A bi-annual update complementing the

Commentary on the Criminal Procedure Act


NO 1 OF 2018


Andrew Paizes, Author (Editor)

Steph van der Merwe (Author)


For a printable .pdf version of this Review click here.



The cases discussed in this edition raise interesting and important questions, some of a fundamental constitutional nature. The issues considered in relation to substantive criminal law are:

  • the constitutional status of the common-law defence, raised when a parent is charged with assaulting his or her child, of ‘moderate chastisement’;
  • the importance of distinguishing between liability as an accessory (also called an ‘accomplice’) and liability by reason of the doctrine of common purpose; and
  • whether money in its incorporeal form, in particular by means of electronic transfer, can constitute ‘property capable of being stolen’ for the purpose of attracting liability for the crimes of theft and robbery.

An interesting and important evidential issue is addressed: when can an unrepresented accused be said to have ‘agreed’ to the admission of hearsay evidence tendered against him or her, and how do we safeguard the constitutional rights of an unrepresented accused in such circumstances?

An issue which raises both evidential and procedural challenges is also considered: in what circumstances is an interested party entitled to have access to the record of evidence given at an investigation conducted in terms of s 28(1) of the National Prosecuting Authority Act 32 of 1998, and how is the discretion to disclose such evidence, which vests in the Director of National Prosecutions in terms of s 46(c) of that Act, to be exercised?

The following procedural questions (pre-sentence) are examined:

  • the liability of the police following an unlawful arrest for the further detention of the arrestee after he or she has been remanded to custody by a court;

  • the role of public opinion and public demand in refusing bail;

  • whether a court has a discretion to stay the execution of a warrant of arrest in respect of an accused who fails to appear after bail has been granted;

  • the grounds upon which bail may legitimately be cancelled;

  • the power of a court to amend the charge;

  • the judicial review of a decision by the NDPP to withdraw changes on the grounds that the NDPP’s decision did not comply with the requirements of legality and rationality;

  • the importance of separating notionally the inquiries into whether a witness is competent to testify at all and whether he or she is competent to take the oath or affirmation; and

  • the need to respect the constitutional rights of an accused at a summary inquiry under s 170 of the Criminal Procedure Act where he or she fails to appear after an adjournment; and the unconstitutional status of a provision in s 170(2) that requires the accused to satisfy the court that the failure was ‘not due to fault on his part’.

In respect of sentencing, there is a discussion of procedural and evidential issues concerning ‘planned or premeditated’ murder as referred to in Part I of Schedule 2 of the Criminal Law Amendment Act 105 of 1997, as well as the meaning of the term ‘planned and premeditated’ in that Schedule. There is also a note on the need to prove the value of drugs, where an accused is charged with dealing in drugs, as set out in Schedule 2 of that Act.

Andrew Paizes





No legislation of note


(a) Criminal Law

Accessory (or accomplice) liability and liability by reason of common purpose
Assault: Defence of parent’s right of moderate chastisement of a child unconstitutional
Robbery: Can the forced transfer of money by electronic means be the subject of the crime of robbery?

(b) Criminal Procedure and Evidence

i. Pre-sentence

Section 41(6) of the National Prosecuting Authority Act 32 of 1998: The discretion of the National  Director of Public Prosecutions

Section 40(1)(b): Unlawful arrest: Is the Minister of Police liable, following an unlawful arrest, for the further detention of the arrested person after he or she has been remanded to custody by a court?

Section 60(4)(e), (8A) and (9): Bail and the public demand that bail be refused

Section 67(1): Bail and failure to appear: Validity of court order staying execution of warrant

Section 68: The irregular cancellation of bail

Section 86: Court’s power to amend the charge

Judicial review of a decision by the National Director of Public Prosecutions to withdraw charges: Legality and rationality

Sections 164, 192 and 194: Separating the inquiries into competence to testify and competence to take the oath or affirmation

Section 170: Need to respect constitutional rights of accused at summary inquiry where accused fails to appear after an adjournment; unconstitutional reverse onus in s 170(2)

Section 216: Hearsay evidence: s 3(1)(a) of the Law of Evidence Amendment Act 1988: When can an unrepresented accused be said to have ‘agreed’ to the admission of hearsay evidence tendered against him or her?

ii. Sentencing

The meaning of ‘planned or premeditated’ in s 51(1) read with Part I of Schedule 2 of the Criminal Law Amendment Act 105 of 1997

Procedural and evidential issues concerning ‘planned or premeditated’ murder as referred to in Part I of Schedule 2 of the Criminal Law Amendment Act 105 of 1997

Dealing in drugs and proof of value for purposes of Part II of Schedule 2 of the Criminal Law Amendment Act 105 of 1997



No legislation of note

 (a) Criminal Law

Accessory (or accomplice) liability and liability by reason of common purpose

S v Phetoe 2018 (1) SACR 593 (SCA)

It is important to keep apart notionally the discrete principles underlying, on the one hand, criminal liability as an ‘accessory’ (also known as an ‘accomplice’) and, on the other, liability by reason of the doctrine of common purpose. The doctrine of common purpose operates to attribute to one party (the ‘remote’ party) the conduct performed by another party (the ‘immediate’ party) in one of two distinct situations. One is where there is prior agreement (a ‘mandate’) between the parties in terms of which they agree, expressly or impliedly, that the conduct will be performed by the immediate party in order to give effect to their common purpose. The other is where the remote party witnesses the performance of the conduct in question by the immediate party and, intending to make common cause with the perpetrator or perpetrators, performs some act of his own by which he associates himself with the commission of that act.

Either way, the conduct of the party performing the act in question becomes, in law, the act of the remote party himself. The result, if the other elements of the crime (such as fault and the presence of a causal link between the imputed act and the prohibited consequence (in the case of murder, the death of the victim)) are present, is that the remote party will have satisfied all the definitional elements of the crime in question. He will thus be liable as a principal offender.

The position is different in the case of an ‘accessory’ (or an ‘accomplice’, as such a person is commonly called).

The Constitutional Court in Minister of Justice and Constitutional Development & another v Masingili & another 2014 (1) SACR 437 (CC) at [21], relying on the leading case of S v Williams en ’n ander 1980 (1) SA 60 (A) at 63B, gave us this pithy definition of an ‘accomplice’ (per Van der Westhuizen J, emphasis added):

‘An accomplice is someone whose actions do not satisfy all the requirements for criminal liability in the definition of an offence, but who nonetheless intentionally furthers the commission of a crime by someone else who does comply with all the requirements (the perpetrator).’

The liability of the ‘accomplice’ is thus contingent upon the perpetrator’s liability. It is ‘accessory’ to the perpetrator’s liability, so that such a person is not liable as a ‘principal’ offender.

As to the meaning of the word ‘furthers’, used by the courts in both Williams and Masingili, it is clear from the judgment of Joubert JA in Williams that ‘there must be a causal connection between the accomplice’s assistance and the commission of the crime by the perpetrator’ (at 63 in Williams). Thus, as some academic commentators have pointed out (see, for instance, RC Whiting (1980) 97 SALJ 199), ‘further’ should be read as meaning ‘causally further’. Conduct which ‘furthers’ the commission of the crime by the principal offender may, as Snyman points out (see CR Snyman Criminal Law 6 ed (2014) at 266), include ‘any conduct whereby a person facilitates, assists or encourages the commission of a crime, gives advice concerning its commission, orders its commission or makes it possible for another to commit it’. But it remains subject to the causal test: the conduct must further causally the commission of the offence by the perpetrator. This means, at the very least, that the assistance rendered by the accomplice/accessory must be a factual cause of the commission of the crime by the principal in the sense that such commission would, but for the participation of the accomplice, not have taken place (the so-called ‘sine qua non’ test).

The question of accomplice liability arose in S v Phetoe in connection with an alleged rape. The evidence in the case did not, however, disclose any conduct performed by the appellant by way of ‘facilitation, assistance or encouragement’. To convict on the basis of mere presence at the scene, said Mocumie JA (Leach JA and Plasket AJA concurring), would be ‘to subvert the principles of participation and liability as an accomplice in criminal law’ (at [15]). The fact that the appellant laughed after being asked why they were ‘doing such a thing’ might, said the court, be conduct that showed his approval of what was happening, but that was not enough to establish his liability as an accomplice (see S v Nooroordien en andere 1998 (2) SACR 510 (NC) at 524f–g). Mocumie JA distinguished the facts in S v Kock en ’n ander 1988 (1) SA 37 (A), where the appellant actually stood guard with a panga while his co-accused was raping the complainant. Being present, laughing and doing nothing to prevent the rape was certainly, at least, ‘morally reprehensible’ (at [16]), but it was not enough to justify a conviction as an accomplice to rape.

In respect of other charges against the appellant (including housebreaking with intent to rob, robbery with aggravating circumstances, common assault, assault with intent to do grievous bodily harm and malicious injury to property), the appellant had been convicted by the trial court even though he had not been identified as being present at the scenes, on the strength of the doctrine of common purpose. The trial court concluded that prior agreement had been reached. It did so by means of inferential reasoning: ‘because so many offences were committed by so many people at so many places, those who were identified must have agreed beforehand to the rampage and everything that it entailed’ (at [18]). This reasoning was, said Mocumie JA, fallacious, since it was not, in the circumstances, the only reasonable inference to be drawn. In respect of the appellant, it could not be said that, because he had been seen at the scene of one of the crimes, he was necessarily a party to a prior agreement or that he was present at all of the other scenes.

Mocumie JA then turned to the alternative leg of the common purpose doctrine—that relating to active association—but found that the necessary conditions for liability in this regard (as set out in S v Mgedezi & others 1988 (1) SA 687 (A) at 705–6) had not been satisfied. There was no evidence to place him at the scenes of the violence; it had not been shown that he was aware of the violence, or that he had manifested his sharing of a common purpose with the perpetrators. Furthermore, the requisite mens rea had not been proved beyond a reasonable doubt. 

The court’s remarks on the criminal standard of proof are worth repeating:

‘When a court finds that the guilt of an accused has not been proved beyond reasonable doubt, that accused is entitled to an acquittal, even if there may be suspicions that he or she was, indeed, the perpetrator of the crime in question. That is an inevitable consequence of living in a society in which the freedom and the dignity of the individual are properly protected and are respected. The inverse—convictions based on suspicion or speculation—is the hallmark of a tyrannical system of law. South Africans have bitter experience of such a system and where it leads to.’

Assault: Defence of parent’s right of moderate chastisement of a child unconstitutional

S v YG 2018 (1) SACR 64 (GJ)

The question before the court was whether the common-law defence of reasonable chastisement in respect of children was compatible with the Bill of Rights. The common law recognises a specific defence to a charge of assault for parents who use the application of physical force to discipline their children, provided it falls within the bounds of ‘moderate’ or ‘reasonable’ chastisement. This defence finds its origins in the common-law rights and duties of parents, described by Burchell and Hunt South African Criminal Law and Procedure 3 ed at 117 as emanating from their ‘uniquely independent authority in rearing children’, which ‘meant that the State did not interfere in the exercise of the rights, duties and responsibilities of parents in rearing children’.

This right, however, found itself, in the post-constitutional era, in potential conflict with a number of fundamental rights. As far as children were concerned, these included: the right to human dignity (s 10); the right to equal protection under the law (s 9(3)); the right to be free from all forms of violence from either public or private sources (s 12(1)(c)); the right not to be treated or punished in a cruel, inhuman or degrading way (s 12(1)(e)); the right of children to be protected from maltreatment, neglect, abuse or degradation (s 28(1)(d)); and the right to the constitutional principle that a child’s best interests are of paramount importance in every matter concerning the child (s 28(2)).

Keightley J (with whom Francis J agreed) pointed out at [45] that there was ‘well-established jurisprudence on the rights of the child under the Constitution facilitated by the inclusion of s 28, and South Africa’s ratification of the United Nations Convention on the Rights of the Child’. The Convention was recognised by the Constitutional Court in S v M (Centre for the Child as Amicus Curiae) 2007 (2) SACR 539 (CC) at [16] as the ‘international standard against which to measure legislation and policies’. That court saw s 28 as having established a set of children’s rights that the courts are bound to enforce: statutes must be interpreted and the common law developed in a manner which favours protecting and advancing the interests of children, and courts ‘must function in a manner which at all times shows due respect for children’s rights’. This passage in S v M at [18]–[19], in the view of Keightley J (at [45]), identified the essence of what children’s rights mean under the Constitution:

‘Every child has his or her own dignity. If a child is to be constitutionally imagined as an individual with a distinctive personality, and not merely as a miniature adult waiting to reach full size, he or she cannot be treated as a mere extension of his or her parents, umbilically destined to sink or swim with them . . . . Individually and collectively all children have the right to express themselves as independent social beings, to have their own laughter as well as sorrow . . . And foundational to the enjoyment of the right to childhood is the promotion of the right as far as possible to live in a secure and nurturing environment free from violence, fear, want and avoidable trauma.’

This dictum, said Keightley J (at [47]), was ‘critical to how the constitutional issue raised in the present case should be approached’. S v M makes it clear that ‘our Constitution imagines children as their own constitutional beings’, and that they hold constitutional rights in their own respect and not through their parents. Children are entitled under the Constitution and legislation like the Children’s Act 38 of 2005 to require their parents to protect their rights, and if their parents fail in this regard, the state ‘bears the overarching obligation to ensure that children’s rights are respected, protected and enforced’ (at [61]).

The existing case law and authorities were, said the court, littered with reference to the parental right to use reasonable and moderate chastisement on their children, but it was pointed out that the defence purported, as well, to recognise the protection and wellbeing of the child in ensuring that the child was brought up in a socially acceptable manner. It was argued that these disciplinary benefits should save the defence from being regarded as an unjustifiable breach of the child’s rights.

Keightley J, however, had a number of difficulties with this submission. First, the common law did not lay down strict guidelines as to what constituted ‘reasonable chastisement’ but merely identified factors to be considered in determining what was reasonable. This was ‘deeply problematic’ as it ‘introduce[d] a level of arbitrariness to the infliction of physical punishment on children’. It left too much latitude to the parent to decide the level of physical force the child both deserved and could withstand as punishment. It was likely that many a child was ‘subjected to levels of physical punishment that, regardless of their parent’s belief, they [were] unable to withstand without harm to their physical and/or emotional state’. Such arbitrariness was ‘out of line with the child-centred model of rights demanded by our Constitution’ (at [68]).

Second, the defence was a clear violation of the rights guaranteed under s 12 of the Constitution, which gave protection from ‘all forms of violence, whether from public or private sources’ (emphasis of the court).

Third, the effect of the defence was to undermine fundamentally the critical concept of children having their own dignity. Dignity was compromised at two different levels: one being the breach of physical integrity, which also involved a measure of degradation or loss of dignity for the child; the other being the treatment by the state of the child with a ‘lesser level of concern’ and a reduced power to protect children, who were ‘effectively treated as second-class citizens by the law in this regard’ (at [72]).

Fourth, the defence treated child victims of assault by their parents differently to adult victims of assault. Children were entitled under s 9(1) of the Constitution to equal protection of the law, and under s 9(3) not to be discriminated against because of their age. The ‘reasonable chastisement’ defence did not give children equal protection under the law in that it failed to protect children from assault in circumstances where adults who were subject to the same level of force were protected (at [75]). This was not a rational differentiation within the borders of s 9. It was ‘antithetical to the constitutional right prioritising the best interests of the child’ and ‘undermine[d] the special duty owed by the state to protect children from all forms of violence and degradation, and to protect their best interests’. It rendered more vulnerable a group of people who had been singled out by the Constitution as being ‘deserving of special protection, and whose best interests [were] expressed to be of paramount importance’.

The final question was whether these breaches of s 9(1) and (3) of the Constitution could be regarded on any basis as being justifiable under s 36 of the Constitution, the so-called limitation provision. Could the defence be saved by the need to continue to permit parents to assault their children for disciplinary purposes? Keightley J could find no justifiable reason for doing so. She took into account the fears that, in doing away with the defence, ‘many well-meaning parents, who genuinely believe they are doing their best for their children, may become criminalised, as they will now be vulnerable to criminal convictions for assault’. These fears were, in the court’s view, ‘out of step with the underlying objectives of the Children’s Act, which is to promote positive parenting and positive discipline, rather than to criminalise errant parental behaviour’ (at [81]). The Children’s Act made provision for the diversion of cases to the children’s court, which has broad powers to make orders to facilitate ‘positive parenting in families’, and it was not envisaged that the demise of the ‘reasonable chastisement’ defence would lead to the ‘overcriminalisation’ of parenting behaviours. Early reporting for inappropriate punishment of children should, according to the draft policy discussed on behalf of the Minister of Social Development, ‘be referred to prevention and early intervention services’. Criminal sanctions would not be imposed ‘willy-nilly in respect of any parent who chastises their child’ (at [81]).

The court accordingly made an order, with prospective and not retrospective effect, that the common-law defence of reasonable chastisement was no longer applicable in our law.

It is difficult to fault the reasoning of the court, and one can have nothing but respect and admiration for the views expressed by Keightley J. The one reservation is the niggling suspicion that there is something wrong about criminalising a vast range of commonly observed and ubiquitous behaviours in the hope that almost none of it will lead to prosecution and conviction in practice. This places a lot of trust in the discretionary powers of those involved in separating diversion procedures from criminal prosecutions. It will be interesting to see how efficiently and appropriately these duties are discharged.

Robbery: Can the forced transfer of money by electronic means be the subject of the crime of robbery?

S v Sishuba 2018 (1) SACR 402 (WCC)

In this case the sum of R14 000 had been taken from the victim by electronic means from a bank account after the victim had been threatened with violence. Did this amount to the crime of robbery?

Henney J was unable to find any case law on the matter, and he decided the issue on the basis of principle and the weight of academic opinion. ‘Robbery’ is defined by Burchell Principles of Criminal Law 4 ed at 706 as consisting in ‘the theft of property by intentionally using violence or threat of violence to induce a person to submit to the taking of the property’. If there was theft of the money in Sishuba, there would clearly be robbery as well.

So, was it theft? Henney J turned at [91] to the academic commentators for guidance. Burchell (at 690) says this:

‘If it is possible to steal money in an incorporeal form, then clearly it is possible to steal money in its credit form. This means that it can be theft to deposit dishonestly a cheque drawn on another person’s bank account, or dishonestly to instruct a computer to transfer money from one account to another. Although there is no actual physical handling of the money, the conduct amounts to theft.’

And Snyman Criminal Law 6 ed at 478 agrees that the taking of money in the form of credit amounts to theft, adding (at 493):

‘The most obvious meaning of money is corporeal notes or coins. However, money may also have a less obvious and more abstract meaning, namely credit. By credit is usually meant a right to claim money from a bank, because the bank is the owner of the money which is in the bank, where the bank’s client only has a right to claim from the bank. In modern business usage cash is seldom used. Money generally changes hands by means of cheques, negotiable instruments, credit or debit entries in books, or registration in the electronic memory of a computer. In these cases one can hardly describe the money in issue as tangible, corporeal articles. It would be more correct to describe it as economic assets, ‘‘an abstract sum of money’’, ‘‘a unit representing buying power’’, or . . . ‘‘credit.’’’

On the strength of these views, Henney J was prepared to conclude (at [94]) that the crime of robbery could be committed by the theft of an incorporeal thing through violence or force.

There can be no doubting the correctness of Henney J’s conclusion. It is, however, worth taking note of a caveat expressed by Snyman at 478 and cited in the judgment at [91]. It is that this notion differs from the ordinary principles governing theft to such an extent that it cannot be accommodated under the traditional definition of that crime ‘without radically amplifying the ordinary meaning of the words’.

Burchell defines ‘theft’ as consisting in ‘an unlawful appropriation with intent to steal of a thing capable of being stolen’. The two issues raised in Sishuba are, first, whether there has been an ‘appropriation’, and, second, whether the credit was something ‘capable of being stolen’. On the first question it is significant that the word ‘contrectatio’, which denotes a touching or a handling, is no longer favoured by modern commentators, although it was required by Roman law and by the earlier cases. Snyman defends this choice on the ground that while ‘contrectatio’ ‘might have been a satisfactory criterion centuries ago when the economy was still relatively primitive and primarily based on agriculture’, today’s world, with its much more complicated economic structure, is better suited to the more abstract concept of appropriation to describe the act of theft. He is probably correct. To insist on a ‘contrectatio’ leads to a more and more synthetic understanding of that notion, so that it ends up, as Snyman observes, as ‘merely . . . a technical ‘‘erudite-sounding’’ word’ to describe that act. The term ‘appropriation’, says Snyman, describes precisely what our courts in practice understand by the act of stealing.

The second question relates to the kind of property that can be stolen. Snyman is right in pointing out that abstract, incorporeal things such as ‘ideas’ or ‘board and lodging’ cannot be stolen. But he points out further that this rule has to be viewed ‘circumspectly’. Not to recognise that money in its incorporeal form such as credit can be stolen would not, he suggests, be in keeping with judicial practice. It would, too, it is submitted, fly in the face of commercial custom, common sense and the fundamental dictates of justice.

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

Section 41(6) of the National Prosecuting Authority Act 32 of 1998: The discretion of the National Director of Public Prosecutions

Maharaj & others v Mandag Centre of Investigative Journalism NPC & others 2018 (1) SACR 253 (SCA)

Section 41(6)(c) of the National Prosecuting Authority Act 32 of 1998 (NPA Act) provides that no person shall without the permission of the Director of Public Prosecutions (NDPP) disclose to any other person the record of any evidence given at an investigation conducted in terms of s 28(1) of the NPA Act. Section 28 regulates investigations by an Investigating Director appointed in terms of the NPA Act. See Chapter 5 (ss 26–29) of the NPA Act for the powers, duties and functions of the Investigating Directorates.

Maharaj (supra) concerned the proper exercise of the NDPP’s discretion in terms of s 41(6)(c) to grant or refuse permission to publish the record of any evidence given at an investigation conducted in terms of s 28(1). In this case the NDPP had denied a national newspaper, Mail & Guardian (hereafter ‘M&G’), permission to publish information which formed part of the record of a s 28 investigation. This denial frustrated the publication of an M&G article alleging that in the course of the relevant s 28 investigation, a cabinet minister and his wife had failed to disclose information and had, furthermore, provided false information.

On 12 May 2016 the High Court had reviewed and set aside the NDPP’s refusal to grant permission. It also granted the M&G permission ‘to publish the s 28 record’. See M and G Centre for Investigative Journalism NPC & others v National Director of Public Prosecutions & others (unreported, GP case no 37510/2012, 12 May 2016) at [128].

In dealing with the NDPP’s appeal against the High Court’s order, the Supreme Court of Appeal—in a unanimous judgment written by Ponnan JA—made three essential observations regarding the purpose, interpretation and application of s 46(1)(c) of the NPA Act. First, the purpose of the requirement that prior permission to publish must be sought and obtained from the NDPP ‘is . . . to protect the integrity of the criminal justice system’ (Maharaj at [21]). Second, this requirement is a limitation on the right to freedom of expression contained in s 16 of the Constitution and does, in the context of a case like Maharaj, also limit media freedom ‘and the correlative right of the public to receive and impart information’ (at [21]). Third, in exercising the discretion conferred by s 41(6), the NDPP must take the ‘utmost care’ to ensure that in each case the appropriate balance is struck ‘between securing the integrity of the criminal justice system and upholding freedom of expression’ (at [21]).

Ponnan JA pointed out that whereas the M&G wanted to publish information on an issue of high public interest, there was ‘no valid countervailing concern regarding the integrity of the administration of the criminal justice system’ (at [22]). In his view the integrity of the criminal justice system itself required that the M&G be allowed to inform the public (‘this country’s citizenry’) of the responses of a senior public official to allegations of unlawful conduct involving public money (at [22]). The NDPP had advanced nothing of substance to counter this consideration. At [23] Ponnan JA said:

‘The NDPP approaches the matter so as to elevate the policy considerations underlying the relevant provisions of the Act to the level of directives from which no deviation is permitted. The Act confers a discretion on the NDPP. That discretion must be properly exercised. The express conferral of a discretion clearly contemplates that there will be circumstances where disclosure is appropriate.’

The NPA Act does not state or list the factors or considerations which the NDPP must take into account in exercising the discretion conferred by s 41(6). But Ponnan JA had no problem in identifying ‘a consideration of the s 28 record’ as ‘the first and most obvious factor’ (at [24]). The NDPP thought otherwise. Some four months after having deposed to her answering affidavit, the ‘startling revelation’ was made that the NDPP had not considered the s 28 record in coming to her decision to refuse permission to the M&G to publish. Ponnan JA found as follows (at [24], emphasis added):

‘Without a consideration of the s 28 record, the discretion conferred could not have been properly exercised. It was not sufficient for the NDPP to state, as she subsequently did, that she was aware of the s 28 interviews in ‘‘general terms’’. What was meant by that rather vague expression remained unexplained. The NDPP is no ordinary litigant. She is an officer of the court, who is duty-bound to take the court into her confidence and fully explain the facts so that an informed decision can be taken.’

In the absence of a careful study by the NDPP of the contents of the s 28 record, it was difficult to comprehend how the NDPP ‘could have performed the vital balancing exercise of weighing the public interest in publication against the likelihood of harm’ (at [25]). Permission to publish should not have been refused ‘without a proper consideration of the record at the time of making her decision’ (at [25]). The NDPP’s failure to consider the record rendered her decision ‘irrational’ (at [25]). This finding alone was sufficient to set the NDPP’s refusal aside. However, Ponnan JA also examined some of the reasons advanced by the NDPP in support of her decision to refuse permission.

The NDPP’s reliance on public interest and the interests of third parties was rejected. The NDPP’s ‘public interest’ argument was that ‘the interests of the public also extend to ensuring that legitimate institutions . . . effectively serve their intended purposes’ (at [10.9]). However, without having examined the s 28 record the NDPP could not contend that she did engage in a delicate balancing exercise (at [20]). Ponnan JA said: ‘Any consideration of the public interest based on a general awareness of the investigation suggests that it had to have been superficial’ (at [29]). The NDPP’s assessment was unsatisfactory given the fact that the M&G’s request for permission to publish concerned ‘the probity of a senior public office bearer’ and involved constitutional values of accountability and the role of the media in reporting on corruption (at [27] and [28]).

Ponnan JA gave three reasons for rejecting the NDPP’s claim that she had refused permission to publish in order to protect the interests of third parties mentioned in or affected by the s 28 record. First, given the NDPP’s failure to consider the s 28 record, ‘third party interests could not have featured as a consideration’ and there could not have been a proper determination of the question whether such interests required any protection (at [31]). Second, at the time the NDPP decided to refuse permission to publish, ‘the information in question was already a matter of public knowledge’ and there was, accordingly, ‘no longer any basis for confidentiality’ (at [31]). Third, even on the assumption that some third party interests could have been affected by disclosure, there was ‘no reason in principle why disclosure could not have been permitted subject to certain conditions, such as protecting the identities of those third parties’ (at [31]).

In support of her refusal to permit publication, the NDPP also stated that it was not the policy of the NPA to disclose the record of evidence given at a s 28(1) investigation (at [10.3]). According to the NDPP this policy was underpinned by paragraph 13(c) of the United Nations Guidelines on the Role of Prosecutors (hereafter the ‘UN Guidelines’). This paragraph states that prosecutors shall in the performance of their duties keep ‘matters in their possession confidential, unless the performance of duty or the needs of justice require otherwise’. However, Ponnan JA found that the UN Guidelines did not provide a proper basis for ‘the policy of non-disclosure’ relied upon by the NDPP. The UN Guidelines, said he at [32], could not trump s 41(6) of the NPA, which ‘expressly envisages publication with the necessary permission’. He also pointed out that paragraph 13(c) of the UN Guidelines was, in any event, not phrased in absolute terms in that the paragraph itself acknowledged exceptions where performance of prosecutorial duties or the needs of justice might require disclosure (at [32]). The NDPP’s policy of non-disclosure was also inconsistent with the very purpose of s 41(6) which requires the NDPP to exercise ‘an appropriate discretion on a case-by-case basis’ (at [32] and [39]).

Ponnan JA also found it necessary to make broad but important observations regarding the role of the NDPP in the context of our criminal justice system, the role of the media and the need to combat crime in our constitutional democracy (at [38], emphasis added):

‘The objective of policing state officials to guard against corruption and malfeasance in public office forms part of the constitutional imperative to combat crime. The NDPP is an important bulwark in that regard. The NDPP is there to inspire confidence that all is well and, if there are corruption and malfeasance in high public office, that it is being effectively dealt with. The public needs to be assured that there is no impropriety in public life and that if there is, then it should be exposed. In that sense, the media play a vital watchdog role. One of the aspirational goals of the media is to make governmental conduct in all of its many facets transparent.’

In conclusion it was noted that the factors identified, considered or weighed by the NDPP could neither individually nor collectively survive scrutiny (at [40]). The NDPP’s appeal was dismissed.

Section 40(1)(b): Unlawful arrest: Is the Minister of Police liable, following an unlawful arrest, for the further detention of the arrested person after he or she has been remanded to custody by a court?

De Klerk v Minister of Police [2018] 2 All SA 597 (SCA)

The appellant had been arrested without a warrant following an alleged assault. The police relied on s 40(1)(b) of the Criminal Procedure Act. That section permits a peace officer to arrest, without a warrant, any person ‘whom he reasonably suspects of having committed an offence referred to in Schedule 1, other than the offence of escape from lawful custody’. It was argued that ‘assault when a dangerous wound is inflicted’, which is listed in Schedule 1, was in point in that the peace officer effecting the arrest entertained a reasonable suspicion that this offence had been committed. The evidence, however, did not support this. There was no evidence that an investigation had been carried out to ascertain the nature and extent of the wound; the medical report (J88), which was presented to the court a quo, was illegible and the original could not be located; and the arresting officer relied on only the complainant’s statement and the medical report when she made the decision to arrest. This was clearly insufficient, and it could not be said that the jurisdictional requirements of the section had been satisfied or that the necessary discretion had been exercised in a proper manner.

The more difficult part of the inquiry was whether the minister should be liable for the ensuing detention of seven days which occurred after the appellant had appeared in court and been remanded in custody. This question drew a divided response from the Supreme Court of Appeal.

The majority (Shongwe ADP, with whom Majiedt JA and Hughes AJA agreed) took the view that what happened in court and thereafter could not ‘be placed before the doorstep of the respondent’ (at [12]). The majority judges followed what was said by Harms DP in Minister of Safety and Security v Sekhoto 2011 (1) SACR 315 (SCA) at [42]–[44]. Harms DP was of the opinion that once an arrest had been effected, and the arrestee had been brought before a court within the required 48 hours, the authority to detain that was inherent in the power to arrest had been exhausted, and the authority to detain the suspect further was then within the discretion of the court. The arrestor, he considered, was not called upon to determine whether the suspect ought to be detained pending a trial, since that was the role of the court or, in some cases, a senior officer. The purpose of the arrest was no more than to bring the suspect before the court (or senior officer) ‘so as to enable that role to be performed’.

In this case the police actually recommended bail and took the suspect to court within two hours of his arrest. It was then, said Shongwe ADP, for the court to inquire from the prosecution why it was necessary to detain the suspect further and to ensure that his rights under s 35(1)(e)(f) of the Constitution were not undermined. A failure to inquire into the reasons for further detention ‘is clearly a contravention of [these] constitutional imperatives and therefore the further detention of a suspect without just cause would be arbitrary and unlawful’ (at [14]). The police, however, were simply ‘doing their job by taking the suspect to court’ (at [15]). From that point it would be the justice department who ‘would be responsible and liable for the further detention because of its failure to observe the constitutional rights of a detained person’ (at [14]). The decision in Minister of Safety and Security v Tyokwana 2015 (1) SACR 597 (SCA) was distinguished by the majority on its facts: the complainant in that case knew that there was simply no evidence upon which the arrestee could successfully be prosecuted; he was aware that the initial statements of the state witnesses were obtained under duress and were false; and the arrestee had been seriously assaulted by the police.

The respondent, said Shongwe ADP, could be held liable only for the two hours until the appellant appeared in court, during which time he had not even been locked up in a holding cell. For this he was awarded a sum of R30 000 in damages.

Rogers (AJA) (with whom Leach JA agreed), took a different approach. He viewed the matter through the lens of the ‘usual rules of delictual liability’, according to which ‘a wrongdoer is liable for all the harmful consequences of his or her wrongful act’ (at [29]; emphasis added). It was necessary, in this regard, to take particular notice of how the test for causation had been applied by the courts to situations where ‘the act of a third party, itself having causal effect, intervenes between the act of the wrongdoer and the harmful consequence but where the wrongdoer is still held liable for the harmful consequence’ (at [32]). If, for instance, A injures B so that B requires surgery, he may still be liable if B dies during the course of lawful surgery: ‘one does not . . . automatically conclude that the wrongdoer is not liable’; ‘[r]ather, one asks whether, in accordance with the well established requirements for delictual liability, the wrongdoer can be held responsible for the harmful consequence’ (at [32]). There was, said Rogers AJA (at [33]), ‘no reason for not following the same approach in determining the harmful consequences of an unlawful arrest for which the police may be held liable’.

After considering the decisions in Zealand v Minister of Justice and Constitutional Development 2008 (4) SA 458 (CC), Woji v Minister of Police 2015 (1) SACR 409 (SCA), Tyokwana (supra) and Thandani v Minister of Law and Order 1991 (1) SA 702 (E), the minority judges concluded as follows: first, the test for factual causation was plainly satisfied in that, but for the unlawful arrest, the appellant would not have been brought before the court and there would have been no occasion for the court to remand him in custody. Second, as to legal causation, the ‘direct consequences’ test was satisfied since it was a legal requirement, once the appellant had been arrested, that he be brought before the court. The court could then make one of only two decisions—to remand him in custody or grant him bail—either of which outcomes would be a direct consequence of the arrest. Third, as to foreseeability, either outcome was foreseeable to the arresting officer who actually admitted that she subjectively foresaw that he would be remanded in custody until his next appearance: she knew that at the first appearance the remand would be a routine or mechanical act rather than a considered judicial decision, as it ought to have been. Fourth, there were no reasons of policy for not holding the police liable for further detention in such circumstances. The test for a so-called ‘novus actus interveniens’ involved a value judgement since it was ‘impossible, by applying any absolute objective test, to say whether an intervening act ‘‘breaks the causal chain’’’. Fifth, an arresting officer should not be held liable for detention ordered by a court pursuant to a deliberative or considered judicial process unless the arresting officer has the full animus iniuriandi required for malicious deprivation of liberty’ (at [44]). There were, however, ‘no reasons of policy for disregarding . . . those harmful consequences which flow from the arrest mechanically or as a matter of routine’ (at [45]). ‘Only where a judicial officer has applied his or her mind to the question whether the subject is entitled to bail does one have an intervening act which can be said, from the perspective of policy, to neutralise the harmful effect of the wrongful arrest’.

On the facts of De Klerk’s case, it was the minority’s view that there was no considered decision by the magistrate as to whether the appellant should be released on bail. The evidence indicated that the prosecutor and magistrate had failed to comply with their duties, whereas the appellant was not legally represented and seemed to have been overwhelmed by the circumstances. Moreover, since intended or foreseen consequences can never be too remote to attract liability (see Thandani supra at 705F–G), the minority held that the appellant’s damages should be assessed with reference to the full period of detention.

The minority, it should be noted, did not consider it necessary to find the actual detention following the judicial remand to be ‘unlawful’ before concluding that it was a harmful effect for which the wrongful arrester could be held responsible. As long as the necessary causal nexus between that effect and the initial wrongful act (the arrest) could be established, liability would ensue under the actio iniuriarum, which is the action that governs liability for non-patrimonial damages for wrongful arrest and detention. This action is, as Rogers AJA pointed out (at [43]), a general one relating to infringements of personality rights, of which liberty is one.

The reasoning of Rogers AJA is attractive. It is rooted in the fundamental principles of the law of delict and it invokes well established notions and precepts of the common law. If there is any weakness in his well-considered and thoroughly researched judgment, it may be in his application of some of those principles, in particular, his assertion that a novus actus interveniens may be said to exist in such circumstances ‘only where a judicial officer has applied his or her mind to the question whether the suspect is entitled to bail’ (emphasis added). The fact remains that a judicial officer has a legal duty to apply his or her mind properly to this question. If he or she does so, I think Rogers AJA is correct in considering the discharge of the judicial duty to be a novus actus interveniens. Even if the judicial officer, having applied his or her mind to the question properly, reaches the wrong result, it would plainly be wrong to regard the ensuing harm effected by the detention as being ‘caused’ by the initial wrongful arrest. But what of the converse situation, where the judicial officer does not properly apply his or her mind to the question of bail? A failure to do so would be a clear breach of the judicial duty, but how can such a situation be distinguished from the situation (alluded to by Rogers AJA himself) where surgery is negligently (and thus improperly) performed on a person following an assault by the initial wrongdoer? The doctor’s failure properly to discharge the duty to the patient seems, in this context, no difference in principle to the judicial officer’s failure properly to discharge the duty to the arrested person. And if the latter failure will ordinarily constitute a novus actus interveniens, why should we take a different approach to the former?

On the facts of De Klerk’s case, this objection does not threaten what is submitted to be the correct ultimate finding of the minority judges. Because it was clear that the arresting officer subjectively foresaw that the magistrate would be in breach of the duty in question, and because it is accepted in our law that a foreseen event cannot constitute a novus actus, the integrity of the causal nexus would remain intact. Had this subjective foresight not been found to exist, however, it is difficult to see how the minority decision could have stood.

Section 60(4)(e), (8A) and (9): Bail and the public demand that bail be refused

S v Nel & others 2018 (1) SACR 576 (GJ)

Section 60(4) of the Criminal Procedure Act 51 of 1977 provides that the interests of justice do not permit an accused’s release on bail where one or more of five possible grounds are established. In terms of s 60(4)(e) one of these grounds is ‘where in exceptional circumstances there is the likelihood that the release of the accused will disturb the public order or undermine the public peace or security’. Section 60(8A), in turn, provides that in considering whether the ground identified in s 60(4)(e) has been established, a bail court may, where applicable, take into account factors identified in s 60(8A)(a) to (f). One such factor is whether the nature and circumstances of the offence are ‘likely to induce a sense of shock or outrage in the community where the offence was committed’ (s 60(8A)(a)); another is ‘whether the shock or outrage of the community might lead to public disorder if the accused is released’ (s 60(8A)(b)). See further the discussion of s 60 in Commentary, sv Where in exceptional circumstances there is the likelihood that the release of the accused will disturb the public order or undermine the public peace or security: s 60(4)(e) as read with s 60(8A).

Almost two decades ago the Constitutional Court held that s 60(4)(e) and (8A) passed constitutional muster. See S v Dlamini; S v Dladla & others; S v Joubert; S v Schietekat 1999 (2) SACR 51 (CC). It was held that although the provisions of s 60(4)(e) and (8A) infringed the detained person’s liberty interest as protected by s 35(1)(f) of the Constitution, this infringement was a constitutionally permissible limitation as provided for in s 36(1) of the Constitution (at [55]). However, the fact that s 60(4)(e) and (8A) has survived a constitutional challenge does not mean that there are no real inherent risks and difficulties in applying these subsections. This much was indeed acknowledged and foreseen by the Constitutional Court in Dlamini itself. At [56] Kriegler J, writing for a unanimous Constitutional Court, pointed out that courts would ‘no doubt be alive to the danger of public sentiment being orchestrated by pressure groups to serve their own ends’. It was also observed that as far as s 60(4)(e) and (8A) is concerned, ‘the field of application . . . will be extremely limited’ and that bail courts would in respect of these sections proceed ‘with great circumspection in the knowledge that the Constitution protects the liberty interests of all’ (at [57]). Kriegler J also stressed that the exceptional circumstances referred to in s 60(4)(e) must be established on a preponderance of probabilities and—if so established—s 60(4)(e) must still be weighed against the considerations enumerated in s 60(9) before a decision to refuse bail can be taken (Dlamini at [57]). Section 60(9) requires that the interests of justice be weighed against the accused’s right to personal freedom and the possible prejudice he may suffer should bail be refused. See the discussion of s 60 in Commentary, sv Section 60(9).

The recent decision in S v Nel & others 2018 (1) SACR 576 (GJ) illustrates how easily a bail court can be led astray by s 60(4)(e) and (8A) if it fails to follow the guidelines set out by Kriegler J in Dlamini.

In Nel (supra) the three appellants appeared in the district court on three serious charges: attempted murder, pointing a firearm and assault with intent to do grievous bodily harm (at [1]). They appealed against the magistrate’s refusal to grant them bail. One of the magistrate’s main reasons for refusal of bail was her finding that, having regard to s 60(4)(e) and the factors in s 60(8A), there was a ‘likelihood that the release of the appellants would disturb the public order or undermine public peace and security’ (at [26]).

At the time of the bail hearing in the lower court, there was a public outcry on social media and call by protesters that bail be refused. All this was triggered by the fact that the charges against the appellants were based on incidents which, in the words of Pietersen AJ, ‘manifested racial connotations or undertones’ and which involved what is ‘often described as so-called ‘‘white on black’’ violence’ (at [23]). In these circumstances, noted Pietersen AJ, the public outcry and call for bail to be refused were ‘understandable’ (at [23]). However, the essential question was whether the bail court had, upon proper proof of all relevant facts, correctly interpreted and applied s 60(4)(e) as read with s 60(8A). At [23] reference was made to S v Schietekat 1999 (1) SACR 100 (C) at 104h–i, where Slomowitz AJ noted the unacceptability of ‘lynch law’ and stressed the duty of courts ‘to ensure the maintenance of law, order and justice’.

In S v Nel & others (supra) Pietersen AJ remarked and concluded as follows as regards the bail magistrate’s refusal to grant bail to the three appellants (at [24]):

‘No objective facts of the likelihood, and not possibility, of the eventualities envisaged s 60(8A) were presented to the court from which the magistrate drew her inferences. The magistrate appears clearly to have been influenced by the events which manifested themselves in the social media, comments emanating from the Minister of Police on Twitter, and protesters who had gathered, opposing the release of the accused on bail. On what public outcry constitutes, the magistrate indicated that she did not need a dictionary to tell her what public outcry was, but had merely to have regard to s 60(8A). It is apparent that the magistrate paid lip service to the statutory provision.’

In Nel Pietersen AJ identified a further misdirection by the bail magistrate. The likelihood that the granting of bail to the appellants would disturb the public order or undermine the public peace and security was never weighed against the considerations set out in s 60(9). This omission, concluded Pietersen AJ at [26], was ‘on its own a material misdirection’ because it was inconsistent with the Constitutional Court’s decision in Dlamini (supra) at [57]. The magistrate’s refusal of bail in Nel was set aside on account of this misdirection as well as other misdirections which Pietersen AJ identified at [11], [19] and [21]. The appeal court was at liberty to give the decision which the bail court should in the first instance have given. Bail in the amount of R5 000 was fixed in respect of each appellant. Certain bail conditions were also determined.

There can be no doubt that magistrates are under tremendous pressure when they have to decide bail applications in situations where protesters who are members of the community in which the offence was committed, are outside court demanding that bail should be refused. This happens most frequently where racism plays a real or alleged role in the commission of a violent crime, or where a child is a rape victim, or where an alleged serial rapist has been apprehended. However, the community’s demand and sentiments can never usurp the bail court’s judicial duty to consider release on bail in view of constitutional principles and rights, all relevant statutory provisions as well as possible precedents established in case law. Another problem regarding the provisions of s 60(4)(e) and (8A) is that they appeal to ‘an uninformed or ignorant public’. See Van der Berg Bail—A Practitioner’s Guide 3 ed (2012) 150. In 1999 Kriegler J also noted that there was ‘widespread misunderstanding regarding the purpose and effect of bail’ and that ‘much must still be done to instil in the community a proper understanding of the presumption of innocence and the qualified right to freedom pending trial under s 35(1)(f) [of the Constitution].’ See Dlamini (supra) at [55]. Mokoena A Guide to Bail Applications (2012) 84–5 has confirmed that the public’s response to the granting of bail ‘is primarily based on ignorance of the legal implications at play in general, and of the bail regime in particular’. It is unlikely that the position will change any time soon. However, magistrates—who are really at the coalface when it comes to making bail decisions when the community demands that bail be refused—can take some comfort in the decision by Pietersen AJ in Nel (supra), which is to the effect that the provisions of s 60(4)(e), (8A) and (9) form a coherent whole and can never be interpreted to mean that the public’s opinion must or can be elevated to the point where facts and applicable legal rules must be ignored. The rule of law must prevail. It would be constitutionally impermissible for it not to.

Section 67(1): Bail and failure to appear: Validity of court order staying execution of warrant

S v Lerumo & others 2018 (1) SACR 202 (NWM)

On 23 January 2017 one NM, who was accused no 4 in Lerumo (supra), was released on bail in the amount of R1 000. On 5 May 2017 NM failed to appear in court. It was common cause that on this date he was indeed required to appear in court. His attorney informed the presiding magistrate that non-appearance was due to the fact that NM ‘was attending an initiation school’ (at [1]). The attorney suggested that should a warrant for the arrest of NM be issued, the execution thereof should be stayed. However, the presiding magistrate, acting in terms of s 67(1) of the Criminal Procedure Act 51 of 1977, ‘issued a warrant for the arrest of the accused with immediate execution’ (at [1]). Relying on s 67, the presiding magistrate also declared NM’s bail provisionally cancelled and his bail money provisionally forfeited to the State. The matter was then remanded to 19 May 2017.

A senior magistrate sought reasons from the presiding magistrate for her decision to issue an arrest warrant with immediate execution. The latter’s response was that she had no discretion to stay the execution of the warrant (at [2]). According to her, s 67 required ‘as an immediate action’ (as she put it) that a warrant be issued in the event of non-appearance (at [14]). The senior magistrate disagreed. In sending the matter on special review for the High Court’s ‘guidance’ and making of an ‘appropriate order’, the senior magistrate stated that over the years ‘a rule of practice’ had developed in terms of which ‘the execution of a warrant of arrest is stayed in situations where the accused . . . is unable to attend the court proceedings due to illness, hospitalization and other unforeseen and compelling circumstances’ (at [2]).

‘The issue for determination’, said Hendricks ADJP in Lerumo at [3], ‘is whether a magistrate can issue a warrant of arrest, but order that its execution be stayed’. At [5]–[6] it was found that s 67 is ‘peremptory and mandatory’ and that in the event of non-appearance of an accused who was on bail, a court does not have a discretion to stay the execution of the warrant. The relevant portion of s 67(1) provides that ‘the court before which the matter is pending shall declare the bail provisionally cancelled and the bail money provisionally forfeited to the State, and issue a warrant for the arrest of the accused’.

In Lerumo Hendricks ADJP (Djaje J concurring) pointed out that s 67 provides a comprehensive procedure which regulates and determines the consequences of non-appearance (at [6]). Non-appearance, said the court at [9], has the ‘inevitable consequence that a warrant of arrest be authorised’. After having referred to several cases dealing with the provisions of s 67 and the warrant of arrest to be issued in the event of non-appearance, the review court concluded that the presiding magistrate was ‘correct’ and that ‘this matter is therefore not reviewable’ (at [15]). This finding is most certainly correct in the sense that on the facts in Lerumo the attendance of an initiation school as the reason for NM’s non-appearance in court left the presiding magistrate with no option but to issue the warrant ‘with immediate execution’, despite an attempt by NM’s attorney to have the execution of the warrant stayed (at [1]). There was nothing on record to indicate that NM’s non-appearance was perhaps due to circumstances beyond his control, for example, that he had been kidnapped or otherwise forced to attend the initiation school on the date scheduled for his court appearance. However, it would appear that in Lerumo the review court also targeted the validity of the rule of practice that the senior magistrate referred to in submitting the matter on special review. At [16] Hendricks ADJP held:

‘I am of the view that the practice of issuing warrants of arrest for accused persons and staying the execution thereof is not in accordance with the prescripts of s 67 of the CPA, and should be done away with, unless the legislature amends the said section. Until then, this practice must be stopped.’

This finding is sweeping and wide-ranging. It has serious implications for our courts in their application of s 67 and for bail as a constitutionally entrenched means of securing and protecting liberty pending the final outcome of a criminal trial. Does the finding in Lerumo at [16] mean that an accused’s non-appearance must—regardless of the information available to the court—result in an arrest warrant with immediate execution? Is there no discretion to stay the execution of the warrant?

A good example of the rule of practice referred to by the senior magistrate in Lerumo when he submitted the matter for special review, can be found in S v Porritt & another (unreported, GJ case no SS 40/2006, 21 July 2017) where accused no 1, who was on bail, failed to appear in court on 12 June 2017 for the continuance of his trial. On the basis of information furnished to the court regarding the health of accused no 1, Spilg J authorised a warrant for his arrest but also ordered him ‘to show cause on Monday, 19 June 2017 at 10 am why the warrant should not be issued and his bail estreated’ (at [6]). Indeed, in Porritt at [38] Spilg J referred to ‘the current practice of only authorising or issuing but not executing a warrant of arrest for non-appearance’.

In Lerumo the review court referred to and quoted from six cases: Terry v Botes & another 2003 (1) SACR 206 (C); Da Costa v The Magistrate, Windhoek & others 1983 (2) SA 732 (SWA); S v Cronje 1983 (3) SA 739 (W); S v Mudau 1991 (1) SACR 636 (W); S v Nkosi en andere 1987 (1) SA 581 (T) and S v Engelbrecht 2012 (2) SACR 212 (GSJ). All these cases dealt, directly or indirectly, with the consequences of a failure to appear and the fact that s 67(1) requires that a warrant be issued. However, none of these cases addressed the legality of the rule of practice as applied in Porritt (supra) and seemingly abolished in Lerumo at [16].

The practice ‘of only authorising or issuing but not executing a warrant of arrest for non-appearance’ (Porritt at [38]) does not ignore the fact that s 67(1) requires that non-appearance be met with a warrant. It merely means that a court has a discretion—to be exercised on the basis of information made available to it—to remand its finalisation of the warrant of arrest to a specified date when evidence or further information regarding the reason for non-appearance might be available. A typical situation is the one that arose in Porritt (supra).

It is submitted that the rule of practice finds a statutory basis in the provisions of s 168 of the Criminal Procedure Act. This section provides as follows:

‘A court before which criminal proceedings are pending, may from time to time during such proceedings, if the court deems it necessary or expedient, adjourn the proceedings to any date on the terms which to the court may seem proper and which are not inconsistent with any provision of this Act.’

The words ‘on the terms which to the court may seem proper’ are wide enough to accommodate a situation where a court, in remanding the matter to a specific date, stays the execution of a s 67(1) warrant so that a final decision can be made on further information (which may or may not be available) regarding the reason(s) for the non-appearance which triggered the arrest warrant.

An argument that non-appearance must in all circumstances be met with an arrest warrant with immediate execution and that a court has neither the power nor the discretion to stay the execution of the warrant pending a finding based on further information or evidence, is in conflict with the right to freedom and security of the person as entrenched in s 12(1)(a) of the Constitution: ‘Everyone has the right to freedom and security of the person, which includes the right . . . not to be deprived of freedom arbitrarily or without just cause’. Even an accused who was released on bail and then fails to attend court, may not be deprived of his freedom arbitrarily or without just cause. See also generally S v Matitwane 2018 (1) SACR 209 (NWM) which is discussed elsewhere in this edition of Criminal Justice Review. The rule of practice as applied in Porritt accommodates constitutionally required procedural and substantive fairness, without compromising the essential principle that wilful non-attendance must result in arrest. This approach is entirely consistent with the provisions of s 67. To conclude—as was done in Lerumo at [16]—that ‘the practice of issuing warrants of arrest . . . and staying the execution thereof is not in accordance with the prescripts of s 67’, is to overlook or ignore the fact that s 67 cannot be read and interpreted in isolation from s 168 of the Criminal Procedure Act and provisions such as those entrenched in s 12(1)(a) of the Constitution.

Section 68: The irregular cancellation of bail

S v Matitwane 2018 (1) SACR 209 (NWM)

It is an established principle of our law that in bail matters a court may be proactive: it can take the initiative in raising the question of release on bail and it may act inquisitorially to secure evidence to determine this question. See the discussion of s 60 of the Criminal Procedure Act 51 of 1977 in Commentary, sv Section 60(3). However, it is also true that a bail inquiry remains ‘an ordinary judicial process’ which must, despite its own necessary peculiarities, ‘be conducted impartially and judicially and in accordance with the relevant statutory prescripts’ (Nugent JA in S v Mabena & another 2007 (1) SACR 482 (SCA) at [7]).

In Mabena (supra) neither the prosecution nor the defence had any advance notice that, on the day that the accused’s case was set down for remand, the presiding judge would raise and deal with the matter of bail in respect of the two accused charged with robbery and murder. The prosecutor’s objections were ‘summarily’ dismissed by the presiding judge (at [29]). The prosecutor’s request for an adjournment to consider and assess the matter of bail was also refused by the presiding judge, who granted each accused bail of R1 000, with certain conditions. The State’s application for leave to appeal was dismissed by the presiding judge and the matter reached the Supreme Court of Appeal only after the latter court had granted the required leave to appeal. On appeal Nugent JA (Harms and Streicher JJA concurring) concluded that the proceedings ‘were not conducted judicially’ and amounted to ‘no inquiry at all’ as contemplated by Act 51 of 1977 (at [28]). The granting of bail by the court a quo was set aside, especially since it appeared that the judge had already ‘made up his mind’ before raising the matter (at [28]).

In contradistinction to Mabena (supra) which dealt with the granting of bail, S v Matitwane 2018 (1) SACR 209 (NWM) concerned the presiding judicial officer’s summary decision to cancel bail.

In Matitwane (supra) the applicant sought a review and setting-aside of a regional court’s order cancelling his bail. At the commencement of his trial in the regional court, the applicant had pleaded guilty. However, this plea was contrary to the instructions given to his legal representative. They were granted an opportunity to solve this problem. However, they ‘could not resolve their misunderstanding’ and the legal representative withdrew from the case (at [4]). The regional magistrate once more explained the right to legal representation to the applicant, who insisted on conducting his own case ‘as he did not want to be seen as delaying the matter’ (at [5]). He repeated his plea of guilty but the regional court invoked s 113 of Act 51 of 1977 and corrected his plea to one of not guilty. The trial resumed some 14 days later. When given an opportunity to cross-examine the first state witness, the applicant wanted legal representation and indicated that ‘he would not be able to cross-examine the witness on his own’ (at [6]). He made the frank admission that at his earlier appearance he had made a mistake by electing to conduct his own defence. This led to some argument or disagreement between the court and the applicant about the latter ‘causing a delay’ (at [6]). The result was that the applicant cross-examined the witness on his own, after which he requested a remand so that he could brief a legal representative. The request was granted. However, the regional court also mero motu cancelled the applicant’s bail which was granted to him almost a year earlier and which he had as yet not paid. According to the regional court magistrate, the applicant’s conduct amounted to delaying tactics and indicated that he had no intention of finalising the matter (at [6] and [8]). She—in her own words—feared that the applicant ‘might pay bail and disappear’ (at [6]).

On appeal it was found that the regional court magistrate had acted outside s 68 of Act 51 of 1977. She had given no consideration to this section which governs the cancellation of bail and sets out the grounds for cancellation of bail. Section 68(1) deals with the situation where ‘a charge is pending in respect of which bail has been granted, whether the accused has been released or not’. Several grounds for cancellation of bail are stipulated in s 68(1), for example, where the accused has become a flight risk (s 68(1)(a)) or has attempted to interfere with witnesses (s 68(1)(b)) or where it would be ‘in the interests of justice’ to cancel bail (s 68(1)(g)). The grounds upon which bail can be cancelled are in principle linked to the grounds that must be considered when the issue is whether bail should be granted: compare s 68 and s 60(4) and see generally S v Kyriacou 2000 (2) SACR 704 (O) at 711b. Section 68(1) also requires that ‘information on oath’ be received by the court before a finding can be made regarding the existence of one or more of the grounds in s 68(1)(a) to (g).

In Matitwane (supra) Kgoele J concluded that the regional court’s cancellation of the applicant’s bail was inconsistent with s 68(1). There was no information on oath in support of the cancellation and the magistrate had ‘acted wholly mero motu without being authorised by any legislation to take the step in the manner she did’. The regional court magistrate had, furthermore, taken her decision without giving the applicant an opportunity to respond or object to the cancellation of his bail (at [13]).

In Matitwane the applicant’s request for a legal representative after he had earlier waived the right to have one, could hardly have served as the basis for an inference that he was delaying matters so that he could pay bail and evade trial. Kgoele J also concluded that ‘the decision taken was clearly biased because the audi alteram partem rule was not adhered to’ (at [13]).

In Matitwane the regional court magistrate not only acted outside the scope of s 68(1) but also failed to appreciate that cancellation of bail—like an application for bail—is a judicial process or inquiry that must be conducted impartially and in accordance with common-law and constitutional fair trial requirements.

Section 86: Court’s power to amend the charge

S v Thakeli & another 2018 (1) SACR 621 (SCA)

A charge (or indictment, as the case may be) must ‘correctly and adequately reflect all the necessary averments’ (S v Mashinini & another 2012 (1) SACR 604 (SCA) at [28]). Devlin The Criminal Prosecution in England (1958) at 133 has pointed out that the charge is ‘the . . . instrument giving legal precision’ to what the accused has to meet. An accurate charge is necessary to ensure that an accused is in terms of s 35(3)(a) of the Constitution ‘informed of the charge with sufficient detail to answer it’. This section, said Bosielo JA in S v Msimango 2018 (1) SACR 276 (SCA) at [16], ‘goes to the very heart of what a fair trial is’ and its principal aim is ‘to banish trial by ambush’.

However, the above background and principles must also be read in the context of s 86(1) of the Criminal Procedure Act 51 of 1977. This section authorises a court to make certain amendments to a charge ‘at any time before judgment, if it considers that the making of the relevant amendment will not prejudice the accused in his defence’. For an analysis of the permissible types of amendments and the measures required to avoid prejudice, see the discussion of s 86 in Commentary, sv General and also sv In what circumstances and at what stage (s 86(1)).

The most important limitation on the power of the court to act in terms of s 86(1) is that the amendment in question may not prejudice the accused in his defence (S v Kariko & another 1998 (2) SACR 531 (Nm); R v Baxter & another 1928 AD 430).

The issue of prejudice was addressed by a full bench of the Supreme Court of Appeal in S v Thakeli & another 2018 (1) SACR 621 (SCA). In this case the two appellants had appeared in the regional court on a charge of murder read with s 51(2) of the Criminal Law Amendment Act 105 of 1997 and Part II of Schedule 2 of the same Act. This meant that in the event of a conviction the minimum sentence, for a first offender, was 15 years’ imprisonment unless there were substantial and compelling circumstances justifying a lesser sentence. After hearing all the evidence, the regional court—relying on s 86 of Act 51 of 1977—amended the charge by replacing the reference to s 51(2) with a reference to s 51, that is, by simply deleting the reference to subsection ‘(2)’. The regional court’s view was that this amendment could cause no prejudice to the appellants (at [4]). The regional court then proceeded to deliver its judgment on the basis of the amended charge and convicted the appellants of murder in terms of s 51(1) of Act 105 of 1997, as read with Part I of Schedule 2, which carried with it the sentence of life imprisonment. However, each appellant was sentenced to 28 years’ imprisonment on the basis that there were substantial and compelling circumstances justifying a departure from the prescribed minimum sentence of life imprisonment.

On appeal to the Free State High Court (hereafter ‘the court a quo’) the appellants argued that the regional court had erred in amending the charge and that this amendment had caused undue prejudice. Their appeal was dismissed. However, the Supreme Court of Appeal granted the appellants leave to appeal to it on sentence only.

Writing for a unanimous full bench of the Supreme Court of Appeal, Saldulker JA pointed out that the court a quo had found that the amendment effected by the regional court ‘was akin to curing a ‘‘typing error’’ which did not go to the substance of the charge nor the sentencing regime’ (at [6]). The full bench rejected this finding. At [4] Saldulker JA observed that by changing ‘s 51(2)’ to ‘s 51’ the charge became ‘vague’ in that reference must be made to either s 51(1) or 51(2) so that there could be clarity as to the sentencing provisions which would govern the situation in the event of a conviction. Of fundamental concern was the principle that ‘an accused person must be apprised from the outset what charge he . . . has to meet, so that he . . . not only appreciates properly and in good time what the charges are that he . . . is facing but also the consequences’ (at [6]). Indeed, an accused’s constitutional right to a fair trial demands certainty and clarity so that important decisions regarding the conduct of his case can be made. The appellants, it was held at [7], might well have conducted their defence differently had they been informed from the outset that they were being charged with a murder within the ambit of s 51(1), which involved premeditation and acting in common purpose in killing the deceased. The regional court’s amendment of the charge also exposed the appellants to life imprisonment as the prescribed minimum sentence, as opposed to 15 years’ imprisonment, which would have been the prescribed minimum sentence had the amendment not been made. All this, said Saldulker JA at [7], ‘was done after all the evidence had been led and without affording the appellants any opportunity to address the court on the question of prejudice, and whether the amendment should be effected’. Denying the appellants ‘a full and proper opportunity’ to argue the matter ‘constituted a fundamental irregularity that infringed the fair trial rights of the appellants, and destroyed the validity of the amendment’ (at [7]).

Saldulker JA accordingly concluded that it was impossible to determine with certainty that the appellants had suffered no prejudice as a result of the amendment to the charge (at [7]). It was held that the court a quo had erred in dismissing the appeal against sentence (at [8]). The regional court’s sentence was set aside, and each accused was sentenced to 15 years’ imprisonment.

It should be noted that in Thakeli (supra) at [8] the Supreme Court of Appeal specifically pointed out that its decision should be distinguished from those cases in which it was held that an irregularity was not a fatal one vitiating the proceedings. Specific reference was made to S v Kolea 2013 (1) SACR 409 (SCA), where it was held that the accused in that case had known at the outset what charges he faced.

Judicial review of a decision by the National Director of Public Prosecutions to withdraw charges: Legality and rationality

Freedom Under Law (RF) NPC v National Director of Public Prosecutions & others 2018 (1) SACR 436 (GP)

In the above case—hereafter referred to as the ‘FUL case’—the applicant, relying on grounds of legality and rationality, sought a review and setting-aside of a prosecutorial decision to withdraw criminal charges against a member of the National Prosecuting Authority (NPA). In March 2015 Mr Nxasana, the then National Director of Public Prosecutions (NDPP), preferred charges of fraud and perjury against Ms Jiba, a Deputy National Director of Public Prosecutions (DNDPP). The charges against Ms Jiba arose as a result of her conduct in a High Court case and at a stage when she was acting NDPP: see Booysen v Acting National Director of Public Prosecutions & others 2014 (2) SACR 556 (KZD) at [1], [34] and [43], and see further the general background provided in cases like General Council of the Bar of South Africa v Jiba & others 2017 (1) SACR 47 (GP) and Corruption Watch (RF) NPC & another v President of the Republic of South Africa & others 2018 (1) SACR 317 (GP) at [32]–[33].

On 1 June 2015 Mr Abrahams replaced Mr Nxasana as NDPP (FUL case at [35]). Upon assuming office as NDPP, Mr Abrahams—acting without any representations from Ms Jiba—inquired from the then regional head of the Specialised Commercial Crime Unit (SCCU) what his view was regarding the charges against Ms Jiba. This inquiry was made despite the fact that Mr Nxasana, as predecessor of Mr Abrahams, had already taken the decision to prosecute, and despite the fact that all the parties, including the state, had informed the trial court of their readiness to proceed.

On 17 August 2015 the regional head of the SCCU delivered, in the words of Mothle and Tlhapi JJ, ‘an opinion to Abrahams . . . in which he recommended that the charges against Jiba be withdrawn’ (FUL case at [54], emphasis in the original). On 18 August 2015, two days before the commencement of Ms Jiba’s trial and one day after having received the recommendation by the regional head of the SCCU, Mr Abrahams held a press conference where he informed the media that the withdrawal of the charges against Ms Jiba was justified in terms of s 78 of the Prevention of Organised Crime Act 121 of 1998 (POCA). Mr Abrahams said that this section ‘clearly absolves Ms Jiba from any liability, including criminal prosecution, for having exercised the power in terms of the empowering provisions of POCA’ and that there were accordingly ‘no prospects of a successful prosecution’ (at [37]).

Section 78 of POCA provides that any person generally or specifically authorised to perform any function in terms of POCA, ‘shall not, in his . . . personal capacity, be liable for anything done in good faith’ under POCA. However, Mothle and Tlhapi JJ held that reliance on s 78 in withdrawing the charges was ‘a material error of law’ (at [44]) in that ‘when Jiba deposed to and filed an affidavit containing information misleading to the court, she was not performing any functions under POCA’ (at [42]). In a separate judgment Wright J also noted that s 78 formed an appreciable or significant part of Mr Abrahams’ reasoning in the withdrawal of the charges (at [163]). At [164] Wright J said that this reason ‘is wrong in law’ and that s 78 of POCA

‘is clearly intended to protect the person referred to, from liability for things done in good faith. The charges against Ms Jiba are criminal rather than civil in nature and allege intentional wrongdoing. It cannot have been the intention of the legislature that s 78 would provide immunity against the criminal charges preferred against Ms Jiba.’

On the basis of this error alone, concluded Wright J at [165], Mr Abrahams’ decision to withdraw the charges could not stand. In their joint judgment, Mothle and Tlhapi JJ also concluded that the reasons given for the withdrawal of the charges were based ‘on a material error of law, which falls short of the legality expected in a rational decision’ (at [59]).

The respondents in the FUL case also argued that the application for a review of the NDPP’s decision should not be entertained because the applicant had failed to exhaust internal remedies. The respondents relied on s 22(2)(c) of the NPA Act. This section determines, in accordance with s 179(5)(d) of the Constitution, that the NDPP may review a decision to prosecute or not to prosecute, after having consulted the relevant Director of Public Prosecutions (DPP) and after having taken representations—within the period specified by the NDPP—of the accused person, the complainant and any other person or party deemed relevant by the NDPP. The respondents accordingly argued that the applicant should at some earlier stage have approached the NDPP to review the decision not to prosecute. However, all three members of the court concluded that it was Mr Abrahams as NDPP who had taken the decision to withdraw the charges (at [53], [59] and [151]). The regional head of the SCCU made a recommendation to the NDPP and the latter took the decision (at [54]–[55]). It was accordingly held that there was no merit in the argument that the applicant had not exhausted his internal remedies by failing, first, to seek a review by the NDPP in terms of the NPA Act.

After pointing out that the NPA derives its power from s 179 of the Constitution and the provisions of the NPA Act and that the mandate of the NPA is to institute criminal proceedings, Mothle and Tlhapi JJ concluded as follows (at [60]–[61], emphasis added):

‘Once a decision has been made to prosecute, the NPA may review that decision in a manner prescribed by s 179(5)(d) of the Constitution. The exercise of that power must not be manifestly at odds with the purpose for which the power was conferred . . . . Thus, this court concludes that the means selected to withdraw the charges against Jiba are not rationally related to the NPA’s objectives which it sought to achieve. Consequently, it is the finding of this court that the decision to withdraw charges against Jiba was irrational and is set aside.’

At [62] it was observed that the consequence of the review and setting-aside of the NDPP’s decision was that the charges and proceedings were automatically reinstated. In this regard reliance was placed on what Brand JA said in National Director of Public Prosecutions & others v Freedom Under Law 2014 (2) SACR 107 (SCA) at [51]. See also the remarks made by Wright J at [166] in the FUL case. For a discussion of case law dealing with legality and rationality as grounds for reviewing prosecutorial decisions, see the introduction to Chapter 1 ‘Prosecuting Authority’ in Commentary, sv Review of prosecuting authority’s decision to withdraw charges, and the validity of a mandatory interdict to prosecute and sv Judicial review of the decision not to prosecute: Prosecutorial integrity and rationality.

Sections 164, 192 and 194: Separating the inquiries into competence to testify and competence to take the oath or affirmation

The issue in S v Macinezela [2018] ZASCA 32 (unreported, SCA case no 550/2017, 26 March 2018), where the appellant appealed against his conviction for rape in terms of s 3 of the Criminal Law (Sexual Offences and Related Matters) Amendment Act 32 of 2007), was the admissibility of the evidence of the complainant, who was alleged to be ‘mentally unstable’ or ‘mentally unsound’.

Before the trial commenced, the prosecutor had asked that the charge sheet be amended to reflect that the complainant was ‘not mentally stable’. This was done. At the trial the complainant and her uncle testified on behalf of the State. The uncle testified that the complainant was not mentally sound. The magistrate rejected the appellant’s case that the complainant had consented to sexual intercourse and convicted the appellant on the ground that he had been aware that the complainant was ‘not mentally sound’, and that, even if ‘she may have consented to sexual intercourse such consent was not recognised by virtue of her mental illness or mental retardness’. The magistrate also remarked that he had also observed that the complainant was ‘not completely sane’, pointing to the fact that she appeared not to know her age and had read incorrectly the date on which she was alleged to have sent a text message to the appellant. He accepted the evidence of the complainant and rejected the evidence of the appellant that she had consented to intercourse.

After an unsuccessful appeal to the High Court, the appellant appealed to the Supreme Court of Appeal on two broad grounds. The first concerned the manner in which the complainant’s mental condition was introduced into the proceedings: the appellant was not given an opportunity to make submissions on the proposed amendment to the charge sheet and there was no ruling by the court on the proposed amendment. It was argued for the appellant that the trial court’s finding that the charge sheet had been duly amended and that the complainant’s mental disability had been established was wrong. The second was that the trial court and the High Court had failed properly to consider whether the complainant was in fact ‘mentally disabled’ as envisaged in the Criminal Procedure Act.

Dambuza JA, with whom the other judges agreed, pointed to a ‘primary issue’ that was ‘regrettably’ not identified by the magistrate, the prosecutor or the defence: the fact that proper procedure was not followed when it became apparent, right at the outset, that the complainant might not understand the nature and import of the oath or affirmation as set out in s 164 of the Criminal Procedure Act. Further, it was not apparent that the distinction between the competence of the complainant to testify and her competence to take the oath or make the affirmation had been understood.

It was trite, said Dambuza JA (at [13]), that the ‘principal method of adducing evidence in a trial is by oral evidence of a competent witness’ (see, said the judge, Zeffertt & Paizes The South African Law of Evidence 2 ed (2009) at 805). The general rule is that everyone is presumed to be a competent and compellable witness (see s 192), subject to certain specified exceptions. One such exception is contained in s 194, which provides that ‘[n]o person appearing or proved to be afflicted with mental illness or to be labouring under any imbecility of mind due to intoxication or drugs or the like, and who is thereby deprived of the proper use of his reason, shall be competent to give evidence while so afflicted or disabled’. In S v Katoo 2005 (1) SACR 522 (SCA) at [11] the Supreme Court of Appeal stressed that the two requirements set out in s 194 ‘must collectively be satisfied before a witness can be disqualified from testifying on the basis of incompetence’.

In Macinezela, said Dambuza JA (at [15]), there was no indication from the record whether, apart from the allegation by the prosecutor at the start of the trial, the magistrate had formed a view in respect of the complainant’s mental capacity. The application to amend the charge sheet ‘clearly called for vigilance in considering the proper approach to her evidence’. A court confronted with a potentially mentally ill witness may opt to seek expert medical evidence on the effect of that illness on the cognitive faculties of the witness or it may allow the witness to testify in order to assess his or her competency. If the court follows the latter path, as the magistrate did in this case, the provisions of ss 162, 163 and 164 of the Act come into play (see Commentary for a discussion of these sections). Section 164(1) applies when a court is dealing with the admission of the evidence of any person who is found not to understand the nature and import of the oath or affirmation. Such a witness must, instead of being sworn in or affirmed, be admonished by the judicial officer to speak the truth.

In S v Matshivha 2014 (1) SACR 29 (SCA) at [10]–[11] the court set out the obligations of a court in such circumstances (see further Commentary on s 164 sv General). It stressed that, for s 164(1) to be triggered, there had to be a finding that the witness did not understand the nature and import of the oath, and this finding had to be preceded by ‘some sort of enquiry by the judicial officer, to establish whether the witness understands the nature and import of the oath’. If the judicial officer should find after such an inquiry that the witness does not possess the required capacity to understand the nature and import of the oath, he or she should, said Zondi AJA in Matshivha, establish whether the witness can distinguish between truth and lies. If so, the witness should be admonished to speak the truth.

Although the remarks of Zondi AJA were made in respect of child witnesses, Dambuza JA correctly observed that they were equally applicable to witnesses who were mentally ill, since s 164(1) applies to ‘[a]ny person, who is found not to understand the nature and import of the oath or affirmation’ (emphasis added), irrespective of whether that failure to understand is the result of youth, defective education or mental illness (see S v Motsisi [2012] ZASCA 59 (unreported, SCA case no 513/11, 2 April 2012)).

Dambuza JA added this gloss on the two inquiries identified by Zondi AJA after citing the views expressed in Commentary (notes to s 164 sv General):

‘An inquiry into whether a potential witness can distinguish between truth and falsity goes to whether the witness is competent in the first place. On the other hand, a question directed to a witness on whether he or she understands the nature and import of the oath and affirmation goes to whether the witness should be caused to take the oath or affirmation, or should be admonished to speak the truth in terms of s 164(1).’

In this case, said Dambuza JA (at [20]), the oath or affirmation could not, in the circumstances, be administered in the ordinary course. At the very least, an inquiry in terms of s 164 should have been conducted. This did not happen and, as a result, the magistrate did not inquire further into whether the complainant could even distinguish between truth and falsehood. This omission was fatal to the prosecutor’s case, and the appellant’s appeal was accordingly upheld.

The insistence of Dambuza JA on separating (at least notionally) the two inquiries is welcomed. There have been a number of cases (see Commentary supra) where the courts have either failed to appreciate that there are two separate inquiries at all, or conflated them in a manner that makes it difficult to evaluate whether both have been addressed adequately or properly. Such inaccuracies should not be tolerated, and the judgments in Matshivha and Macinezela suggest strongly that they will not be. Dambuza JA was at pains to point to the ‘injustice that can be suffered by both complainants and accused as a result of failure by courts to properly ascertain whether a witness is able to distinguish between truth and falsehood’. Understanding that there are two discrete inquiries, with two different sets of concerns and conditions, is a sound starting point to avoid falling into the kind of trap that leads to such injustice.

Section 170: Need to respect constitutional rights of accused at summary inquiry where accused fails to appear after an adjournment; unconstitutional reverse onus in s 170(2)

Cooper v District Magistrate, Cape Town 2018 (1) SACR 369 (WCC)

Section 170(1) of the Criminal Procedure Act provides that ‘[a]n accused at criminal proceedings who is not in custody and who has not been released on bail, and who fails to appear at the place and on the date and at the time to which such proceedings may be adjourned or who fails to remain in attendance at such proceedings as so adjourned, shall be guilty of an offence and liable to the punishment prescribed under subsection (2)’. Subsection (2) provides that ‘[t]he court may, if satisfied that an accused referred to in subsection (1) has failed to appear at the place and on the date and at the time to which the proceedings in question were adjourned or has failed to remain in attendance at such proceedings as so adjourned, issue a warrant for his arrest and, when he is brought before the court, in a summary manner enquire into his failure so to appear or so to remain in attendance and, unless the accused satisfies the court that his failure was not due to fault on his part, convict him of the offence referred to in subsection (1) and sentence him to a fine not exceeding R300 or to imprisonment for a period not exceeding three months’.

The nature of the inquiry held under s 170(2) and the duty of the judicial officer to give effect to the fundamental constitutional rights of the accused in such an inquiry were explained and explored in Cooper’s case. Andrews AJ (with whom Baartman J agreed) cited with approval the view expressed in Commentary in the notes to this section that ‘[w]here the court holds a summary inquiry in terms of sub-s (2) into a failure to attend, considerations of justice and common sense demand that the court must inform the accused that there is an onus on him to tender a reasonable excuse—in the unlikely event, that is, that this reverse onus passes constitutional muster (see S v Ngubeni [2009] 1 All SA 185 (T) at [15]). The accused must also be advised that he may call witnesses and give evidence himself (S v Bkenlene 1983 (1) SA 515 (O); see also S v Du Plessis 1970 (2) SA 562 (E) at 564–5)’.

In Cooper, it was clear that significant errors had been made by the magistrate. First, the warrant of arrest was authorised in terms of s 55, which applies only where a person appears in court on a summons. The applicant had been warned to appear on the material date, and so the warrant of arrest should have been authorised in terms of s 170. The applicant had thus incorrectly been found guilty of contravening the provisions of s 55 instead of s 170(1).

Second, the applicant had not been informed of the charge, the nature of the proceedings or of his constitutional rights. The magistrate did not even make it clear to the applicant that he was being charged with a criminal offence or what the commensurate penalties upon conviction might be.

Third, the magistrate failed to afford the applicant a fair opportunity to prepare for the hearing, to present a defence or to call witnesses.

Fourth, the magistrate failed to explain the position regarding the onus of proof. The section purports to impose a reverse onus on an accused in such circumstances. However, the magistrate was held (at [20]) to have ‘materially misdirected himself by applying the reverse onus’ (emphasis added). Andrews AJ drew attention to the similarities between the inquiry directed by s 170(2) and that set out in s 72(4), and pointed to the finding by the Constitutional Court in S v Singo 2002 (2) SACR 160 (CC) that the imposition of a reverse onus in s 72(4) was unconstitutional. The Constitutional Court in that case held that s 72(4) had to be read as though the words ‘there is a reasonable possibility that’ appeared between the words ‘that’ and ‘his failure’. Although the court in Cooper did not go this far, its rejection of the reverse onus in s 170(2) would suggest that this interpretation would have met with its approval. Such an approach would, it is submitted, be appropriate.

Fifth, the applicant was not treated with dignity or respect. The ‘robust manner in which the inquiry was conducted’ suggested to the court that the magistrate was ‘not alive to these basic and fundamental human rights firmly entrenched in South Africa’s constitutional democracy’ (at [19]). The applicant had presented the court with a medical certificate to account for his failure to appear. If the magistrate had any reservations as to the veracity or authenticity of that certificate, its author should have been called as a witness to clarify any aspects for the benefit of the court.

Sixth, although the applicant had the benefit of legal representation at the inquiry, his legal representative was ignored and his presence not even acknowledged during the inquiry.

These irregularities and misdirections led to the major concern expressed by Andrews AJ (at [20]) that the magistrate, who was ‘called upon in exercise of his judicial function to administer the law without fear, favour or prejudice, had a complete disregard for the rights of the accused’, who had suffered prejudice as a result. The proceedings were thus set aside.

Section 216: Hearsay evidence: s 3(1)(a) of the Law of Evidence Amendment Act 1988: When can an unrepresented accused be said to have ‘agreed’ to the admission of hearsay evidence tendered against him or her?

S v Flobela (unreported, WCC case no 17258, 12 March 2018)

Section 3(1)(a) of the Law of Evidence Amendment Act 45 of 1988 provides that hearsay evidence will be admissible if ‘each party against whom the evidence is to be adduced agrees to the admission thereof as evidence at such proceedings’. Where, however, the party against whom the hearsay evidence is adduced is an accused in criminal proceedings, and, in particular, where the accused is unrepresented, great care has to be exercised by a judicial officer who is asked to receive the evidence in terms of s 3(1)(a).

It has been submitted in Commentary in the notes to s 216 (sv Section 3(1)(a) of Act 45 of 1988) that, since consent may, in such circumstances, sometimes be given without a full appreciation by the accused of the nature and extent of the prejudice that might ensue if the evidence is admitted, it would be a ‘salutary practice’, before the evidence is received, for the courts to warn the accused of this danger and to insist on his or her express consent and not to construe a failure to object to its reception as implied or tacit consent. It is pointed out in Commentary that there is a duty on a judicial officer, where he or she becomes aware that a witness is—either deliberately or out of ignorance—giving hearsay evidence, to explain the rule against hearsay to the accused, especially where he or she is not represented (see S v Zimmerie en ’n ander 1989 (3) SA 484 (C) at 492F–G, S v Congola 2002 (2) SACR 383 (T) at 386c–e and S v Ngwani 1990 (1) SACR 449 (N)).

These submissions were accepted by Henney J (with Samela J concurring) in S v Flobela (supra at [29]), where he found that the magistrate had fallen far short of these requirements in admitting hearsay evidence in terms of s 3(1)(a). The accused had been charged with housebreaking with intent to steal and theft. Crucial evidence on the count of housebreaking was a statement made by the deceased complainant, which the prosecution sought to have admitted under s 3(1). The magistrate warned the accused that the State was tendering against him the hearsay evidence of the deceased complainant and explained that his consent was being asked for the handing in of the statement. In his answer, the accused said that he would like to continue with his case because he had ‘been in jail for a long time’. The magistrate twice asked him if he had any objection to the handing in of the statement and he answered, on each occasion, that he did ‘not have a problem with that’ (at [19]).

The magistrate then admitted the evidence, although it was not clear on what basis, since he said to the accused that, if he agreed to the handing in of the statement, there would be no need for the prosecutor to bring an application in terms of the provisions of the Law of Evidence Amendment Act to have it so admitted. It was thus not clear to the court whether the magistrate wanted the accused to make a formal admission in terms of s 220 of the Criminal Procedure Act or whether the statement was being received under s 3(1)(a) of Act 45 of 1988. If his purpose was to extract an admission in terms of s 220, said Henney J, he failed to comply with the provisions of that section because the admission was not formally made and, further, was not formally recorded. The magistrate also ‘did not explain the content and the consequences that would follow if the statement [was] admitted against the accused’ (at [27]). Even greater care had to be exercised where the accused lacked legal representation, and it was a principle of long standing that the judicial officer in such cases had to satisfy himself or herself that the accused’s decision to make the formal admission had been ‘taken with the full understanding of its meaning and effect’ (at [28]); see S v Daniëls en ’n ander 1983 (3) SA 275 (A) and S v Mavundla 1976 (4) SA 731 (N)).

The court then turned to s 3(1)(a) and held that the same principles should be applicable where an undefended accused is asked whether he or she agrees to hearsay evidence being admitted against him or her in terms of that paragraph. It was held in Ngwani (supra at 449) that an unrepresented accused has to have the effect of the relevant paragraph of s 3(1) (in Ngwani’s case it was s 3(1)(c)) fully explained to him. In fact, said Henney J, the court should go even further. It should make an assessment of the facts the accused places in issue; consider the hearsay evidence the State wishes to present against the accused; and consider further whether that evidence would serve to prove those facts which the accused places in issue and which the State bears the onus of proving. The court should then ‘explain to the accused . . . the nature of the hearsay evidence, the purpose for which the prosecutor wants to tender such evidence and the prejudice and consequences that might flow from the admission of such evidence’ (at [31]). It must then ‘inform the accused that there is no obligation upon him or her to agree or to admit such evidence, because the onus rests on the State to prove such evidence beyond reasonable doubt’.

The magistrate did not in any way explain to the accused the consequences of his decision. He did not make clear to him that he was not obliged to agree to the admission of the evidence against him or that the onus of proving the facts contained in the deceased’s statement was on the State. Further, in view of his defence, which was a denial of both the housebreaking and the theft, the magistrate should have explained to him that the reception of the evidence would be prejudicial to his case and harmful to his defence. His failure to do all of the above infringed the accused’s right to a fair trial in terms of s 35(3) of the Constitution. The accused agreed to the admission of the evidence only because he had been in custody for a long time and wanted to finalise the matter. This may have expedited his right to a speedy trial, but it adversely affected his right to a fair trial.

The magistrate thus clearly ‘committed a serious misdirection by admitting the hearsay evidence on the basis which he did’ (at [35]).

The decision in Flobela is undoubtedly correct. But it gives one pause for thought. Why would any unrepresented accused in circumstances such as those in Flobela, who has pleaded not guilty and who makes no relevant formal admissions, agree to the admission against him of hearsay evidence that fundamentally undermines his defence? Would not every represented accused ordinarily endorse his or her counsel’s flat refusal to agree to its admission? Would the prosecution, in the ordinary course in such circumstances, even bother to ask whether the accused would be prepared to agree to the admission of the highly damaging hearsay evidence?

It appears almost as if prosecutors are preying on the ignorance and naivety of unrepresented accused in seeking to have them agree to the reception of evidence of this kind. Would it be going too far in favour of the accused in such cases if the judicial officer were to make it quite plain to the accused that, if they were represented, counsel would undoubtedly not agree to the reception of the evidence as it would be very bad for their case if it were admitted? After all, if the reception of the evidence is in the interests of justice, there is nothing to stop the prosecution from having the evidence received in terms of s 3(1)(c) of the 1988 Act. If the State does not rely on paragraph (c) but only on paragraph (a), it is not fanciful to infer that the State has little confidence in its ability to convince the court that the interests of justice would be well served by admitting the evidence. If its only chance of getting the evidence received is to trade on the ignorance of the accused, it would seem that judicial officers would be well within their rights to be at least as protective as Henney J was in Flobela.


 (ii) Sentencing

The meaning of ‘planned or premeditated’ in s 51(1) read with Part I of Schedule 2 of the Criminal Law Amendment Act 105 of 1997

S v Jordaan & others 2018 (1) SACR 522 (WCC)

In Jordaan (supra) the indictment in respect of one of the counts of murder relied on the minimum sentence legislation referred to above. This meant that if convicted of ‘planned or premeditated’ murder, the prescribed sentence had to be life imprisonment unless there were substantial and compelling circumstances justifying a lesser sentence. See also the discussion of S v Taunyane 2018 (1) SACR 163 (GJ) elsewhere in this edition of CJR.

At [123] in Jordaan Binns-Ward J remarked that ‘[t]he import of the term ‘‘planned or premeditated’’ is inherently imprecise’. In this regard reference was made to Stevens (2015) 28 SACJ 347. It should be noted that Act 105 of 1997 (the minimum sentence legislation) does not provide any definition of a ‘planned or premeditated’ murder. What was clear though, said Binns-Ward J at [123], was that it was unlikely that the legislature had in mind that ‘planned or premeditated’ should be interpreted as equivalent to ‘direct intention’. Murder with dolus directus can be committed without involving planning or premeditation. See also S v Taunyane (supra) at [30].

In Jordaan at [124] Binns-Ward J referred to the decision in S v Raath 2009 (2) SACR 46 (C) where a full bench (per Bozalek J) had pointed out at [16] that ‘only an examination of all the circumstances surrounding any particular murder, including not least the accused’s state of mind, will allow one to arrive at a conclusion as to whether a particular murder is ‘‘planned or premeditated’’’. The Raath decision indicated that the time lapse between forming the intention to murder and the execution of that intention, could be of cardinal importance but was not necessarily decisive in some cases. See further the discussion of Raath and S v Kekana [2014] ZASCA 158 (unreported, SCA case no 629/2013, 1 October 2014) in the notes under the repealed s 277 in Commentary, sv Murder: Absence or presence of premeditation.

In Jordaan Binns-Ward J relied on Raath in support of his statement that the term ‘planned or premeditated’ is ‘inherently imprecise’ (at [124]). In Raath there was also no pertinent decision that ‘planned’ and ‘premeditated’ were two different concepts. Indeed, in S v PM 2014 (2) SACR 481 (GP) at [35] Thulare AJ said that the full bench in Raath had treated ‘planned or premeditated’ as a ‘concept, meaning it as one idea’. According to Thulare AJ this approach was wrong, because ‘the use of the word ‘‘or’’ indicates that the legislature did not favour a composite description’ (S v PM (supra) at [36]). Thulare AJ also explained at [36] what he understood the meaning of ‘planned’ to be, as opposed to ‘premeditated’.

However, Binns-Ward J found the ‘dichotomous connotation’ that Thulare AJ gave to the phrase ‘planned or premeditated’, unconvincing. At [127] he said that

‘if regard is had to the object of the provision—ie the creation of a criterion for the attraction of the mandatory sentence of life imprisonment if substantial and compelling circumstances to deviate from the prescribed sentence are not present—no evident statutory purpose would be served in making the distinction. It is a sterile exercise to seek the meaning of a word without regard to the context in which it has been employed. When the word lies in a statute, the evident scope and object of the instrument are critical contextual considerations.’

In the view of Binns-Ward J, the clear object of the provision actually supported an interpretation which requires that ‘or’ should be read ‘in its frequently acknowledged possible sense of ‘‘and/or’’’ (at [128]). This interpretation would mean that any possible basis for a distinction is rendered ‘immaterial for practical purposes’ (at [128]). Relying on the Oxford Dictionary of English, Binns-Ward J also pointed out that ‘the concept of planning’ fell within the meaning of ‘premeditation’ (at [128]).

In Jordaan at [129]) reference was made to S v Montsho [2015] ZASCA 187 (unreported, SCA case no 20572/2014, 27 November 2015). In this case—which was an appeal from S v PM (supra)—the Supreme Court of Appeal held that on the facts of the case there could be no advantage in formulating a general test or definition to determine whether the phrase ‘planned or premeditated’ denoted a single concept. Petse JA observed that the presence or absence of a plan or premeditation could ‘properly be determined on a case by case basis’ (Montsho at [14]). According to Binns-Ward J in Jordaan at [129], Montsho cannot be reconciled with ‘the notion of any practical dichotomy between planning and premeditation’. He pointed out that Montsho did in fact rely on S v Kekana (supra) ‘which accepted planning as indicative of premeditation’ (at [129]).

In Jordaan the conclusion was reached that in respect of count 7 the prosecution had failed to discharge the onus of proving that the shooting was planned or premeditated. There was ‘no direct proof of planning or premeditation’ and it was ‘reasonably possible that the accused could have decided on the assault virtually on the spur of the moment’ (at [130]).

It is submitted that, in the final analysis, the words ‘planned or premeditated’ refer to a reprehensible state of mind which—from a sentencing point of view—increases the moral blameworthiness of the offender: a murder which is plotted (planned and/or premeditated) in terms of means, opportunity and execution, is more reprehensible than a murder committed impulsively. This is what the minimum legislation sought to achieve and, in terms of the decision in Jordaan, any attempt to distinguish between ‘planned or premeditated’ would serve no practical purpose.

Procedural and evidential issues concerning ‘planned or premeditated’ murder as referred to in Part I of Schedule 2 of the Criminal Law Amendment Act 105 of 1997

S v Taunyane 2018 (1) SACR 163 (GJ)

An accused convicted of ‘planned or premeditated’ murder must be sentenced to life imprisonment unless there are substantial and compelling circumstances justifying a departure from this prescribed sentence: see s 51(1) of the Criminal Law Amendment Act 105 of 1997, as read with s 51(3) and (6) and Part I of Schedule 2 of the same Act. For an overview of some important cases dealing with ‘planned or premeditated’ murder as identified in Part I of Schedule 2 to Act 105 of 1997, see the discussion under the repealed s 277 of the Criminal Procedure Act 51 of 1977 in Commentary, sv Murder: Absence or presence of premeditation

In S v Taunyane (supra) a full bench of the Gauteng Local Division addressed some procedural and evidential matters concerning planned or premeditated murder in the context of Act 105 of 1997.

In Taunyane the appellant was charged with and convicted of ‘murder read with s 51(1) of Act 105 of 1997’ (at [4]). In his judgment on the merits, the trial judge made no mention of planning or premeditation (at [2] and [9]). The conclusion that ‘the killing . . . was premeditated’ was reached only in the judgment on sentence (at [3]). This conclusion led to the imposition of a sentence of life imprisonment as required by Act 105 of 1997.

The first issue in Taunyane concerned the indictment. Part I of Schedule 2 provides for at least eight situations where murder would in the absence of substantial and compelling circumstances attract life imprisonment. A murder which is ‘planned or premeditated’ is one of these: see paragraph (a) of Part I. The summary of substantial facts which accompanied the indictment made no specific mention of the prosecution’s intention to rely on ‘planned or premeditated’ murder as referred to in paragraph (a) of Part I of Schedule 2. Satchwell J (Van Oosten and Masipa JJ concurring) pointed out that—in the absence of further particulars requested by and furnished to the defence—‘there was no indication as to which portion of part I [was] alleged to be of application’ (at [6]). At [7] Satchwell J stated that she was ‘somewhat concerned’ that it should be expected of an accused and his counsel ‘to infer which portion of part I of sch 2 and, therefore, of s 51(1), is the subject-matter of a prosecution’. Having regard to all the circumstances, she was nevertheless satisfied that the appellant had not been prejudiced in conducting his defence: he had legal representation and the indictment contained sufficient details in the sense that ‘none of the other circumstances of murder set out in part I of sch 2 could ever have been under consideration’ (at [7]).

The second issue in Taunyane related to the trial judge’s failure to deal with the question of planning or premeditation in the course of his judgment on conviction. The sentencing stage, said Satchwell J at [9], ‘was not the time when the matter should have been considered or argued or adjudicated for the first time’. At [10] she pointed out that s 51(1) of Act 105 of 1997 requires the court to have convicted an accused of an offence referred to in Part I of Schedule 2: ‘It is only when such a conviction is determined and full reasons given in respect of an identified offence ‘‘referred to in Part I of Schedule [2]’’, that an appropriate sentence can be handed down’. To put the matter differently: the jurisdictional fact requiring the court to impose life imprisonment was absent. See also generally S v Ndlovu 2017 (2) SACR 305 (CC) which is discussed in Commentary under the repealed s 277 of Act 51 of 1977, sv Sentencing jurisdiction and minimum sentence legislation.

In Taunyane Satchwell J pointed out that Act 105 of 1997 did not create ‘a new class of murder’, but does require a court to find at the time of conviction whether or not the relevant murder was committed within one of the categories or situations set out in Part I of Schedule 2 (at [12]). She observed that for purposes of this finding at the conviction stage, the court was required to apply the criminal standard of proof beyond reasonable doubt ‘as to the existence or otherwise of planning or premeditation’ (at [13]). It was accordingly held that the trial judge had misdirected himself ‘in setting out a conviction of planned or premeditated murder in the judgment on sentence’ when the matter was not addressed and adjudicated in the judgment on conviction (at [14]). This misdirection, it was found at [15], caused prejudice because there was no indication ‘of what standard of proof was applied and how such standard was discharged’ when the trial judge found that premeditation was present. At [38] Satchwell J held:

‘In the result I would uphold the appeal against the conviction of murder as contemplated in s 51(1), ie murder committed with planning or premeditation, as set out in part I of sch 2 to Act 105 of 1997. I would qualify the conviction of ‘‘murder’’ with a finding of ‘‘murder with dolus directus’’. I would set aside the finding in the judgment on sentence of a murder which was premeditated.’

In coming to her conclusion that there was no planned or premeditated murder, Satchwell J noted that whilst it is of importance to consider the time which may have elapsed between an accused’s decision to murder and the murder itself, it should also be kept in mind that the period of time cannot provide a ready-made answer and that time was not the only consideration (at [28]). In this regard reference was made to S v Raath 2009 (2) SACR 46 (C) and S v PM 2014 (2) SACR 481 (GP). These two cases are discussed in detail in Chapter 28 of Commentary, sv Murder: Absence or presence of premeditation where it is also pointed out that S v PM (supra) later went on appeal to the Supreme Court of Appeal where it was referenced as S v Montsho [2015] ZASCA 187 (unreported, SCA case no 20572/2014, 27 November 2015). In Montsho at [14] Petse JA said that the presence or absence of planning or premeditation ‘can properly be determined on a case by case basis’. And that, indeed, is what Satchwell J did in Taunyane (supra): it was noted that ‘motive . . . alone’ is insufficient to constitute premeditation (at [31(b)]); that firing four shots into the body of the deceased while he was lying on the ground demonstrated dolus directus and not premeditation (at [30]); that the fact that the appellant had after the event phoned his wife to tell her that after the killing he had intended to burn the body of the deceased, ‘may or may not be indicative of premeditation or of machismo or braggadocio . . . [and] . . . is hardly conclusive of planning or premeditation’ (at [36]).

Relying on some criteria or guidelines identified in Raath (supra) and PM (supra), Satchwell J formulated the following broad test or approach (at [30]):

‘In deciding whether or not appellant killed the deceased in circumstances where such killing was planned or premeditated, the test is not whether there was an intention to kill. That had already been dealt with in finding that the killing was an act of murder. The question now is whether or not appellant ‘‘weighed-up’’ his proposed conduct either on a thought-out basis or an arranged-in-advance basis (as set out in Raath supra . . . para 16), or whether or not appellant ‘‘rationally consider[ed] the timing or method’’ of the killing, or prepared a ‘‘scheme or design in advance’’ for achieving his goal of killing the deceased (as set out in PM supra [27]).’

The conclusion was reached that on the available evidence it could not be said that there was proof beyond reasonable doubt of ‘a deliberate course of action which was so planned as to increase the likelihood of success or enable evasion of apprehension thereafter’ (at [32]). The full bench decision in Taunyane makes a solid contribution to our growing jurisprudence concerning the interpretation of the words ‘planned or premeditated’. In this case the sentence of life imprisonment was replaced by a sentence of 18 years’ imprisonment. 

In S v Mokgalaka 2017 (2) SACR 159 (GJ)—which was decided some four months after Taunyane—Kuny AJ (Mashile J concurring) also held that the burden was on the prosecution to prove beyond reasonable doubt that a murder was planned or premeditated (at [33]). See further the discussion of S v Jordaan 2018 (1) SACR 522 (WCC) elsewhere in this edition of Criminal Justice Review. See also the comments on Mokgalaka by A van der Merwe (2017) 3 SACJ 422 at 430–2.

Dealing in drugs and proof of value for purposes of Part II of Schedule 2 of the Criminal Law Amendment Act 105 of 1997

S v Klaas 2018 (1) SACR 643 (CC)

In S v Klaas (supra) the trial court had convicted the applicant (hereafter ‘the accused’) of two contraventions of the Drugs and Drug Trafficking Act 140 of 1992 (hereafter ‘the Drugs Act’): the unlawful manufacturing of drugs (count one) and dealing in drugs (count two). In respect of count one he was sentenced to five years’ imprisonment. This term of imprisonment was ordered to run concurrently with the prescribed minimum sentence of 15 years’ imprisonment which—in the absence of substantial and compelling circumstances—was imposed in respect of count two, namely dealing in a drug (‘dependence producing substance’) to the value of more than R50 000. See s 51(2) of the Criminal Law Amendment Act 105 of 1997 as read with Part II of Schedule 2 of the same Act and s 13(f) of the Drugs Act. These provisions are hereafter referred to as the ‘minimum sentence provisions’.

All the attempts of the accused to obtain leave to appeal against his conviction and sentence were unsuccessful (at [11]–[13]) except for his application to the Constitutional Court which granted him leave to appeal against sentence on the basis that the matter raised ‘the constitutional issue of the right to a fair trial’ (at [21]) in the sense that there might not have been ‘a fair trial during the sentencing stage’ (at [27]). In granting leave to appeal, the Constitutional Court also noted that the accused had ‘reasonable prospects of success in relation to sentence’ (at [21]).

Writing for a unanimous Constitutional Court, Mhlantla J pointed out that whilst the relevant minimum sentence provisions had been brought to the accused’s attention at the plea stage, the prosecution had in the course of its case failed to prove the market value of the drugs which formed the subject-matter on count two. Proof of such value was necessary ‘before conviction in order for the minimum sentence legislation to be applicable’ (at [28]). In its judgment on the merits the trial court had stated that the accused was charged with contravening the Drugs Act. However, the judgment made no mention of the minimum sentence provisions (at [28]). After conviction neither defence counsel nor the prosecutor had addressed the trial court on the applicability of the minimum sentence provisions and whether there were substantial and compelling circumstances justifying the imposition of a lesser sentence than the prescribed minimum sentence (at [29]). In a brief argument on sentence the prosecutor asked for a sentence of direct imprisonment, whereas counsel for the defence submitted that a long term of imprisonment wholly suspended was appropriate. This, said Mhlantla J at [29], was an indication that the prosecutor as well as defence counsel believed that the minimum sentence provisions were inapplicable (at [29]). The parties were also not invited by the trial court to address it on the applicability of the minimum sentence provisions (at [29]). However, in its judgment on sentence, the trial court had declared for the first time that the minimum sentence provisions were applicable in respect of count two. After having come to the conclusion that there were no substantial and compelling circumstances, the trial court imposed the prescribed minimum sentence of fifteen years on count two (at [30]).

Relying on S v Legoa 2003 (1) SACR 13 (SCA), Mhlantla J pinpointed a fundamental flaw in the trial proceedings in Klaas: the minimum sentence provisions could have been activated only upon proof by the prosecution prior to conviction that the value of the drugs in question exceeded R50 000 (at [31]). Our courts have made it clear that ‘value’ in the minimum sentence provisions means ‘market value’ (Legoa (supra) at [11]; S v Sithole 2005 (2) SACR 504 (SCA) at [12]; S v Umeh 2015 (2) SACR 395 (WCC) at [61]). This approach was accepted by Mhlantla J in Klaas (at [31]–[33]). At [34] it was pointed out that in the trial court’s judgment on the merits no reference was made to the market value of the drugs in question. The trial record indicated that the investigating officer testified that he did not know the market value. However, he estimated that the value of the drugs, chemicals and all items seized was in the region of R18 000 000 (at [8] and [34]). According to the investigating officer experts could provide ‘the actual value of the drugs found’ (at [8]). But no such expert evidence was adduced by the prosecution (at [34]). At [35] Mhlantla J held that the prosecution

‘had a duty to prove the value of the drugs before conviction stage. It omitted to do so or adduce evidence in that regard. The main reason that the trial court could not apply the minimum sentence was not because the applicant had not been informed of the minimum sentence, but because the market value of the drugs had not been proven before conviction, as required by the Criminal Law Amendment Act. In order for the minimum sentence set out in part II of schedule 2 to the Criminal Law Amendment Act to apply, the state would have had to prove that the value of the drugs was greater than R50 000. An estimate of the value is not sufficient. Therefore, the jurisdictional fact entitling the trial court to impose the minimum sentence was absent. This results in an irregularity insofar as sentence is concerned.’

It was accordingly held that the trial court had misdirected itself in imposing the minimum sentence. This meant that the Constitutional Court was entitled to interfere with sentence. Sentence was considered afresh. The trial court’s sentence of 15 years’ imprisonment on count two was set aside and replaced by a period of 12 years’ imprisonment, backdated to 10 December 2013 which was the date of the trial court’s sentence (at [11] and [47]). The long term of imprisonment imposed by the Constitutional Court was based on our case law and explained as follows by Mhlantla J (at [46], emphasis added).

‘These cases make it clear that our courts have deemed sentences of imprisonment of 10 years or longer to be appropriate for convictions relating to the manufacturing of and dealing in dangerous drugs. In my view, the applicant’s personal circumstances pale into relative insignificance when regard is had to the seriousness of the offences and the need to protect the public. It is common cause that the value and weight of the drugs were not proved, but 2 920 tablets were seized. It has to be accepted that the applicant did not only deal in drugs but also manufactured them. Even without evidence of the precise value of the drugs seized, this quantity of tablets had the potential to affect the lives of many people.’

The 12 years’ imprisonment imposed in respect of count two was ordered to run concurrently with the five years’ imprisonment which the trial court had imposed in respect of count one. This concurrency order was considered to be in the interests of justice given ‘the advanced age of the accused’, namely 58. The effective sentence was therefore 12 years’ imprisonment. On advanced age as a possible mitigating factor see the discussion in paragraph 12 under the repealed s 277 in Commentary, sv Imposing imprisonment in the absence of minimum sentence legislation—general principles. 



 Table of Cases

Click here to see the full Table of Cases of the Criminal Justice Review No 1 of 2018 


Click here for a printable .pdf version of this Review


Criminal Law Pocket Library
Essential Legal Practitioner Bundle, The
Juta’s Statutes of South Africa 2017/18 (print)
South African Criminal Law Reports (1990 to date), The (Online)


Click here to register a profile and opt-in to receive these newsletters.

Click here to opt-out from this communication (please note that users are still required to have a profile in order to opt-out from receiving marketing communications from this website).


> Jutastat e-publications user helpdesk

> Juta Customer Services

> Juta Law Marketing


Jutastat electronic subscribers and print subscribers may receive important service-related information relating to their Juta publications from time to time.

Visit the Juta Law Website

© Juta and Company (PTY) LTD  2019