy and that todignity (see s 10 of the Constitution): privacy fostershuman dignity and a right to make intimate deci-sions, and to have one’s personal autonomy pro-tected is central to individual identity. Privacy islinked, too, to the right to freedom; it ‘fosters andencourages the moral autonomy of citizens which isa central requirement of governance in a democracy’(at [24]).

On the general question of the limitation of rightsunder s 36(c), Davis J made these observations:First, the burden of justification rested upon the State(see Moise v Greater Germiston Transitional LocalCouncil: Minister of Justice and ConstitutionalDevelopment Intervening 2001 (4) SA 491 (CC) at[19]). Secondly, the provisions of Chapter 2 of theConstitution (including s 36) had to be interpreted inthe light of values underlying an open and demo-cratic society based on human dignity, equality andfreedom. Privacy, in this context, was clearly deserv-ing of constitutional protection, absent a clear justi-fication to the contrary.

Was there, then, a substantial State interest to war-rant such a limitation? Davis J traced the history ofthe prohibition against cannabis (which began in1928) and found it to be ‘replete with the racism thathas saturated South African society for more thanthree centuries’ (at [33]). He insisted that it wasnecessary for the State to provide ‘clear and plau-sible evidence tailored to the contemporary contextto justify an infringement of an important right . . .viewed in terms of autonomous acts in one’s home’(at [34]). The ‘private moral views of a section of thecommunity’ would not suffice.

The State’s evidence pivoted around an affidavit by apharmacist and registrar of medicines who attestedto the acute effects of cannabis, including neurologi-cal impairment of driving skills and its harmfuleffects on the brain and the body, evident in bothlearning and pregnancy. There was also an affidavitby a general medical practitioner, whose qualifica-tions as an expert in this field were ‘hardly beyondcontest’.

Against these was a detailed report by an expert incriminology, who noted that countries with morepunitive policies did not tend to have lower drug useprevalence levels than those with more liberal poli-cies. He contested claims regarding aspects of theeffects of cannabis on cognitive function, andasserted that many of those who wished to endprohibition did so ‘out of a keen appreciation’ of theharms at play—and out of a belief that the mostsuccessful management of those harms requires thatthe task be brought into the ‘sphere of legal, trans-parent and constitutionally guided public institu-tions’ (at [50]). There was, the expert maintained, noevidence that the decriminalisation of cannabis incertain states in the United States (for instance,Colorado) had resulted in any spike in criminalconduct.


Davis J, on a balance of the medical evidence,accepted (at [61]) that it was clear that ‘uncontrolledconsumption of cannabis, especially where it isconsumed in large doses poses a risk of harm to theuser’. But the evidence was equally clear that therewas ‘a level of consumption that is safe in that it isunlikely to pose any risk of harm’, although themedical evidence on record was silent on what thatlevel of consumption was.

The court then turned to the position in othercomparable democratic societies for guidance. Therespondents relied on the decision of the SupremeCourt of Canada in R v Malmo-Levine [2003] SCC74, in which a constitutional challenge to the crimi-nal provisions under s 7 of the Charter (providing forthe right to life, liberty and security of the person)was dismissed by the majority on the basis that thecriminalisation of possession of cannabis was ‘notarbitrary but [was] rationally connected to a reason-able apprehension of harm’ (at 642–3). The ‘compel-ling’ minority judgment of Arbour J, on the otherhand, examined the ‘harm principle’, which JohnStuart Mill described thus: ‘The only purpose forwhich power can be rightfully exercised over anymember of a civilized community, against his will, isto prevent harm to others.’ The record, said Arbour J,showed that ‘the prohibition of simple possession ofmarihuana attempts to prevent a low quantum ofharm to society at a very high cost’.

In view of the fact that s 7 of the Charter was not theequivalent of s 14 of our Constitution, which pro-vided for a clearly demarcated right to privacy, DavisJ found the approach of Arbour J to be moreapplicable to the issue before him than that of themajority. After examining decisions in Argentina,Mexico and Alaska, and considering legislationlegalising possession of cannabis in small quantitiesfor personal use in eight states in the United Statesand soon, it seems, in Canada, as well as severalother jurisdictions around the world, he found (at[90]) that there was ‘no longer a consensus that canregard such limitations as justifiable’. The evidence,in his view, ‘supports the argument that the legisla-tive response to personal consumption and use isdisproportionate to the social problems caused as aresult thereof’ (at [102]).

Both ss 4 and 5 of the Drugs Act had, accordingly, tobe amended to ensure that those provisions did notapply to those who used ‘small quantities of can-nabis for personal consumption in the privacy of ahome’, since the present position unjustifiably lim-ited the right to privacy. It would be for the Legisla-

ture to determine the extent of what would constitute‘small quantities in private dwellings’. Even if therespondents were correct in identifying the objec-tions of the legislation, they would still have to showwhy a less restrictive means to achieve them did notexist.

The court expressed itself to be acutely aware of theperils of drug abuse, and so was at pains to stress thenarrow confines of its judgment. It maintained, onthe other hand, that there were several options tofight this problem other than the ‘blunt use of thecriminal law’. The impugned legislation was, as aresult, disproportionate to the harms which the Leg-islature sought to curb. The quantity of cannabis in aperson’s possession constitutes, said Davis J at[109], ‘an objective, established and readily enforce-able basis upon which to distinguish possession forpersonal consumption from dealing or other, moreserious conduct’, so it was both practical and objec-tively possible for future legislation to distinguishthe two forms of use or possession.

The contested statutory provisions (ss 4(a) and 5(b)of the Drugs Act and s 22A(9)(a)(i) of the MedicinesAct, read with their respective relevant schedules)were accordingly declared invalid, but ‘only to theextent that they prohibit the use of cannabis by anadult in a private dwelling where the possession,purchase or cultivation of cannabis is for personalconsumption by an adult’ (at [132]).

(b) Criminal Procedure and Evidence

(i) Pre-sentence

ss 7 and 8: Private prosecution ands 6(2)(e) of the Societies for thePrevention of Cruelty to Animals Act169 of 1993National Society for the Prevention of Cruelty toAnimals v Minister of Justice and ConstitutionalDevelopment & others 2017 (1) SACR 284 (CC)

In the above Constitutional Court case (hereafter the‘CC case’) it was declared that the National Societyfor the Prevention of Cruelty to Animals (hereafterthe ‘NSPCA’) has the statutory power of privateprosecution conferred upon it by s 6(2)(e) of theSocieties for the Prevention of Cruelty to AnimalsAct 169 of 1993 (hereafter the ‘SPCA Act’) as readwith s 8 of the Criminal Procedure Act 51 of 1977.

The above declaration of the Constitutional Courtwas preceded by decisions of the High Court and

25Issue 1, 2017

Supreme Court of Appeal. See National Society forthe Prevention of Cruelty to Animals v Minister ofConstitutional Development & another (unreported,GP case no 29677/2013, 8 October 2014) andNational Society for the Prevention of Cruelty toAnimals v Minister of Justice and ConstitutionalDevelopment & another 2016 (1) SACR 308 (SCA).These two cases are discussed in the notes on s 7 inCommentary, sv Section 7(1)(a): Constitutionality ofexcluding juristic persons from the right to institute aprivate prosecution. In both cases the NSPCA’sconstitutional challenge to s 7(1)(a) of Act 51 of1977, on the basis that it permitted only a privateperson (and not a juristic person) to institute aprivate prosecution, was unsuccessful. In terms ofs 7(1)(a) only private persons (natural persons) areallowed to institute a private prosecution—and thenonly in limited circumstances and after havingobtained a certificate nolle prosequi from the Direc-tor of Public Prosecutions who has jurisdiction overthe alleged offence.

In the CC case it was, however, ultimately notnecessary to make a finding concerning the constitu-tionality or otherwise of s 7(1)(a). Writing for aunanimous Constitutional Court, Khampepe Jpointed out that the NSPCA’s claim that it had a rightprivately to prosecute animal cruelty offences, had tobe determined with reference to s 8 of Act 51 of1977 and s 6(2)(e) of the SPCAAct, as read with theAnimals Protection Act 71 of 1962 (hereafter the‘AP Act’) which provides ‘the groundwork for theanimal protection-regime’ (at [43]). Section 8 of Act51 of 1977 provides for private prosecutions understatutory right: ‘Any body upon which or personupon whom the right to prosecute in respect of anyoffence is expressly conferred by law, may instituteand conduct a prosecution in respect of such offencein any court competent to try that offence’ (emphasisadded).

The crucial issue in the CC case was whethers 6(2)(e) of the NSPCA Act gave the NSPCA thestatutory power to institute private prosecutions asenvisaged in s 8 of Act 51 of 1977. Section 6(2)(e)permits the NSPCA ‘to institute legal proceedingsconnected with its functions. . .’. At [36] KhampepeJ noted that this phrase in s 6(2)(e) can include thepower to prosecute privately because ‘the languageis broad and permissive. . .[and]. . .does not distin-guish between civil and criminal proceedings’. TheSPCAAct, it was said at [38], had to be read with theAP Act. Sections 2 and 2A of the latter Act containan extensive list of offences that constitute animal

cruelty. It was therefore clear that legal proceedingsstemming from the AP Act ‘are most likely to becriminal’ (at [46]). Khampepe J concluded as follows(at [47]–[48], emphasis added):

‘Functionally, the NSPCA is best placed toconduct a private prosecution and give effect topreventing and enforcing the offences set out inthe animal-protection regime. To understandthe SPCA Act as conferring the power of privateprosecution is to give effect to the objects andpurposes of the regime. This construction. . .gi-ves effect to the NSPCA’s primary purpose: toprotect animal welfare. . .. To read s 6(2)(e) asexcluding the right of private prosecutionwould render the regime a toothless tiger. . ..The term ‘‘institute legal proceedings’’ takes ona specific and nuanced meaning in this context. . .including the power to institute private pros-ecutions.’

However, the fact that the NSPCA is now considereda ‘body’ with the right to prosecute privately asprovided for in s 8(1) of Act 51 of 1977, does notmean that the NSPCA’s exercise of this power isbeyond the control or oversight of the Director ofPublic Prosecutions (DPP) concerned. The NSPCA,for example, may only exercise its right after consul-tation with the DPP and after the latter has with-drawn his or her right to prosecute. See s 8(2) of Act51 of 1977. In withdrawing the right to prosecute,the DPP may set conditions he or she deems fit,including a condition that the NSPCA’s appointmentof a prosecutor shall be subject to the DPP’sapproval. See s 8(3) of Act 51 of 1977. Anotherpossible condition mentioned in s 8(3) is that theDPP ‘may at any time exercise with reference to anysuch prosecution any power which he might haveexercised if he had not withdrawn his right ofprosecution’.

It must, finally, be pointed out that the CC specifi-cally refrained from determining the constitutional-ity of s 7(1)(a). The matter was no longer ‘a livedispute’ (at [63]) since the NSPCA’s right to pros-ecute privately was decided without relying ons 7(1)(a). The result is that the Supreme Court ofAppeal’s decision as referred to earlier must stand,namely that the exclusion of juristic persons ins 7(1)(a) is not unconstitutional. However,Khampepe J made it clear that nothing in the CCjudgment ‘should be construed as barring a futurechallenge to that provision, if the appropriate factualscenario arises’ (at [64]).


s 60: Bail: Right to a fair bail hearingS v Kobese (unreported, ECG case no CA&R34/2017, 23 February 2017)

In the above matter the appellant appealed unsuc-cessfully against the refusal of a magistrate in theGrahamstown district to release him on bail pendingthe outcome of his trial on charges of having raped a14-year-old girl on three separate occasions andhaving assaulted another child with intent to dogrievous bodily harm. At the bail hearing and onappeal it was common cause that on account of therape charges, the burden of proof was on the accusedas provided for in s 60(11) of Act 51 of 1977. On thenature of the bail applicant’s burden in this regard,see the discussion of s 60 in Commentary, sv Section60(11) and the burden of proof.

In Kobese Malusi J noted, with reference to s 65(4),that a court hearing an appeal against a refusal of bailwill not set aside the bail court’s refusal ‘unless-. . .satisfied that the decision was wrong’ (at [10]).See also the discussion of s 65 in Commentary, svApproach on appeal: s 65(4). Indeed, in Kobese itwas clear that the appellant had failed to dischargehis burden of proving on a balance of probabilitiesthat there were exceptional circumstances which inthe interests of justice permitted his release (at [14],[17] and [19]).

However, in dismissing the appeal (at [23.1]),Malusi J also made an order requiring the Registrarto ensure that a copy of the High Court’s judgment inthe bail appeal be delivered to the Chief Magistrate,Grahamstown, so that the latter could—within areasonable time from delivery of the High Courtjudgment—facilitate ‘sensitivity training’ for themagistrate who had presided at the bail hearing andhad refused bail (at [23.2]).

The above order concerning ‘sensitivity training’was a consequence of certain unacceptable remarksor observations made by the bail magistrate. In herex tempore judgment the bail magistrate, havingstated that in the Grahamstown community ‘childrenare abandoned or raped by their parents’, askedrhetorically whether it would be safe if the appellantwere released on bail in this ‘situation’. The magis-trate claimed that her remarks were based on her fouryears residence in Grahamstown as well as herexperience in the Children’s Court. However, MalusiJ pointed out that her remarks were ‘unwarranted’and that there was ‘no evidence of such a situationnor. . .[was]. . .it such a notorious fact for her to takejudicial notice of it’ (at [20]).

In her ex tempore judgment the bail magistrate alsostated that ‘we are living in a country-. . .[where]. . .we have. . .men who are animals. . .’.According to Malusi J this observation was ‘[t]hemost egregious comment by the magistrate’ andamounted to gender stereotyping that could ‘well bein breach of the equality clause in section 9(3) of theConstitution’ (at [21]).

The order regarding sensitivity training was abso-lutely necessary. However, it would appear that thereis ample room for an argument that the bail magis-trate’s unacceptable and objectionable remarks in thecourse of her ex tempore judgment not only fatallytainted her impartiality but also constituted a grossirregularity that was of such a fundamental naturethat there was a failure of justice which rendered thebail proceedings a nullity. See, for example, S vSibeko en ’n ander 1990 (1) SACR 206 (T). InSibeko a bail magistrate had earlier during the policeinvestigation taken down a confession by one of thebail applicants. It was held that the magistrate couldnot thereafter have presided at the bail hearing. Theirregularity committed was of such a nature that thebail proceedings were declared a nullity. It wasfundamental to the administration of justice, saidKriegler J in Sibeko at 207i–j, that a presidingjudicial officer should be an impartial and open-minded (‘onbevange’) adjudicator who decides mat-ters on the basis of information placed before him orher in the usual way.

It is furthermore submitted that the bail hearingpresided over by the magistrate in Kobese fell wellshort of what was said in S v Majali (unreported,GSJ case no 41210/2010, 19 July 2011) at [33]: ‘Abail inquiry is a judicial process that has to beconducted impartially and judicially and in accor-dance with relevant statutory and constitutional pre-scripts.’ It must also be appreciated that the constitu-tional right to a fair trial includes proceedings suchas bail proceedings. Fairness and the constitutionalright to a fair bail hearing are manifestly under threatwhere a bail magistrate—like the one in Kobese—compromises her impartiality by making the remarksand observations which she did in the course of an extempore judgment dealing with the question whethera male bail applicant charged with the rape of ayoung girl should be released on bail in the district ofGrahamstown. What our law requires is that a bailmagistrate, like any judicial officer presiding over atrial, should conduct proceedings open-mindedly,impartially and fairly, and that such conduct mustindeed be manifest to all concerned, especially the

27Issue 1, 2017

bail applicant. See the instructive pronouncementsmade by Ponnan JA in S v Le Grange & others 2009(1) SACR 125 (SCA) at [14] as regards trial pro-ceedings. At [18] Ponnan JA also made the followingtelling observation: ‘A judicial officer can onlyperform his demanding and socially important dutyproperly if he also stands guard over himself, mind-ful of his own. . .personal views.’

Regrettably, it would seem that in Kobese the bailmagistrate’s personal views on rapes of children inthe Grahamstown district as well as her views on theconduct of some men in South Africa, had taintedher impartiality to the extent that the bail applicant’sright to a fair bail hearing was fatally compromised.The bail proceedings should have been declared anullity, accompanied by an order—like the one inSibeko (supra) at 208c—that a fresh bail applicationbe heard by another magistrate on a specific dayshortly after the date of the order of nullity.

The fact that there were sufficient grounds in Kobesefor the refusal of bail, need not necessarily be anobstacle to a declaration of nullity. An objective testcomes into play. It is well known that a vital elementof a fair trial is that justice must not only be done,but must also be seen to be done. In the context ofKobese this would mean that a reasonable andinformed person’s perception of bias on the part ofthe bail magistrate, would have the same impact asactual bias in that the proceedings are vitiatedwhether the bias is perceived or real. On the natureof the objective test in assessing the impact ofalleged bias on fair trial principles, see generally thediscussion of s 145 in Commentary, sv Guidelinesfor the recusal of judicial offıcers (including asses-sors).

Kobese (supra) calls for one more comment. Ourcourts, and especially the Supreme Court of Appeal,have on numerous occasions for purposes of sen-tencing noted that rape has become prevalent notonly in the specific jurisdiction of the court con-cerned but countrywide as well. See S v Mgibelo2013 (2) SACR 559 (GSJ) at [4]; S v SMM 2013 (2)SACR 292 (SCA) at [14]; S v Nkunkuma & others2014 (2) SACR 168 (SCA); S v Mhlongo 2016 (2)SACR 611 (SCA) at [25]. Indeed, in S v Hewitt 2017(1) SACR 309 (SCA) it was noted for purposes ofsentencing that Interpol has identified South Africaas ‘the rape capital of the world’ (at [9]). However,there is a vast and fundamental difference betweenthe situation where a sentencing court notes theprevalence of rape for purposes of determining asuitable sentence for a convicted rapist and the

situation where a bail court which has to decide onthe pre-trial liberty of an alleged rapist, makesobservations like the ones made by the bail magis-trate in Kobese (supra).

It is, in the final analysis, submitted that Malusi J’sorder regarding sensitivity training for the bail mag-istrate in Kobese was appropriate for purposes ofassisting her to realise that judicial duties must beperformed with sensitivity and circumspection. Butthis order only addressed the future conduct of themagistrate. It is respectfully submitted that the courtof appeal in Kobese should also have addressed abroader but very relevant issue, namely the impact ofthe bail magistrate’s comments on the appellant’sconstitutional right to a fair bail hearing. In fact, at[20] it was noted that the comments of the bailmagistrate could have caused accused personsappearing before her to experience ‘a sense ofinjustice’. Must one assume that the bail applicanthimself had not experienced such a sense of injus-tice? Or that the reasonable person in the position ofthe bail applicant would not have experienced such asense of injustice? It is respectfully submitted thatthe appeal court in Kobese should also have exam-ined and answered the following question: did thecomments of the magistrate constitute a fatal irregu-larity that nullified the bail proceedings?

ss 151 and 166: Onus ofproof—implications of. Judicial officerdescending into the arena—how to dealwith inexperienced prosecutionS v Quayiso 2017 (1) SACR 470 (ECB)

The accused had been charged with two counts ofassault with the intention to cause grievous bodilyharm arising out of an incident in which he wasalleged to have assaulted two men by stabbing themwith a knife. The men had allegedly robbed him ofhis cell phone the previous day. The magistratefound the evidence of the second complainant to beunreliable and acquitted the accused on the secondcount. On the first count, the magistrate convictedthe accused, even though she accepted that hisversion that he had acted in self-defence was reason-ably possibly true, on the basis that he had exceededthe bounds of self-defence by causing excessiveharm to the first complainant.

This, said Mbenenge J (with whom Van Zyl Jagreed), the magistrate was not entitled to do.Having accepted the version of the accused—that hehad acted in lawful self-defence—she was notentitled to reject it on the basis that the accused had


exceeded the bounds of that defence. That was neverthe case of the prosecution in the first place and itflew in the face of the accused’s version (which sheaccepted) that he had been threatened by the com-plainants and their companions and even received astab wound in his thigh from one of them.

Mbenenge J cited this helpful and important test laiddown in S v Sithole & others 1999 (1) SACR 585(W) at 590g–i:

‘There is only one test in a criminal case, andthat is whether the evidence establishes the guiltof the accused beyond reasonable doubt. Thecorollary is that an accused is entitled to beacquitted if there is a reasonable possibility thatan innocent explanation which he has profferedmight be true. These are not two independenttests, but rather the statement of one test,viewed from two perspectives. In order toconvict, there must be no reasonable doubt thatthe evidence implicating the accused is true,which can only be so if there is at the same timeno reasonable possibility that the evidenceexculpating him is not true. The two conclu-sions go hand in hand, each one being thecorollary of the other. Thus in order for there tobe a reasonable possibility that an innocentexplanation which has been proffered by theaccused might be true, there must at the sametime be a reasonable possibility that the evi-dence which implicates him might be false ormistaken.’

The propositions stated in Sithole may be axiomatic,but it is remarkable how often they are overlooked orwrongly applied. It may be helpful, then, to set outthis further explanation by Nugent and SchwartzmanJJ in Sithole (at 590–1), which was not repeated inQuayiso:

‘Whichever way one phrases the test, it is to beapplied upon an assessment of all the evidence,and not by a process of piecemeal reasoning. Inother words, it cannot be applied by lookingonly at the evidence of the State, or the accused,in isolation. It may be that the evidence of theState is such that the conflicting evidence of theaccused must, by a process of logical reasoning,be untrue, or it may be that the evidence of theaccused is such that the possibility that it maybe true cannot be said to be excluded by theState’s evidence; but in either event a courtmust bear in mind all the evidence when reach-ing the appropriate conclusion. Thus, for

example, where there is evidence that theaccused committed a particular act, and hisevidence is that he was not present at the timethe act was committed, quite obviously the twoassertions cannot both be true. There cannoteven be a reasonable possibility that they areboth true. There can only be a reasonablepossibility that the accused’s alibi is true if thereis at the same time a reasonable possibility thatthe State’s evidence is false or mistaken.’

This pronouncement encapsulates the correct posi-tion with admirable clarity.

Apart from the erroneous application of the lawrelating to the onus, the court in Quayiso had otheranxieties with the approach of the magistrate.

First, a number of questions were put to the accusedabout his version after he had been cross-examinedby the prosecutor. The questions were ‘hardly ques-tions to get clarity on certain unclear issues, butconstituted cross-examination, with the result thatthe magistrate descended into the arena, transgress-ing the guidelines stipulated in S v Rall [1982 (1) SA828 (A) at 831–2]’ (at [25]). Rall’s case and severalothers dealing with this issue are discussed in Com-mentary in the notes to s 166 sv Questioning by thecourt.

The impression that the magistrate had descendedinto the arena was further strengthened by herunjustified interference in the cross-examinationconducted by the defence. The stabbing incident hadbeen triggered by the alleged robbery involving thecell phone. The magistrate disallowed cross-exami-nation of the first complainant on the issue of therobbery on the basis that the case was not about acell phone, but an assault. This was clearly unwar-ranted, and by so acting she had, said the court (at[23]), ‘descended into the arena and associatedherself with the prosecution, thus contaminating theproceedings by not acting in an open-minded andimpartial fashion’.

The ‘magistrate’s lopsided approach’, saidMbenenge J (at [26]), ‘once again reared its uglyhead’ when she thwarted his attempt to show thecourt the stab wound he had sustained to his thigh.Since the accused claimed to have been acting inself-defence, this was an important part of his case.The magistrate’s refusal was a sign that she‘favoured the State in a way that could ‘‘only lead tothe administration of justice falling into disreputeand a perception of bias’’ on her part’ (followingClaasen J in S v Du Plessis 2012 (2) SACR 247(GSJ) at [25]).

29Issue 1, 2017

The final irregularity concerned the re-opening of theState’s case, which had been done improperly afterthe closure of the case for the State, with no formalapplication having been made to re-open the case,and without an opportunity being afforded to thedefence to cross-examine the recalled witness, thesecond complainant. The State’s case was re-openedbecause there was no proof that the witness had beenstabbed, as the medical form (J88) was not in thedocket. The witness was recalled to show the courtthe stab wound to his shoulder, a ‘benevolence . . .not extended to the accused’ (at [30]), who was notpermitted to show the court the wound to his thigh.

To make matters worse, the magistrate seemed toshift the blame for this irregularity to the ‘inexperi-ence of aspirant prosecutor’ (at [28]), as she put it ina response to a query by the regional magistrate. Thisresponse, said Mbenenge J, left ‘much to be desired’.It remained the duty of magistrates to ensure thatproper procedures were followed at all stages in atrial. There was a ‘duty cast on the magistrate toguide the inexperienced prosecutor where there wereindications of him floundering, as she is not merely afigurehead’. Her duty was ‘not only to direct andcontrol the proceedings according to recognisedrules of procedure, but to see that justice [was] done’(at [29], the court relying on R v Hepworth 1928 AD265 at 277, followed in S v Rall supra and S vMusiker 2013 (1) SACR 517 (SCA) at [18]).

In the trial justice was conspicuously not done; theproceedings were vitiated and the conviction was setaside.

s 152: Reconstruction of the record: Canthe accused waive his or her rights toparticipate in the process?S v Schoombee & another 2017 (5) BCLR 572 (CC)

Whether an accused can waive his or her constitu-tional rights in respect of the proper procedures to befollowed in reconstructing a lost record of the trialwas considered, but not answered, by the Constitu-tional Court in this case. The applicants inSchoombee, when they had sought to appeal againsttheir convictions for murder, discovered after muchsearching that the record of their trial proceedingshad been lost. A ‘reconstructed record’ had beenprepared by the trial judge on the basis of the noteshe had taken during the proceedings. After somehesitation, both applicants proceeded to appeal onthis record—one against conviction and sentence,the other only against sentence. Both appeals were

dismissed by the full court, before which neitherapplicant raised the issue of the inadequate record.

It was contended before the Constitutional Court thatthe criminal process was flawed, the inadequatereconstruction of their trial leading to a violation oftheir constitutional right to a fair trial. Without aproper trial record, it was argued, there could be noappeal; with no appeal, there could be no fair trial.The decision to proceed with the appeal using thereconstructed record, they contended, was taken outof desperation after years of delay, and after beingadvised by counsel that they had reasonable pros-pects of success on the record as it stood.

The Constitutional Court set out various aspects ofthe law relating to the relation between the recon-struction procedure and a fair trial, the methods thatmay be employed in reconstructing a record and theresponsibilities of the State, the accused and thecourt in engaging in such a process (as to which seethe notes on s 152 of Commentary, as well as thenotes on ss 158 and 309). The appellant (or legalrepresentative) ‘carries the final responsibility toensure that the appeal record is in order’ (at [21],quoting from S v Sibelelwana [2012] ZAWCHC 150(unreported, WCC case no A401/2011, 3 August2012) at 10). At the same time, said the court, it wasfor the reviewing court ‘to ensure that an accused isguaranteed the right to a fair trial, including anadequate record on appeal, particularly where anirregularity is apparent’ (see, too, S v Zuma 1995 (2)SA 642 (CC) at [16]).

In this case, by choosing not to pursue a reconstruc-tion process, could it be said that the applicants hadwaived their right to a fair trial regarding theirparticipation in such a process? The State argued thattheir decision to launch their appeal using the‘judge’s imperfectly reconstructed’ record was sucha waiver. Was it? This was the Constitutional Court’sanswer (at [25]):

‘Perhaps. This Court has emphasised thatwaiver of a constitutional right is difficult. Thebar is high. To waive a right, a party mustintentionally and knowingly abandon it. Theonus to prove waiver is strictly on the partyasserting it—here, the State. Even so, this Courthas questioned whether waiver is applicable inrelation to constitutional rights. And it hasnoted the distinction between waiver in thecontractual sense and a mere choice not toexercise a constitutional right.’


Although the facts here did ‘suggest a possiblewaiver’ (at [26]), the court found it unnecessary, onthe facts, to determine either whether the applicantsdid waive their rights or whether it was even possiblefor them to have done so. This was because theapplicants had a fair trial, including a fair appeal.Although the record of their trial had been improp-erly and imperfectly reconstructed, it was ‘more thanadequate to ensure the applicants exercised theirconstitutional right to appeal’ (at [27]): the notestaken by the trial judge were ‘unusually full anddetailed’ and were not ‘scrappy, telegram-style anno-tations’; they appeared to be a complete narrativeaccount of the evidence led in the trial; they recordedthe witnesses’ evidence-in-chief as well as theircross-examination; and they gave a full picture ofboth the events on the night of the murder as well asthe trial proceedings themselves.

As the Supreme Court of Appeal held in S v Chabedi2005 (1) SACR 415 (SCA), a reconstructed recordneed not be perfect; it need only be adequate for aproper consideration of the appeal. Whether or not itis so is a question that ‘cannot be answered in theabstract’, but depends, inter alia, on ‘the nature ofthe defects in the particular record and on the natureof the issues to be decided on appeal’ (Chabedi at[5]–[6]). Thus, said the Constitutional Court (at[29]), it was ‘practical and sensible and just’, wherethe adjudication of an appeal on an imperfect recordwould not prejudice the appellants, not to set asidetheir convictions ‘solely on the basis of an error oromission in the record or an improper reconstructionprocess’ (see, too, S v Machaba 2016 (1) SACR 1(SCA) at [4]–[5]; see, too, Commentary in the notesto s 152 sv Reconstruction of the record).

These principles, said the court, applied inSchoombee’s case for these reasons: the reconstruc-tion was detailed and specific; the applicantsreviewed this record; they took the advice of coun-sel; they chose to proceed with their appeal on theimperfect record in accordance with counsel’sadvice; and, even if they did not waive their right toparticipate in reconstruction, they ‘certainly signifiedtheir assent to the substantive recital contained in thereconstructed record’ (at [30]). The court warned,however, that ‘[n]one of this detracts from themagnitude of the lapses that took place in construct-ing the record’ (at [38]). It stressed that the HighCourt had failed to ensure, as it was obliged to do,that the reconstruction process involved both parties.The duty to ensure that the process complied withthe right to a fair trial was one that had to be

‘undertaken scrupulously and meticulously in theinterests of criminal accused as well as their vic-tims’.

It noted, finally, that the loss of trial court recordswas a ‘widespread problem’ which ‘raised seriousconcerns about endemic violations of the right toappeal’. Reconstruction, it lamented, ‘should not bethe norm in providing appellants with their trialrecords’.

s 208: Identification evidence guidelinesand pitfallsS v Kotze [2017] ZASCA 27 (unreported, SCA caseno 776/16, 27 March 2017)

S v Madondo (unreported, KZP case no AR 350/2016, 8 December 2016)

‘Authorities dealing with the dangers of incorrectidentification’, said Mocumie JA in Kotze, ‘arelegion’. This is true, and it is instructive to considerhow the court in that case navigated through thoseauthorities in determining whether the complainanthad made an accurate and reliable identification ofthe appellant.

The appellant had been convicted of housebreakingwith intent to commit indecent assault, as well asindecent assault itself. Much turned on the complain-ant, a 13-year-old girl. The trial court and the fullcourt found the complainant to be an honest andimpressive witness who had no reason falsely toimplicate the appellant. In the appeal to the SupremeCourt of Appeal, the accuracy of the identificationwas attacked.

Mocumie JA, who delivered the judgment of thecourt, referred to the ‘locus classicus’ in this regard,S v Mthetwa 1972 (3) SA 766 (A) at 768A, where thecourt warned that ‘because of the fallibility of humanobservation, evidence of identification is approachedby the Courts with some caution’. Reference wasalso made to this passage in the court’s judgment inR v Dladla & others 1962 (1) SA 307 (A) at 310C–E(see, too, S v Arendse [2015] ZASCA 131 (unre-ported, SCA case no 089/15, 28 September 2015)):

‘One of the factors which in our view is of thegreatest importance in a case of identification isthe witness’ previous knowledge of the personsought to be identified. If the witness knows theperson well or has seen him frequently before,the probability that his identification will beaccurate is substantially increased. Even in thecase when a witness has some difficulty in thewitness-box in giving an accurate description of

31Issue 1, 2017

the facial characteristics and clothes of theperson whom he has identified, the very factthat he knows him provides him with a pictureof the person in the round which is a summaryof all his observations of the person’s physiog-nomy, physique and gait, and this fact willgreatly heighten the probability of an accurateidentification. . .. In a case where the witnesshas known the person previously, questions ofidentification marks, of facial characteristics,and of clothing are in our view of much lessimportance than in cases where there was noprevious acquaintance with the person sought tobe identified. What is important is to test thedegree of previous knowledge and the opportu-nity for a correct identification, having regard tothe circumstances in which it was made.’

In Kotze’s case there were circumstances thatenhanced the reliability of the identification. First,there was unrefuted evidence that the complainantknew the appellant prior to the incident. There wasalso unchallenged evidence that of all the peoplewho visited at the house in question on the eveningof the assault, the complainant knew only the appel-lant.

Secondly, the appellant was able to identify theappellant on three separate occasions on the night inquestion. The first was when she saw him inside herbedroom the first time, after he had switched on thelight. On that occasion there was instant recognition.The second was when he returned to that room toassault her. On that occasion she called him by hisname and, on the strength of his silhouette, correctlydescribed his physique as ‘tall and chubby’. He had,she said, smelled of liquor, and it was common causethat the appellant was intoxicated that night. It wasclear, further, that she could not have known thisunless her evidence was true. The third occasion waswhen she was hiding in the garden after she had fledfrom her attacker. She saw him leave the main housethrough the kitchen door and described him as beingunsteady on his feet, adding that he then walked to anearby flatlet and sat on a bench next to the slidingdoor. The light at the back door was on and so wasthe light at the door of the flatlet. The appellant waslater found asleep on this bench and it was clear fromthe evidence that the complainant could never haveknown that the appellant had gone to sit on the benchunless she had seen him.

On the strength of this evidence and this rigorousanalysis of that evidence, the court accepted that thecomplainant’s identification of the appellant as the

person who assaulted her was ‘undoubtedly bothhonest and reliable’ (at [20]).

A different conclusion was reached by the court inS v Madondo (unreported, KZP case no AR350/2016, 8 December 2016). The appellant’s convictionfor robbery with aggravating circumstances in thatcase depended upon the identification of him as therobber by the complainant. There were several prob-lems relating to this identification: the complainantdid not know the appellant or his name; he had notseen the faces of his two assailants clearly; the onlyillumination was provided by street lights some eightto nine metres away; it was already dark when theattack occurred; the event was a very emotional andtraumatising one, which ‘would no doubt haveaffected his powers of observation and recall’ (at[5]); and both the estimation by the appellant of thetime of the event and, as a result, the opportunity toobserve, were open to doubt since the ‘harrowingnature thereof probably made it feel much longerthan it actually was’.

But added to these concerns were two extremelytroubling features of the complainant’s account. Thefirst was that when he was taken to the home of theappellant by a co-accused, he claimed to haverecognised the appellant as the one who had robbedhim ‘by his dreadlocks as well as his complexion’.The problem was that he had not provided thisdescription to any law enforcement agency or evento any of the many members of the community whohad been present at his home before this identifica-tion. The second was that the co-accused had takenhim to the appellant’s home in the first place. He hadnot taken him there freely and voluntarily, and itseems that this was the result of the unfortunateconduct of members of the community—numberingbetween 20 and 30—who had decided to take theinvestigation into their own hands and had assaultedthe co-accused before he agreed to lead them to thehouse ‘of the one who had been with him when theycommitted the robbery’ (at [7]).

These circumstances raised the ‘real danger that theseemingly positive identification of the appellantmight have been caused or influenced (possiblyexclusively) by the conduct of [the co-accused]leading the complainant and the others to the appel-lant as a co-assailant’ (at [9]). The position, saidKoen J (with whom Chetty J agreed), would obvi-ously have been different if the complainant had,right from the outset, mentioned the identifyingfeatures to the members of his community.


As for the statement made by the co-accused to thecrowd following the assault, it was obviously inad-missible against the appellant. This was so, apartfrom the fact that it was not made freely andvoluntarily, because it constituted an extra-curialadmission which could, in line with what was held inS v Litako & others 2014 (2) SACR 431 (SCA), notbe admitted against a co-accused (see the discussionin Commentary in the notes to s 219 sv Is anadmission (as opposed to a confession) by anaccused admissible against a co-accused?)

There was, as a result, a reasonable possibility thatthe complainant had mistakenly identified the appel-lant. The conviction was set aside.

ss 217, 219, 166: Admissibility ofconfessions under the CriminalProcedure Act and the ConstitutionS v Mudau [2017] ZASCA 34 (unreported, SCA caseno 1148/2016, 29 March 2017)

The trial of the two appellants, in which they wereconvicted of murder, assault with intent to commitgrievous bodily harm and robbery with aggravatingcircumstances, contained so many serious misdirec-tions, gross procedural irregularities, infringementsof the statutory provisions governing confessionsand violations of their constitutional right to a fairtrial, that the Supreme Court of Appeal had no optionbut to exclude the confessions in question and to setaside the convictions.

The recording of the confessions in writing by themagistrate was ‘replete with omissions, incoherentand contradictory recording of answers by the appel-lants to questions, and serious non-adherence tosome of the fundamental principles governing con-fessions’ (at [15]). Some of these irregularities were:

(1) In the first appellant’s confession form, theparagraph setting out his constitutionalright to remain silent and the right to legalrepresentation, was left blank and notcompleted. So, too, was the paragraphwhich questioned whether the deponentwas in his sound and sober senses and theparagraph which inquired whether theaccused wished to make a statement not-withstanding what had been explained tohim.

(2) When questioned about a further blankparagraph, the magistrate, in his viva voceevidence, conceded that the first appellanthad undoubtedly answered in the negative

to the question whether or not he wished tomake a statement. In the form itself, hesaid expressly that he did not wish orprefer to make a statement. The fact thatthe magistrate still proceeded to record theconfession and the fact that the trial courtruled the confession to be admissible‘resulted in a serious miscarriage of justiceand rendered unfair the first appellant’strial’ (at [19]), per Mbha JA, who deliv-ered the judgment of the court).

(3) The trial court also ‘inexplicably’ ignoredor overlooked the testimony of a policeinspector who conceded that he ‘dulyadvised the first appellant of the advan-tages of making a confession, namely, thatsuch confession was going to allow thematter to proceed quickly’ (at [19]). This,said Mbha JA, amounted to undue influ-ence.

(4) The statement of the first appellant was,moreover, wrongly classified as a ‘confes-sion’. He claimed that he had acted underthe coercion of the second appellant, withthe result that the statement did not consti-tute an unequivocal admission of everyelement of the offence.

(5) The form relating to the second appellant’sconfession also showed material omis-sions. The magistrate omitted, for instance,to record the answer given to the questionwhether he wished to make a statementand whether or not he had been assaultedor coerced into making a statement. Themagistrate also failed to record his ‘obser-vation whether the second appellant was inhis sound and sober senses with specificreference to anxiety, nervousness, jovialityand demeanour’. These qualities were rel-evant since the appellant claimed in histestimony that, when he was brought to themagistrate to make a confession, he was‘dizzy because he felt he was being forcedto admit an offence which he did notcommit’ (at [22]). Even more disturbing,in the view of Mbha JA, was the answergiven by the second appellant to the ques-tion of how it came about that the he wasbrought to make the confession. Theanswer—that he had been advised by thepolice inspector to report to the magis-trate—was simply recorded by the magis-

33Issue 1, 2017

trate, who should have seen this answer asa ‘red flag’ and followed up by askingquestions to ensure that the appellant wasnot influenced in any way into making theconfession.

Mbha JA then turned to the irregularities committedby the trial judge:

(1) The trial judge had made an erroneousruling on the onus of proof by maintaining,at the commencement of the trial within atrial, that the appellants had first to adduceevidence to prove that the confessionswere not freely and voluntarily made orwere not made without undue influence.The onus, as Mbha JA pointed out, restsalways on the State, and the ‘erroneousshifting of the onus of the appellants ren-dered their trial unfair’ (at [23]).

(2) The trial judge ignored the fact that thesame magistrate recorded both appellants’confessions, which were made a weekapart. There was nothing in the record toindicate that he made any attempt to checkwhether another magistrate was availableto take down the second confession, and itwas ‘doubtful’ whether he had sufficientlycautioned himself not to be influenced byhis prior knowledge of the first confession.

(3) The trial judge applied the confessions andadmissions made by the (then) threeaccused against one another in clear viola-tion of s 219 of the Criminal ProcedureAct (see Commentary on s 219). Instead oftreating each confession separately againstits specific maker, he treated all of them ‘inblanket fashion against all the accused’ (at[28]).

(4) In addition to these irregularities in rela-tion to the confessions, there was onefurther irregularity that sustained the con-clusion that the judge ‘was not open-minded, impartial and fair during the sec-ond appellant’s trial within a trial’ (at[27]). After the second appellant had givenhis evidence-in-chief at the trial within atrial, and before the prosecutor couldcross-examine him, the trial judge said tothe prosecutor:

‘I do not know if you would like tocross-examine this witness. He madea very bad impression. I do not know,

as a witness really. I do not knowwhether it would be worth for you tocross-examine him, but you can havethat time. But he made a bad impres-sion, really, I do not know. Do youhave any questions for the witness?’

This ‘unfortunate remark’ showed that the trialjudge had already decided that he was not goingto accept the second appellant’s evidence, and,said Mbha JA, ‘a perception of bias on the partof the trial judge by any reasonable person was. . . inevitable’ (at [25]). See, further, S v Basson2007 (3) SA 582 (CC) at [27], Van Rooyen &others v The State & others (General Council ofthe Bar of South Africa intervening) 2002 (5)SA 246 (CC) at [32]—which followed whatwas said by the Canadian Supreme Court inValente v The Queen [1985] 2 SCR 673—andS v Le Grange & others 2009 (1) SACR 125(SCA) at [16]. See also, on the need for thejudicial officer to avoid descending into thearena, Commentary on s 166 sv Questioning bythe court.

(ii) Sentencing

Sentencing and trial fairness: Noreference in charge-sheet to minimumsentence legislationS v Tshoga 2017 (1) SACR 420 (SCA)

Tshoga (supra) was an appeal against life imprison-ment imposed by the High Court on the appellant forthe rape of a 10-year-old girl. This sentence was—inthe absence of substantial and compelling circum-stances—the prescribed sentence in terms of s 52(1)as read with s 51(1) and Part I of Schedule 2 to theCriminal Law Amendment Act 105 of 1997 (at [1]and [5]).

One of the main grounds of appeal was that theappellant was at no stage prior to conviction warnedthat he faced a possible sentence of life imprison-ment in terms of s 51(1) of Act 105 of 1997 (at [8]).This, it was argued, constituted an unfair trial.

The following facts were common cause: the charge-sheet did not refer to Act 105 of 1997; at no stageprior to conviction did the trial magistrate or pros-ecutor refer to Act 105 of 1997; the first time themagistrate mentioned s 51 of the Act was afterconviction when he informed the appellant thathaving been convicted of the rape of a girl under theage of 16 years, it was in terms of s 51 of Act 105 of


1997 required that the appellant’s case be referred tothe High Court for purposes of sentence (at [9]). Itshould be noted that at the time of the appellant’strial this was indeed the procedure that was pre-scribed by Act 105 of 1997. See generally thediscussion under the repealed s 277 of Act 51 of1977 in Commentary, sv Minimum sentences forcertain serious offences: A brief overview of Act 105of 1997 as amended by Act 38 of 2007.

In Tshoga (supra) the majority (per Schoeman AJA,Dambuza JA and Nicholls AJA concurring) foundthat the appellant ‘did have a fair trial’ (at [26]) andhad ‘suffered no prejudice, in the circumstances ofthe case’ (at [27]). The majority referred to an earlierdecision of the court in S v Kolea 2013 (1) SACR409 (SCA) where it was held that if a charge-sheethad incorrectly referred to s 51(2) instead of s 51(1)of Act 105 of 1997, it could not—in the absence ofany prejudice to the accused—preclude a sentencingcourt from imposing the prescribed minimum sen-tence as if there was no incorrect reference in thecharge-sheet. In Kolea a full bench held that theappellant had on account of at least three reasonssuffered no prejudice: he had legal representation athis trial and ‘well knew of the charge he had to meet’and that the prosecution ‘intended to rely on theminimum-sentencing regime created’ by Act 105 of1997 (at [11]); it was never ‘demonstrated’ that theappellant would have conducted his defence differ-ently ‘had the mistake not been made in the charge-sheet’ (at [14]); trial unfairness and prejudice werefor the first time claimed before the Supreme Courtof Appeal (at [15]).

The majority in Tshoga took care to point out that thefacts in Tshoga were not comparable with those inKolea: in Tshoga the charge-sheet made no referenceto Act 105 of 1997 whereas the charge-sheet inKolea did refer to that Act but contained an incorrectreference to s 51(2) instead of s 51(1) (see Tshoga at[20]). However, the majority in Tshoga at [22] wassatisfied that the true principle applied in Kolea wasthat ‘substance was of paramount importance andthat form was secondary’. At [22] in Tshoga Schoe-man AJA also explained as follows:

‘I am of the view that a pronouncementthat. . .[Act 105 of 1997]. . .had to be men-tioned in the charge-sheet or at the outset of thetrial would be elevating form above substance.Every case must be approached on its own factsand it is only after a diligent examination of allthe facts that it can be decided whether anaccused had a fair trial or not.’

The majority in Tshoga, relying on its own interpre-tation of the approach adopted in Kolea, concludedthat the cumulative effect of all the circumstanceswas such that there was indeed no prejudice (at[23]). The appellant had pleaded not guilty to therape charge, had ‘participated fully’ in the trial andhad ‘effective legal representation throughout thetrial until his conviction and the transfer to the HighCourt’. After his conviction there was no objectionwhen the regional court ordered that the case, interms of then-existing procedures, had to be trans-ferred to the High Court for the latter to consider theapplicability of prescribed minimum sentences,which included consideration of life imprisonment.After transfer to the High Court, no complaint ofpossible prejudice was raised despite four furtherproceedings (two sentencing procedures in the HighCourt and two appeals to the full bench of the HighCourt) during which the appellant was legally repre-sented (at [23]). The appellant therefore had legalrepresentation at all relevant stages, from the trial tothe various High Court proceedings, and had neveralleged prejudice.

The complaint of prejudice, said the majority at [23],was ‘only belatedly raised’when the matter reached theSupreme Court of Appeal. At [24] it was pointed out,too, that appellant’s counsel could in argument beforethe court also not identify any prejudice the appellanthad suffered as a result of ‘the failure to mention. . .[Act 105 of 1997]. . .in the charge-sheet or at theoutset of the trial’, except for the possibility of preju-dice in that the appellant, if properly informed, mighthave tendered a plea of guilty. This possibility,observed the majority, was remote since the appel-lant—some eight years after the incident—had ‘stillprofessed his innocence’ to a probation officer who hadcompiled a pre-sentence report for purposes of thesecond sentencing procedure in the High Court, wherethe accused had legal representation but chose not totestify in mitigation (at [6] and [24]).

The minority (per Bosielo JA, Tshiqi JA concurring)took the view that it was grossly unfair to inform anaccused only after conviction of the applicability ofsuch a ‘patently serious’ matter as ‘the applicabilityof the minimum-sentence legislation’ (at [47]). Thetrial in Tshoga, said the minority at [47], was ‘a trialby ambush, which is neither desirable nor permis-sible in a constitutional democracy underpinned by aBill of Rights’ (at [47]). The minority considered thatthe prosecution’s failure to refer to Act 105 of 1997in the charge-sheet, as well as its further failure toapply for the charge-sheet to be corrected as pro-

35Issue 1, 2017

vided for in s 86(1) of the Criminal Procedure Act,inevitably caused prejudice to the accused and hadrendered his trial unfair (at [46] and [52]).

At least two matters can be raised in an attempt toexplain the differences between the minority andmajority judgments.

First, the majority—unlike the minority—acceptedfor purposes of the merits of the case (that is,whether the appellant should have been convicted oracquitted on the charge of raping a 10-year-old girl)that the absence of any reference to Act 105 of 1997could not have caused prejudice, especially since theaccused—assisted by a competent legal representa-tive—had participated fully in his trial. After convic-tion an awareness that the minimum-sentence legis-lation would for sentencing purposes be considered,was essential. But it was precisely at this stage thatthe magistrate had indeed informed the appellant ofthe applicability of such legislation, and the legalrepresentative at the trial had raised no objection tothe referral of the matter to the High Court forpurposes of the latter’s consideration of the mini-mum-sentence legislation. The minority, however,took a firm stance that the appellant’s fair trial rightwas breached right from the start when he wasrequired to plead to a charge that was incomplete andnot later corrected. See Tshoga at [50] and [52].

Second, the majority as well as the minority acceptedthat the facts and issues in Tshoga were not coveredby the decision of the full bench of the SupremeCourt of Appeal in Kolea (supra). See Tshoga at [22]and [51]. However, the majority—unlike the minor-ity—accepted that Kolea indicated that in assessingthe presence or absence of prejudice, form shouldnot be elevated to substance. It was this principlewhich ultimately formed the basis of the majority’sdecision. And it is, with respect, on the basis of thisprinciple that the majority’s decision should besupported.

s 276B: Procedural fairness where courtconsiders a non-parole periodS v Ndlovu [2017] ZASCA 26 (unreported, SCA caseno 925/2016, 27 March 2017)

Section 276B(1)(a) of Act 51 of 1977 provides that ifa court sentences a convicted offender to imprison-ment for two years or longer, the court may as part ofthis sentence fix a period during which the sentencedoffender may not be released on parole. In terms ofs 276B(1)(b) such period ‘shall be referred to as thenon-parole period, and may not exceed two thirds of

the term of imprisonment imposed or 25 years,whichever is the shorter’. On the origin and purposeof s 276B and its context in the parole system, wherethe executive as opposed to the judiciary should havethe say, see Terblanche A Guide to Sentencing inSouth Africa 3 ed (2016) 259–60, Moses Parole inSouth Africa (2012) 40–3 and the discussion ofs 276B in Commentary, sv General and sv Thejudiciary and the executive. See also the introductorypart in the article ‘Perspectives on parole: Two recentdecisions’ which appears in Section A of the presentedition of CJR.

In S v Ndlovu [2017] ZASCA 26 (unreported, SCAcase no 925/2016, 27 March 2017) the sentencingcourt had fixed a non-parole period in terms ofs 276B (at [5]). On appeal it was common cause thatthe court never informed the appellant and the Stateof its intention to invoke the provisions of s 276B (at[12]). The prosecution and the State were also notinvited by the sentencing court to make submissionsas to whether a non-parole period should be fixed.

The basis for the appeal was that the sentencingcourt had misdirected itself when it fixed a non-parole period without first having invited the partiesto make submissions on the matter (at [7]). Writingfor a unanimous full bench, Mbatha AJA confirmedthat a convicted offender is entitled to address thesentencing court when it considers the imposition ofa non-parole period as part of a sentence (at [11]).See also S v Jimmale 2016 (2) SACR 691 (CC) at[16] and [21]. Indeed, a sentencing court should interms of the decision in S v Stander 2012 (1) SACR537 (SCA) at [22] give the parties an opportunity toaddress it on at least the following: should a non-parole period be ordered? And, if so, how longshould the non-parole period be?

In Ndlovu (supra) it was concluded, in line withseveral other decisions, that a failure to afford theparties an opportunity to address the court was amisdirection by the sentencing court. See further thediscussion of s 276B in Commentary, sv Proceduralmatters and the requirement that the court shouldgive reasons’.

In Ndlovu (supra) the appellant also relied on anotherground of appeal, namely that there were no excep-tional circumstances which could have justified thecourt’s decision to fix a non-parole period (at [7]). Inthis regard it must be pointed out that the require-ment that there should be ‘exceptional circum-stances’ is not stipulated in s 276B, but stems fromcase law in which s 276B has been interpreted andapplied. The Supreme Court of Appeal, for example,


stated in Stander (supra) at [16] that an order interms of s 276B should ‘only be made in exceptionalcircumstances, when there are facts before the sen-tencing court that would continue, after sentence, toresult in a negative outcome for any future decisionabout parole’. See also S v Mhlongo (supra) at[5]–[7]; S v Motloung 2016 (2) SACR 243 (SCA) at[18] and further cases referred to in the discussion ofs 276B in Commentary, sv Case law: Aggravatingfactors and the requirement that there must beexceptional circumstances. The reason why ourcourts insist that there should be exceptional circum-stances before a court can fix a non-parole periodmust also be understood in the context of thefollowing remarks by the Constitutional Court in S vJimmale (supra) at [13]:

‘The order should be made only in exceptionalcircumstances, which can be established byinvestigation of salient facts, legal argumentand sometimes further evidence upon which adecision for non-parole rests. In determining anon-parole period following punishment, acourt in effect makes a prediction on what maywell be inadequate information as regards theprobable behaviour of the accused. Therefore, aneed for caution arises because a proper eviden-tial basis is required.’

In Ndlovu Mbatha AJA, having duly noted theapproach as set out in Jimmale (supra), proceeded topoint out that it was not clear from the judgment ofthe sentencing court what factors were consideredfor purposes of imposing the non-parole period (at[12]). This, in turn, also meant that exceptionalcircumstances were not identified. In this respect thesentencing court had also misdirected itself. Indeedin S v Mhlongo (supra) Mocumie JA—having notedthat the right to a fair trial extends to the sentencingprocess (at [9])—also emphasised the importance ofthe principle that proper reasons must in the interestsof justice be furnished whenever a non-parole periodis determined (at [12], emphasis added):

‘Giving reasons for decisions is a longstandingand salutary practice that serves the interests ofjustice. Furthermore, it helps to show the ratio-nale for the decision. Without reasons for ajudgment on sentence, as is the case in thismatter, in respect of the invocation of s 276B,such lack of reasons is highly prejudicial to theaccused person. Thus the court a quo’s failureto state the rationale for its judgment is avitiating factor.’

In Ndlovu (supra) Mbatha AJA accordingly had noproblem in setting aside ‘the imposition of thenon-parole period’ (at [13]).

In conclusion Mbatha AJA also identified anotherirregularity, namely that the non-parole period fixed bythe sentencing court was in conflict with the provisionsof s 276B(2). This section provides that where anaccused who has been convicted of two or moresentences is sentenced to imprisonment and the courtdirects that prison sentences shall run concurrently, thecourt shall—subject to s 276B(1)(b)—‘fix the non-parole period in respect of the effective period ofimprisonment’. The sentencing court in Ndlovu hadindeed made a concurrency order: the sentence of 20years’ imprisonment for murder was ordered to runconcurrently with the sentence of 12 years’ imprison-ment for robbery. The effective period of imprisonmentwas therefore 20 years. The sentencing court, however,had confined its non-parole period of 13 years to thesentence imposed for murder. This, said MbathaAJAat[12], was another ‘misdirection’. See further S vMthimkulu 2013 (2) SACR 89 (SCA) which is alsodiscussed in the notes on s 276B in Commentary, svThe interpretation of s 276B(2).

s280(2): Sentencing court’s discretion to orderconcurrent running of sentences

S v Zulu [2017] ZASCA 207 (unreported, SCA caseno 226/2016, 21 December 2016) and S v Langa[2017] ZASCA 2, unreported SCA case no 640/2016, 23 February 2017)

Section 280(2) of Act 51 of 1977 provides a sentenc-ing court with a judicial discretion to order thatsentences consisting of imprisonment (or of correc-tional supervision) should run concurrently. See thediscussion of s 280 in Commentary, sv Concurrentsentences. A concurrency order is a means of ensur-ing that the cumulative effect of two or moresentences is kept within acceptable levels. See Ter-blanche A Guide to Sentencing in South Africa 3 ed(2016) 200. In S v Chake 2016 (2) SACR 309 (FB)Murray J pointed out that a disregard for s 280 ‘maylead to inhumane and unfair sentences’ (at [8.6]).

Aspects concerning concurrency orders arose in tworecent Supreme Court of Appeal decisions.

In S v Zulu [2017] ZASCA 207 (unreported, SCAcase no 226/2016, 21 December 2016) the applicantwas granted leave to appeal against sentence to a fullbench of the High Court which had dismissed hispetition for leave to appeal. At [6] Wallis JA pointedout that where various charges against an accused

37Issue 1, 2017

arose out of a ‘single criminal enterprise’, a sentenc-ing court is required to consider concurrency.

In Zulu the regional court had convicted the appel-lant of two counts of robbery. On each count ofrobbery he was—in the absence of substantial andcompelling circumstances justifying a lesser sen-tence—sentenced to 15 years’ imprisonment inaccordance with the minimum sentence provisions inthe Criminal Law Amendment Act 105 of 1997.Sentences imposed for firearm charges were orderedto run concurrently with the two sentences forrobbery. However, the regional court made no con-currency order in respect of the two sentences forrobbery, which meant that these two sentences had tobe served consecutively—amounting to an effective30 years’ imprisonment (at [2]).

In Zulu there was indeed, for purposes of sentencing,a single criminal enterprise. The appellant and hisaccomplice had at gunpoint robbed a man of hismotor car and when this man and the woman withhim had stepped out of the car, she was robbed of herhandbag and its contents. In these circumstances thetwo charges of robbery, which resulted in twoconvictions, did not offend the rule against duplica-tion of convictions. For an analysis of this rule andits guiding principles, see the discussion of s 83 inCommentary, sv The rule against duplication ofconvictions and sv Guiding principles to determinewhether a duplication of convictions has occurred.

Wallis JA pointed out that whilst the appellant hadbeen properly convicted of two counts of robberyarising ‘out of a single criminal enterprise’, the‘problem’ to which this gave rise was that the twoconvictions were treated separately for sentencingpurposes (at [5]). The reasons for requiring an orderof concurrency were set out as follows by Wallis JA(at [6], emphasis added):

‘The result was that the chance fact that. . .[thewoman]. . .was present and had her handbagstolen, was the only reason for an increase insentence from 15 years to 30 years imprison-ment. It should have been plain to the magis-trate that this was irrational and resulted in amanifestly excessive sentence being imposed fora single criminal enterprise. The theft of thehandbag added nothing to the moral culpabilityof. . .[the appellant]. . .and did not justify anysignificant increase in the sentence to beimposed upon him.’

Zulu (supra) should be compared with S v Langa[2017] ZASCA 2 (unreported, SCA case no 640/

2016, 23 February 2017). In Langa the regionalcourt had sentenced the appellant to ten years’imprisonment for murder and four years’ imprison-ment for attempted murder. It was specifically notedby the sentencing court that the sentences were not torun concurrently (at [1]). This meant that the appel-lant had to serve an effective 14 years.

One ground of the appeal against sentence was thatthe regional court should have exercised its s 280discretion by making a concurrency order, since ‘thetwo offences were closely connected in space andtime’ (at [13]).

Writing for a unanimous full bench, Potterill AJAsaid that while it was true that the murder andattempted murder were indeed ‘closely linked intime and space’, the evidence showed that theappellant ‘randomly’ shot ‘at patrons in a shebeen inan attempt to kill the deceased execution style’ (at[14]). ‘The appellant’, continued Potterill JA at [14],‘thus intended not only to shoot the deceased, butforesaw that by randomly shooting in the shebeen,he could kill other patrons.’

The full bench concluded that on account of theabove facts, the appellant was ‘not punished twicefor the same actions’ and that the regional court hadnot misdirected itself ‘in not ordering that the sen-tences run concurrently’ (at [14]). In the court’s viewthe cumulative effect of 14 years’ imprisonment formurder and attempted murder was neither disturb-ingly inappropriate nor disproportionate to theoffences. The regional court, it was concluded, hadexercised its discretion judicially. There was nomisdirection and there were no other grounds forinterfering with the sentences on appeal.

It is submitted that there is no conflict between Langaand Zulu (supra). The essential principle is that anotherwise acceptable effective term of imprisonment isnot rendered unacceptable simply because the offencesconcerned were ‘closely linked in space and time’ or‘arose from ‘a single criminal enterprise’.

(iii) Forfeiture and confiscation

s 20: Sections 48(1) and 50(1) of thePrevention of Organised Crime Act 121of 1998: When is residential property an‘instrumentality of an offence’ for thepurpose of a forfeiture order?Brooks & another v National Director of PublicProsecutions 2017 (1) SACR 701 (SCA)


The appellants, who were husband and wife, werethe registered owners of residential immovable prop-erty which had been declared forfeit to the State interms of s 50(1) of POCA on the basis that it was an‘instrumentality of an offence’, illegal diamond deal-ing, as contemplated in POCA. Of the 19 transac-tions in question, ten took place at the property in thespace of less than one year. The second appellant, thewife, had raised the so-called ‘innocent owner’defence to the grant of the forfeiture order relating tothe property, which was owned jointly by thespouses, who were married in community of prop-erty. It was accepted that she had no knowledge ofthe illegal diamond dealing on the residential prop-erty, which took place while she was at work. Thecourt a quo therefore excluded her interest in theproperty from the forfeiture order, and ordered thecurator bonis to pay her one half of the net proceedsfrom the sale of the property.

An appeal against this judgment drew a dividedresponse in the Supreme Court of Appeal. For theminority, Schippers JA (Mocumie JA concurring)held that the property had correctly been held to havebeen an instrumentality of the offence. In the view ofthe minority judges the relationship between the useof the property and the commission of the offenceswas ‘neither tenuous nor remote’ (at [36]). Theinvolvement of the property was ‘not merely inci-dental to the commission of the offences: it was putto use in a positive sense and was a means throughwhich the crimes were committed’. The propertywas the base for a significant, organised and well-funded illicit diamond dealing business, and, but fortheir arrest, the first appellant (the husband) and thedealers would clearly have continued to use theproperty as their base. POCA defines ‘instrumental-ity of an offence’ as meaning, inter alia, ‘anyproperty which is concerned in the commission orsuspected commission of an offence’. The cases havemade it clear (see Commentary on s 20 sv Seizureand forfeitures under the Prevention of OrganisedCrime Act 121 of 1998), however, that the definitionmust be restrictively construed and that the propertyhas to ‘facilitate or make possible the commission ofthe offence in a real and substantial way’ (at [19]).

Schippers JA found that the evidence in support ofinstrumentality was compelling and that the NDPPhad discharged the onus of establishing that fact on abalance of probabilities. He then turned to the issueof proportionality, which he described as an inquiry‘aimed at balancing the constitutional imperative oflaw enforcement and combating crime and the seri-

ousness of the offence, against the right not to bearbitrarily deprived of the property’ (at [39]). Thebalance, in his view, fell in favour of forfeiture,which in this case would serve the ‘broader societalpurpose of: (a) deterring persons from using propertyto commit crime . . . (b) eliminating or incapacitatingsome of the means by which crime may be commit-ted; and (c) advancing the ends of justice by depriv-ing those involved in crimes of the property con-cerned’ (at [43]): see NDPP v RO Cook Properties(Pty) Ltd; NDPP v 37 Gillespie Street Durban (Pty)Ltd & another; NDPP v Seevnarayan 2004 (2)SACR 208 (SCA), discussed in Commentary.

The majority, per Ponnan JA (with Willis and ZondiJJA concurring), disagreed. Ponnan JA assumed,without deciding, that the use to which the propertywas put did render it an instrumentality. But he didso only because this was not crucial in view of theconclusion he reached on proportionality, and evenadded that, because the commission of the crimes onthe property was ‘purely incidental to their commis-sion’ (at [61]), it was ‘difficult . . . to see howproperty bearing no reasonably direct connection tothe offences other than serving as the location for thetransactions could constitute an instrumentality ofthose offences’.

The Bill of Rights, said Ponnan JA (at [62]), pro-vides in s 25(1) that ‘no law may permit arbitrarydeprivation of property’. The Constitutional Courthas stressed that there must be sufficient reason forthe deprivation of assets by forfeiture and that thenature of the relationship between the means andends to satisfy the rationality requirement of s 25(1)depends both on the nature of the affected propertyand the extent of the deprivation (see Mkontwana vNelson Mandela Metropolitan Municipality & others2005 (1) SA 530 (CC) at [35]). Thus, althoughs 50(1) of POCA is couched in peremptory terms byproviding that a court ‘shall’ make a forfeiture orderif it finds on the civil standard of proof that theproperty in question is an instrumentality of anoffence, our courts have consistently interpreted‘shall’ to mean ‘may’ (see Mohunram & another vNDPP & another (Law Review Project as AmicusCuriae) 2007 (4) SA 222 (CC) at [121]). Accord-ingly, said Ponnan JA (at [63]), the courts havedeveloped the notion of proportionality, not as astatutory requirement, but as a ‘constitutionalimperative’ to curb the excesses of civil forfeiture.Its proper application ‘weighs the forfeiture and, inparticular, its effect on the owner concerned, on the

39Issue 1, 2017

one hand, against the purposes the forfeiture serves,on the other’ (see, too, Mohunram at [58]).

The majority examined all the circumstances of thecase and concluded (at [82]) that the forfeiture ordermade under s 50(1) of POCA was disproportionate.What factors are weighed in the inquiry into propor-tionality? According to the Constitutional Court inProphet v NDPP 2006 (2) SACR 525 (CC) at [58],[63] and [68], the nature and gravity of the offence inquestion, the extent to which ordinary criminal lawmeasures are effective in dealing with it, its publicimpact and potential for widespread social harm anddisruption, are all to be considered. Ponnan JAobserved that, according to Mohunram’s case, itwould be wrong for POCA to be used in a mannerwhich blurred the distinction between the purposesand the methods of criminal enforcement, on the onehand, and those of civil law, on the other. There wasthus ‘no justification for resorting to the remedy ofcivil forfeiture under POCA as a substitute for theeffective and resolute enforcement of ordinary crimi-nal remedies’ (at [64]). The forfeiture had to advancethe purposes of POCA itself; if not, the forfeiture,being the means, would be misaligned with thepredominant ends pursued by POCA.

Was, then, the instrumentality of the offence suffi-ciently connected to the main purpose of POCA?The following considerations led the majority toanswer in the negative: First, the appellant andothers were, at the time, being prosecuted for contra-vening the Diamonds Act 56 of 1986, which set outfairly harsh penalties for its contravention and madespecific provision for forfeiture, so that forfeiture interms of POCA might be ‘doubly punitive’ (at [65]).Secondly, the forfeiture had to be ‘rationally relatedto its purposes’. These purposes, set out in the CookProperties case at [(18)], included: (a) removingincentives for crime; (b) deterring persons fromusing or allowing their property to be used in crime;(c) eliminating or incapacitating some of the meansby which crime may be committed; and (d) advanc-ing the ends of justice by depriving those involved incrime of the property concerned. Although thesemotives were laudable, said Ponnan JA, there was‘no escaping the draconian nature of the forfeiture’,particularly in the case of the second appellant (thewife) who had ‘committed no wrong of any sort,whether intentional or negligent, active or acquies-cent’ (at [66], quoting from the Cook Properties caseat [26]).

The third consideration involved the interests of thetwo minor children of the two appellants. In terms of

s 28(2) of the Constitution, the best interests of thechildren are of ‘paramount importance’ in all mattersconcerning them. Their interests had to be given‘specific and separate consideration, in addition tothe attention they might get in the proportionalityanalysis’ (at [72]), so that a court might need toappoint a curator to conduct an independent assess-ment of their interests. No investigation had beenconducted in this case and the interests of thechildren were not even mentioned in the proportion-ality analysis of the High Court. The onus ofestablishing proportionality rested on the NDPP,who made no attempt to adduce any evidence that, inrespect of the wife and the two minor children inparticular, the forfeiture would not be constitution-ally excessive (at [78]).

The closing remarks of Ponnan JA at [81] encapsu-late the sentiments of the majority as to why, inessence, the forfeiture order was disproportionate:

‘There can be no serious claim that the confis-cation order sought by the NDPP is not puni-tive. Fundamental fairness prohibits the punish-ment of innocent people. I regard it asaxiomatic that persons should not be punishedwhen they have done no wrong. The uniquefacts of this case demonstrate that Ms Brooksand her children are entitled to the protection ofthat rule. ‘‘Improperly used, forfeiture couldbecome more like a roulette wheel employed toraise revenue from innocent but hapless ownerswhose property is unforeseeably misused, or atool to punish those who associate with crimi-nals, than a component of a system of justice’’’[quote from Bennis v Michigan 516 US 442(1996) at 456].

(iv) Appeal and review

Appeal: Requirement that court ofappeal gives notice of possible increaseof sentence

S v Joubert 2017 (1) SACR 497 (SCA)

The appellant in the above matter was convicted inthe regional court on 20 counts of fraud and sen-tenced to seven years’ imprisonment, wholly sus-pended on certain conditions for a period of fiveyears. Aggrieved by his convictions but not hissentence, the appellant petitioned the High Court forleave to appeal against his convictions only. How-ever, the High Court granted leave—‘erroneously it


would appear’ (at [1])—to the appellant to appealagainst sentence as well as conviction.

Riding on the back of the High Court’s error, theState ‘grasped the opportunity’ to give notice to theappellant and his attorneys that the State intended toseek an increase of sentence on appeal (at [2]). In hisgrounds of opposition, the appellant referred to theState’s failure ‘to follow the legal prescripts andrequirements’ specified in the Criminal ProcedureAct 51 of 1977. The State never formally soughtleave to appeal against sentence, as required bys 310A of Act 51 of 1977. The High Court hearingthe appeal was aware of this ‘procedural shortcom-ing’ (at [9]), but never notified the appellant that thecourt itself was indeed considering an increase onappeal.

Having heard the appellant’s appeal and havingdismissed the appeal against conviction, the HighCourt set aside the sentence imposed by the regionalcourt, replacing it with a sentence of seven years’imprisonment, with four years suspended on certainconditions for a period of five years (at [1]).

The appellant appealed—with leave of the HighCourt—to the Supreme Court of Appeal against hisincreased sentence, claiming that his right to a fairtrial had been infringed by the High Court’s failureto give him proper prior notice of its intention toconsider increasing his sentence. Indeed, in theabsence of a prior notice from the court, counsel forthe appellant never explained to the appellant thatthe High Court ‘was minded to mero motu increasethe sentence’ (at [6]). It was only in the course ofargument before the High Court that she becameaware of the High Court’s concern over the appropri-ateness of the sentence imposed by the regionalcourt.

Writing for a unanimous full bench, Majiedt JA had‘little doubt’ that the Constitutional Court decision inS v Bogaards 2013 (1) SACR 1 (CC) applied‘squarely to this case’ (at [5]), despite the fact thatthe High Court appeal was concluded beforeBogaards was decided (at [4]). For a summary of theprinciples established in Bogaards, see the discus-sion of s 309 in Commentary, sv Increase in sen-tence. Relevant to the appeal in Joubert were thefollowing rules and principles as formulated inBogaards (at [79]):

‘If the court forms a prima facie view before thehearing, that it is considering an increase insentence, for instance, while reading the record,it should put the accused person on notice prior

to the hearing. If the court forms this opinionduring the hearing, then it must formally informthe accused person that it is considering anincrease and give the accused person sufficienttime, subsequent to the hearing, to make writtensubmissions on this issue. Finally, even if thecourt is contemplating an increase after thehearing it must formally request the parties tomake submissions on this point before makingits final decision.’

It is clear that the High Court in Joubert had actedwithout regard to the rules in Bogaards. However,counsel for the State submitted that whilst its appli-cation to the High Court for an increase in sentencewas irregular, the High Court’s failure to complywith the rules in Bogaards was cured by the fact thatthe State’s notice and application, even thoughirregular, had indeed alerted the appellant to thepossible increase in sentence. Majiedt JA rejectedthis argument on the basis that it was ‘inconceivablethat one fatal irregularity can be called in to cureanother irregularity’ (at [8]).

Counsel for the State also advanced another argu-ment in an attempt to place Joubert outside the ambitof Bogaards. It was argued that there had been noprejudice to the appellant since he was in factafforded an opportunity to make submissions onsentence—in his opposing affidavit and through hiscounsel’s oral submissions in the course of argumentwhen the High Court raised the matter. Having notedthat not every irregularity is necessarily a fatalirregularity, Majiedt JA proceeded to point out thatan accused person is required to demonstrate that theirregularity had caused material prejudice to him,‘such that it. . .led to a failure of justice and aninfringement of the right to a fair trial’ (at [8]). Andfairness, continued Majiedt JA, had to be determined‘upon the particular facts of the case and. . .is con-text-specific’.

Turning to the facts in Joubert, it was concluded thatthe appellant had indeed been materially prejudiced.The prejudice present went beyond a mere inad-equate opportunity to prepare. In preparing theresponse to the State’s notice, the appellant and hislegal representatives obviously focused on the proce-dural defects of the notice and not the merits ordemerits of an increase of the regional court’ssentence; and the possibility of an increase in sen-tence had not been foreshadowed in a prior noticefrom the High Court, as was required (at [9]).

At [10] Majiedt JA stated that there was ‘a furtherand even more compelling reason’ in support of the

41Issue 1, 2017

conclusion that the appellant had been materiallyprejudiced:

‘An accused person who has been given noticeby an appellate court that it intends to increasethe sentence imposed by the trial court has theoption of withdrawing the appeal, with theleave of the appellate court. This practice,together with the requirement of prior notice toan accused person by the appellate court bal-ances the appellant’s right to a fair trial and thecourt’s duty to ensure that the sentence isappropriate and where necessary, to increase aninappropriate sentence. In the present instance,the appellant had not been afforded the opportu-nity to consider such a course of action. Ascounsel for the appellant correctly contended, anotice of intention to increase sentence is a veryweighty consideration, emanating as it doesfrom the judges empowered to increase thesentence. The prejudice and its materiality areself-evident.’

The sentence imposed by the High Court on appealwas accordingly set aside, and the matter was remit-ted to the High Court for consideration of the appealagainst sentence only, in accordance with the guide-lines outlined by the Constitutional Court inBogaards (supra) at [79].

Appeal: Disposal of an appeal in termsof s 19(a) of the Superior Courts Act 10of 2013

S v Zulu [2017] ZASCA 207 (unreported, SCA caseno 226/2016, 21 December 2016)

In the above appeal the Supreme Court of Appeal,having granted the appellant special leave to appealto it, was required to determine whether the appel-lant had reasonable prospects of succeeding in hisappeal against sentence if he were permitted topursue such an appeal in the High Court (at [3]). Inaddressing this matter, the Supreme Court of Appealhad the relatively rare opportunity to exercise thepower it now has in terms of s 19(a) of the SuperiorCourts Act 10 of 2013. This section determines thatthe Supreme Court of Appeal—or any Division ofthe High Court exercising appeal jurisdiction—‘may,in addition to any power as may specifically beprovided for in any other law. . .dispose of an appealwithout the hearing of oral argument. . .’.

In Zulu at [4] Wallis JA explained that ‘[a] reading ofthe record and the heads of argument made it clear toall the members of the court allocated to sit in theappeal that the appeal had to succeed’. It wasaccordingly decided that the appeal—which was setdown for hearing on 16 February 2017—could bedisposed of without the need to hear oral argument.Judgment was delivered on 21 December 2016. Itwas, said Wallis JA, ‘appropriate’ for the court toexercise its s 19(a) power ‘in the interests of theexpeditious disposal of the appeal’. This also meantthat there was ‘an appropriate use of judicialresources’ and the saving of costs ‘that would other-wise have been incurred by the public purse’ (at [4]).A further advantage was that the process of settingdown the appellant’s appeal for hearing in the HighCourt was expedited. The decision in Zulu makes alot of sense. But as a precedent it has and must havea very limited reach.

It is submitted that the Supreme Court of Appeal’sdecision in Zulu (supra) to dispose of the appealwithout the hearing of oral argument should beunderstood not only in the context of the fact that thejudges concerned were all in agreement ‘that theappeal had to succeed’ (at [4]). It must be assumedthat the judges in Zulu had, by implication, alsodecided that s 19(a) of Act 10 of 2013 could beinvoked because they were convinced that nothingadvanced in oral argument would or could havepersuaded them to change their position. This type ofdecision, it is suggested, is not difficult to take wherethe issue happens to be elementary and where thesolution or answer is of such a nature that theappellant’s rights are in no way curtailed orinfringed, and the respondent is not deprived ofarguing the merits of the matter elsewhere. It shouldbe emphasised that in Zulu the issue was whether theappellant had reasonable prospects of success if hewere allowed to have his appeal against sentenceheard in the High Court. Matters of greater complex-ity and finality require a different approach. In theseinstances no court of appeal—however firm itsviewpoint might be after having read the record andheads of argument—should deprive itself of thebenefit of the salutary practice that a court, if it holdsa view adverse to a particular party, should put suchview to the party concerned in the course of oralargument.

Clearly, the decision in Zulu should not be inter-preted as a case which seeks to reduce oral argumentto an unnecessary formality.


Table of Cases

Bennis v Michigan 516 US 442 (1996) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Bernstein & others v Bester & others NNO 1996 (2) SA 751 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Brooks & another v National Director of Public Prosecutions 2017 (1) SACR 701 (SCA) . . . . . . . . . . 38Carter v Canada (Attorney-General) 2015 SCC 5;l [2015] 1 SCR 331 . . . . . . . . . . . . . . . . . . . . . . . . . . 23Chamberlain v The Queen [No 2] (1984) 153 CLR 521 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6Clarke v Hurst NO & others 1992 (4) SA 630 (D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21, 22De Vos NO & others v Minister of Justice and Constitutional Development & others 2015 (2) SACR

217 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Democratic Alliance v President of the Republic of South Africa & others 2013 (1) SA 248 (CC) . . . 12Estate De Wet v De Wet 1924 CPD 341 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Ex parte Die Minister van Justisie: In re S v Grotjohn 1970 (2) SA 355 (A) . . . . . . . . . . . . . . . . . . . . . 22International Tobacco Co (SA) Ltd v United Tobacco Cos (South) Ltd 1953 (3) SA 343 (W) . . . . . . . 14J v National Director of Public Prosecutions & others 2014 (2) SACR 1 (CC) . . . . . . . . . . . . . . . . . . . 17Kelly v Minister of Correctional Services & others 2016 (2) SACR 351 (GJ) . . . . . . . . . . . . . . . . . . . . 11Lornadawn Investments (Pty) Ltd v Minister van Landbou 1977 (3) SA 618 (T) . . . . . . . . . . . . . . . . . 13Mdani v Allianz Insurance Ltd 1991 (1) SA 184 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Minister of Justice & others v Estate Stransham-Ford 2017 (3) SA 152 (SCA) . . . . . . . . . . . . . . . . . . . 21Minister of Safety and Security & another v Carmichele 2004 (3) SA 305 (SCA) . . . . . . . . . . . . . . . . . 9Mkontwana v Nelson Mandela Metropolitan Municipality & others 2005 (1) SA 530 (CC) . . . . . . . . . 39Mohunram & another v NDPP & another (Law Review Project as Amicus Curiae) 2007 (4) SA 222

(CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Moise v Greater Germiston Transitional Local Council: Minister of Justice and Constitutional

Development Intervening 2001 (4) SA 491 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Naidu v Minister of Correctional Services [2017] 2 All SA 651 (WCC) . . . . . . . . . . . . . . . . . . . . . . . . 9National Society for the Prevention of Cruelty to Animals v Minister of Constitutional Development &

another (unreported, GP case no 29677/2013, 8 October 2014) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26National Society for the Prevention of Cruelty to Animals v Minister of Justice and Constitutional

Development & another 2016 (1) SACR 308 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26National Society for the Prevention of Cruelty to Animals v Minister of Justice and Constitutional

Development & others 2017 (1) SACR 284 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25NDPP v RO Cook Properties (Pty) Ltd; NDPP v 37 Gillespie Street Durban (Pty) Ltd & another; NDPP

v Seevnarayan 2004 (2) SACR 208 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Pentree Limited v Nelson Mandela Bay Municipality [2017] 2 All SA 260 (ECP) . . . . . . . . . . . . . . . . 12Prince v Minister of Justice and Constitutional Development & others; Acton & others v National

Director of Public Prosecutions & others (unreported, WCC case no 8760/2013, 31 March 2017) . . 23Prince v The President, Cape Law Society & others 2002 (2) SA 794 (CC) . . . . . . . . . . . . . . . . . . . . . 24Prophet v NDPP 2006 (2) SACR 525 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40R v Baldree 2013 SCC 35 [2013] 2 SCR 520 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15R v Blom 1939 AD 188 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3, 4R v Dladla & others 1962 (1) SA 307 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31R v Hepworth 1928 AD 265 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30R v Malmo-Levine [2003] SCC 74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25R v Manda 1951 (3) SA 158 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6R v Miller 1939 AD 106 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13R v Mlambo 1957 (4) SA 727 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8R v Mtembu 1950 (1) SA 670 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5R v Peverett 1940 AD 213 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21R v Sibanda & others 1965 (4) SA 241 (SRA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6R v Villaroman 2016 SCC 33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Re Conroy 486A 2d 1209 (NJSC 1985) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ?Rex v de Villiers (1944, A.D. 493 at p. 508) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6S v Arendse [2015] ZASCA 131 (unreported, SCA case no 089/15, 28 September 2015) . . . . . . . . . . . 31

43

S v Basson 2007 (3) SA 582 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34S v Bogaards 2013 (1) SACR 1 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41S v Botha 2006 (2) SACR 110 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8S v Chabedi 2005 (1) SACR 415 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31S v Chake 2016 (2) SACR 309 (FB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37S v Cooper 1976 (2) SA 875 (T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8S v De Bellocq 1975 (3) SA 538 (T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22S v Du Plessis 2012 (2) SACR 247 (GSJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29S v Essack & another 1974 (1) SA 1 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8S v Gordon 1962 (4) SA 727 (N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22S v Hartmann 1975 (3) SA 532 (C) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22S v Hewitt 2017 (1) SACR 309 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28S v IJ 2013 (2) SACR 599 (WCC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17S v Jimmale 2016 (2) SACR 691 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8, 36, 37S v Joubert 2017 (1) SACR 497 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40S v Kobese (unreported, ECG case no CA&R 34/2017, 23 February 2017) . . . . . . . . . . . . . . . . . . . . . . 27S v Kolea 2013 (1) SACR 409 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35S v Kotze [2017] ZASCA 27 (unreported, SCA case no 776/16, 27 March 2017) . . . . . . . . . . . . . . . . . 31S v Langa [2017] ZASCA 2 (unreported SCA case no 640/2016, 23 February 2017) . . . . . . . . . . . . . . 37, 38S v Le Grange & others 2009 (1) SACR 125 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28, 34S v Litako & others 2014 (2) SACR 431 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33S v Machaba 2016 (1) SACR 1 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31S v Madolwana (unreported, ECG case no CA&R 436/12, 19 June 2013) . . . . . . . . . . . . . . . . . . . . . . . 9S v Madondo (unreported, KZP case no AR350/2016, 8 December 2016) . . . . . . . . . . . . . . . . . . . . . . 31, 32S v Mahlalela [2016] ZASCA 181 (unreported, SCA case no 396/16, 28 November 2016) . . . . . . . . . 4, 5S v Majali (unreported, GSJ case no 41210/2010, 19 July 2011) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27S v Marengo 1991 (2) SACR 43 (W) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22S v Matlala 2003 (1) SACR 80 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9S v Mgibelo 2013 (2) SACR 559 (GSJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28S v Mhlakaza & another 1997 (1) SACR 515 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8S v Mhlongo 2016 (2) SACR 611 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28, 37S v Morgan & others 1993 (2) SACR 134 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6S v Motloung 2016 (2) SACR 243 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9, 37S v Mthetwa 1972 (3) SA 766 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31S v Mthimkulu 2013 (2) SACR 89 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37S v Mudau [2017] ZASCA 34 (unreported, SCA case no 1148/2016, 29 March 2017) . . . . . . . . . . . . . 33S v Musiker 2013 (1) SACR 517 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30S v Ndlovu [2017] ZASCA 26 (unreported, SCA case no 925/2016, 27 March 2017) . . . . . . . . . . . . . . 9, 36S v Nkuna 2013 (2) SACR 541 (GNP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8S v Nkunkuma & others 2014 (2) SACR 168 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28S v Ntsele 1998 (2) SACR 178 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6S v Pedro 2015 (1) SACR 42 (WCC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19S v Quayiso 2017 (1) SACR 470 (ECB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28S v Rall [1982 (1) SA 828 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29, 30S v Robinson & others 1968 (1) SA 666 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22S v Schoombee & another 2017 (5) BCLR 572 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30S v Shepard & others 1967 (4) SA 170 (W) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6S v Sibeko en ’n ander 1990 (1) SACR 206 (T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27S v Sibelelwana [2012] ZAWCHC 150 (unreported, WCC case no A401/2011, 3 August 2012) . . . . . 30S v Sithole & others 1999 (1) SACR 585 (W) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29S v Smith en andere 1978 (3) SA 749 (A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6S v SMM 2013 (2) SACR 292 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28S v Stander 2012 (1) SACR 537 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36


S v Tshoga 2017 (1) SACR 420 (SCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34S v Van Loggenberg 2012 (1) SACR 462 (GSJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9S v Zulu [2017] ZASCA 207 (unreported, SCA case no 226/2016, 21 December 2016) . . . . . . . . . . . 37, 42S v Zuma 1995 (2) SA 642 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Shepherd v The Queen (1990) 170 CLR 573 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6Valente v The Queen [1985] 2 SCR 673 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Van Rooyen & others v The State & others (General Council of the Bar of South Africa intervening)

2002 (5) SA 246 (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Wright v Doe d Tatham (1837) 7 Ad & El 313 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

45Issue 1, 2017

Documents

A bi-annual update complementing the Commentary on … · A bi-annual update complementing the Commentary on the Criminal Procedure Act ... Is it time to jettison the rules in R v