A bi-annual update complementing the

Commentary on the Criminal Procedure Act


NO 1 OF 2016


Andrew Paizes, Author (Editor)

Steph van der Merwe (Author)



The first of our two feature articles considers the ambit of legal professional privilege when confidential legal advice finds its way into the public domain, and questions the orthodox view that denies a client a remedy to restrain publication of what would have been a privileged communication had the confidentiality of the information not been lost. It is an issue that was raised and addressed by the court in South African Airways SOC v BDFM Publishers (Pty) Ltd & others 2016 (2) SA 561 (GJ).

The second feature article considers two decisions of the Supreme Court of Appeal which pivot around an issue that has received much attention in this Review: dolus eventualis, or legal intention. The first of these cases was the appeal by the State on questions of law reserved in the highly publicised case of Director of Public Prosecutions, Gauteng v Pistorius 2016 (2) SA 317 (SCA), in which the court found that the trial court had wrongly applied the principles relating to dolus eventualis and error in objecto. The second, S v Van Schalkwyk [2016] ZASCA 49 (unreported, SCA case no 680/15, 31 March 2016), is striking for the rigorous critical analysis of Willis JA who, in a dissenting judgment, examined and found wanting the approach of our courts in previous cases to the second limb of the test for dolus eventualis—the so-called ‘volitional element’. Willis JA endorsed, further, the ideas expressed in an article in this Review that the type of activity in which an accused was engaged may be critical in determining the presence of dolus eventualis, and that since dolus eventualis is a ‘tainted intention’, a moral judgment has to be formed in making such a determination.

Two statutory enactments feature in our section on legislation: the Criminal Matters Amendment Act 18 of 2015, which came into force on 1 June 2016, which creates new offences relating to essential infrastructure and contains special bail procedures and sentencing provisions for those offences; and the Judicial Matters Amendment Act 24 of 2015, which came into force on 8 January 2016 (except for certain sections), and which repeals s 384 of the old Criminal Procedure Act (Act 56 of 1955) and amends s 98 of the Child Justice Act.

Topics considered in our review of recent cases include:

  • a consideration of how far an accused has to go in order to effect an uncompleted attempt;
  • the tests for legality and rationality in reviewing a decision to discontinue a prosecution;
  • irregularities caused where a legal representative appears for two accused who have a conflict of interest, or where a co-accused is represented by someone without a right of appearance;
  • whether a judicial officer who wishes to subpoena or recall a witness should give the parties an opportunity to address the court;
  • the need to prevent the abuse of the accused’s right to legal representation;
  • the exclusion of statements made by an accused in the course of a pointing out after having been denied the right to legal assistance;
  • the role of the public interest in determining the admissibility under the Constitution of evidence elicited by undercover operations;
  • the fixing of a non-parole period as part of a sentence of imprisonment;
  • whether two judges sitting as a review court can grant leave to appeal to the Supreme Court of Appeal; and
  • whether the State can appeal to the Supreme Court of Appeal against a sentence imposed by the High Court sitting as a court of appeal.

Andrew Paizes




  • Legal professional privilege and the release of confidential legal advice into the public domain

  • Dolus eventualis: Two more decisions by the Supreme Court of Appeal


  • The Criminal Matters Amendment Act 18 of 2015
  • Background to and objects of the Act
  • Definitions: s 1
  • Offences created by s 3(1)
  • Bail in respect of essential infrastructure-related offences: s 2
  • Amendment of Schedule 5 to the Criminal Procedure Act: s 4
  • Amendments to s 51 of the Criminal Law Amendment Act 105 of 1997: ss 5 to 8
  • Amendment of Schedule 1 to the Prevention of Organised Crime Act 121 of 1998 (‘POCA’): s 9
  • Remarks in conclusion
  • The Judicial Matters Amendment Act 24 of 2015
  • Repeal of s 384 of the Criminal Procedure Act 56 of 1955: s 2
  • Expungement of certain criminal records in terms of the Child Justice Act 75 of 2008: s 20


(a) Criminal Law

  • Attempts: When does the end of the beginning change to the beginning of the end?

(b) Criminal Procedure and Evidence

  (i) Pre-sentence

  • Review of a decision to discontinue a prosecution: Legality and rationality
  • ss 57 and 57A: Admission of guilt and payment of a fine
  • s 60(11)(a): Bail and the relevance of conditions of detention in determining exceptional circumstances
  • s 73: Legal representative appearing for two accused who have a conflict of interest
  • s 73: The regularity of trial proceedings where a co-accused is represented by someone without the right of appearance
  • s 112(1)(b): The undefended accused’s plea of guilty and questioning by the court
  • s 162: Witnesses sworn in by interpreter and not by the presiding judicial officer
  • ss 167 and 186: When should a court subpoena or recall a witness, and should the parties be given an opportunity to address the court?
  • s 168: Right to have trial postponed for legal representation cannot be abused
  • ss 190 and 210: Previous consistent statements and sexual complaints: Complainant not testifying
  • ss 217 and 218: Statements made in the course of a pointing out after accused denied legal assistance
  • s 252A: Undercover operations and the Constitution

       (ii) Sentencing

  • s 276B: The fixing of a non-parole period as part of a sentence of imprisonment
  • s 280(2): Order that a later sentence of imprisonment should run concurrently with an earlier one

       (iii) Appeal and Review

  • Appeal: Can two judges sitting as a review court grant leave to appeal to the Supreme Court of Appeal?
  • s 316B: Can the State appeal to the Supreme Court of Appeal against a sentence imposed by the High Court sitting as a court of appeal?
Legal professional privilege and the release of confidential legal advice into the public domain

Several important aspects of the legal professional privilege were examined by the court in South African Airways SOC v BDFM Publishers (Pty) Ltd & others 2016 (2) SA 561 (GJ), which, as Sutherland J observed, was a case which required the court to ask what remedies are available to a person whose confidential legal advice is, by some or other unauthorised means, released into the public domain. The communication around which the matter pivoted was a document composed by the ‘General Manager: Legal Risk and Compliance’ of South African Airways, the applicant, which contained advice accompanied by legal opinion on the potential acquisition of new aircraft by the airline. The contents of this report by the in-house legal adviser of the applicant were clearly confidential, and no serious challenge was made to the proposition that the contents of the document were eligible to be the subject matter of a claim of privilege. Although the matter was once moot, Sutherland J accepted that the notion that an attorney who is not in private and independent practice but is an employee of an entity may be regarded as a ‘legal adviser’ for the purpose of the privilege, is now recognised in our law (see Van der Heever v Die Meester en andere 1997 (3) SA 93 (T) and Mohamed v President of the Republic of South Africa & others 2001 (2) SA 1145 (C)).

The central issue in this case was whether an order by the High Court obtained by the applicant interdicting the respondents, all media houses, from publishing information derived from this document, should be reconsidered. In particular, the court turned its attention to (1) whether the applicant could invoke a claim of privilege ‘against the world’ to protect confidentiality; and (2) whether the failure of the applicant to claim the privilege, despite its engagement with journalists on four occasions, could be said to amount to an imputed waiver of that privilege.

In respect of the first question, the applicant relied on the decisions in R v Special Commissioner & another, Ex parte Morgan Grenfell & Co Ltd [2002] UKHL 21 and Three Rivers District Council & others v Governor and Company of the Bank of England [2004] UKHL 48 for the proposition that legal professional privilege is a human right. In the former decision Lord Hoffmann (at [7]) described the privilege as ‘a fundamental human right long established in the common law’ and ‘a necessary corollary of the right of any person to obtain skilled advice about the law’. In the latter, Lord Scott went as far as maintaining (at [25]) that ‘if a communication or document qualifies for legal professional privilege the privilege is absolute’ and ‘cannot be overridden by some supposedly greater public interest’. It was argued further by the applicants that once a person has exercised the human right to claim privilege, the right can be invoked as against the world to protect and preserve the confidentiality of the information which is subject to the claim of privilege. It was argued that the effect of this proposition was that even where the confidentiality has been breached, the right to protection is not extinguished but continues in perpetuity. It was for this reason, claimed the applicant, that the interdict was appropriate in order to prevent the further dissemination of the information.

Before considering these arguments, Sutherland J found it necessary to address various questions of terminology. He took issue, first, with the label ‘legal professional privilege’, preferring ‘legal advice privilege’ on the ground that the latter ‘actually tells one what it is about, whilst the former phrase demands further explanation’ (at [45]). He turned, next, to ‘[t]he more interesting question ... the content of the right, ie what does the right which vests in the client entitle the client to do?’ To say that a ‘document is privileged’ he said, ‘suffers from three drawbacks in distilling the exact content of the right’: first, it is not the document but the information contained in it that is privileged; second, even to describe the information as privileged obscures the point that the right vests in the client, not in the information itself which is never more than the subject matter of a claim of privilege; and, third, privilege cannot reside in information because it becomes the subject matter of the claim of privilege only when that right not to disclose is claimed, and not before then. The information, then, can never be more than eligible to be the subject matter of the privilege, and it does so if all the requirements for privilege are met and if it is claimed by the client.

The exact nature of the right, said Sutherland J (at [47]), flows from its rationale, which rests on what has been understood to be the essence of the adversarial legal system: the ‘right of a person to a guarantee of confidentiality over communications with that person’s legal adviser is an indispensable attribute of the right to counsel and the adversarial litigation system’, so that the ‘professional duty of legal practitioners towards their clients is inseparable from the duty to respect their clients’ wishes about the secrets revealed by the clients and the confidential advice given to the clients’. By claiming the privilege, ‘the client invokes a ‘‘negative’’ right, ie the right entitles a client to refuse disclosure by holding up the shield of privilege’. The right to refuse to disclose legal advice in proceedings could not, said the judge, be ‘a ‘‘positive right’’, ie a right to protection from the world learning of the advice if the advice is revealed to the world without authorisation’ (at [48]).

As a result, said Sutherland J, although a client may restrain a thief who takes such a document on delictual grounds, there is, otherwise, no remedy in law to restrain publication by strangers who learn of the information if the confidentiality is lost. This is ‘because what the law gives to a client is a ‘‘privilege’’ to refuse to disclose, not a right to suppress publication if the confidentiality is breached’. A client, then, ‘must take steps to secure the confidentiality and, if these steps prove ineffective, the quality or attribute of confidentiality in the legal advice is dissipated’. This is, he added (at [49]), because ‘[t]he concept of legal advice privilege does not exist to secure confidentiality against misappropriation; it exists solely to legitimise a client in proceedings refusing to divulge the subject matter of communications with a legal adviser, received in confidence’. This proposition, said Sutherland J, was ‘endorsed by the authorities’ and is supported by Wigmore (at paragraphs 2325–6), who was of the view that ‘the risk of insufficient precautions is upon the client’ since the law grants ‘secrecy so far as its own process goes’ but leaves it to the client and attorney ‘to take measures of caution sufficient to prevent being overheard by third persons’. Wigmore goes as far as saying that ‘[s]ince the means of preserving secrecy of communication are largely in the client’s hands and since the privilege is a derogation from the general testimonial duty and should be strictly construed, it would be improper to extend its prohibition to third persons who obtain the knowledge of the communications’ (emphasis added).

Sutherland J rejected the notion that privilege was an absolute right, saying that this was not the law in South Africa (see Thint (Pty) Ltd v National Director of Public Prosecutions & others; Zuma v National Director of Public Prosecutions & others 2009 (1) SA 1 (CC) at [183]–[185] and Midi Television (Pty) Ltd t/a E-TV v Director of Public Prosecutions (WC) 2007 (5) SA 540 (SCA) at [9]). It was held in these cases that the right was not absolute; that it may be outweighed by countervailing considerations; and that a balancing of interests was unavoidable in order to reconcile contending values protected in the Constitution. This was, too, the way in which the Promotion of Access to Information Act 2 of 2000 viewed the privilege, which is subject to the public interest override set out in s 46.

Because the claim to privilege could not succeed, Sutherland J declined to consider whether there had been a waiver of that right. He did turn, however, to the applicant’s assertion of its broader right to confidentiality, of which privilege was ‘simply an example’. He concluded that the applicant did have a right to confidentiality in the disputed information, and that this was its true cause of action. But was this right waived in the circumstances? It was argued that there was imputed waiver of the right, based on the dictates of fairness and quite apart from whether there was any intention, actual or implied, by the applicant to abandon the right. Since there was no proof that the applicant understood that the document was eligible for protection before consulting with its legal advisers, Sutherland J concluded that the circumstances did not justify imputing a waiver of confidentiality. He concluded, however, that further suppression of the information would not be in the public interest. Very little of the information contained in the document was not already in the public domain; the information was not harmful to the applicant’s interests; and the subject matter of the document was a ‘demonstrably obvious topic about which every citizen ha[d] a tangible interest to be informed.’

In any event, said Sutherland J, the facts demonstrated that the order, as granted, was ‘futile even as the ink dried upon it’ (at [30]), and the courts have ‘long recognised that, in general, they should not make orders to which effect cannot be given’. An interdict is an appropriate form of relief to prevent future harm, not afford redress for past harm: ‘once confidentiality is shattered, like Humpty Dumpty, it cannot be put back together again’ (at [38]).

There is no doubt that the judgment in this case on the question of legal advice privilege squares with legal orthodoxy and cannot be faulted for not rendering faithfully the rules, as we have them, pertaining to the privilege, however we might wish to call it (and ‘legal advice privilege’ is as good a name as, if not better than, any other). From the perspective of developing our law further in line with the spirit and purport of the Constitution and its values, however, some questions may be raised.

The first relates to the loss of the privilege by reason of the interception of the information by a third party. The treatment of this question by Sutherland J and the views of Wigmore, with which the judge agreed, echo the stance taken by our courts before the Constitution was enacted when the privilege was seen in its more limited incarnation as an evidentiary rule. Much has happened since then. The decision in S v Safatsa & others 1988 (1) SA 868 (A) ushered in a shift in perspective and emphasis, with the privilege coming to be viewed as a ‘fundamental right of a client’; a ‘mere manifestation of a fundamental principle upon which our judicial system is based’; and a ‘doctrine which is based upon the view that confidentiality is necessary for the proper functioning of the legal system and not merely the proper conduct of particular litigation’ (see Commentary in the notes to s 201). These principles have, since Safatsa, taken even deeper root in the post-Constitutional jurisprudence. There has, in this time, been a consolidation of the shift of the privilege from an evidentiary rule to a fundamental right that is seen as being central to the effective working of the adversarial system and the guarantee of a fair trial (see, for instance, Mahomed v National Director of Public Prosecutions & others 2006 (1) SACR 495 (W) at [7]; Bennett & others v Minister of Safety and Security & others 2006 (1) SACR 523 (T) at [55]; S v Tandwa & others 2008 (1) SACR 613 (SCA) at [17]; Governing Body, Hoërskool Fochville & others v Centre for Child Law 2014 (6) SA 561 (GJ) at [21]; and Craig Smith and Associates v Minister of Home Affairs & others 2015 (1) BCLR 81 (WCC)). It has, now, both greater reach and greater legal significance, being seen as applying extra-curially to defeat seizure by warrant and as a corollary of the crucial constitutionally enshrined right to counsel in criminal proceedings (see s 35(3)(f) of the Bill of Rights and see Commentary). The privilege may, further, be seen as a buttress to other important rights: the right of access to the courts (s 34 of the Bill of Rights); the right not to be compelled to give self-incriminating evidence (s 35(3)(j); and the right to privacy (s 14), which includes the right of all persons not to have the privacy of their communications infringed. What has emerged, then, is a notion which, even if we continue to call it a ‘privilege’, is without doubt far more important and weighty than the evidentiary rule which Wigmore and the earlier cases described.

How to blend the two streams—the old learning surrounding the evidentiary rule and the new jurisprudence embracing the servicing of a multitude of fundamental rights and the integrity of the system of justice as a whole—is not a simple task. One method would be to jettison many if not all of the old rules and give the new concept an entirely new treatment within the canon of modern constitutional discourse. Another would be to keep the old structures but to develop them, as the Constitution demands, in a way that promotes more vigorously the spirit, purport and objectives of the Bill of Rights.

The courts have not expressly committed themselves to one of these policies. Some cases implicitly reflect the first perspective, and others the second. What the courts have accepted though, as Sutherland J pointed out, is that the privilege is not absolute and must allow for a broad balancing of contending rights and values (see, as an example, the approach of Davis J in the Craig Smith case, above).

Given the crucial role played by the ‘privilege’ in the modern era, how then should the balance be resolved when otherwise privileged information is intercepted by a third person? At the one extreme is the situation where the client is entirely without blame and where the document is stolen by a thief who uses either violence or stealth to acquire it. It is difficult to see how the imperatives of the Constitution would be served by stripping the client of the right in these circumstances. Apart from the delictual remedy to which Sutherland J alluded, it would seem necessary, if we are to take the right at all seriously, to conclude that it is not terminated by such an event and that the client is entitled to continue to enjoy, as much as this may be practically possible, the legal remedies and entitlements enjoyed before the confidentiality of the communication was compromised. If the integrity of the adversarial system and the fairness of the trial would be impaired if the attorney were forced by law to testify to the contents of the document, why would the position be otherwise if a witness sought to testify to them after members of a powerful gang used improper means such as force or threats or financial power to gain access to the information?

At the other end of the spectrum is the situation where the client actually abandons the right or is careless in his custody of the right. At its most extreme, such behaviour will lead to the conclusion that there has been a waiver in one of its forms (express, implied or imputed). But even if waiver has not taken place, it should be incumbent on a court to engage in the broad balancing of interests in order to determine whether the ‘privilege’ has been lost. Waiver itself is (or, at least, should be) an instance of this broad balancing in which the values protected by the privilege are compared and measured against the dictates of fairness and other interests and rights that militate against the protection of the information.

This balancing exercise was undertaken by the court in the South African Airways case, but only in the context of the broader right to confidentiality. It was not engaged in the context of the privilege, where the court assumed—on the strength of the authorities—that the right is always lost when there is ‘misappropriation’. It is submitted that this pre-constitutional approach should be revisited in view of the developments described above. It is submitted, too, that the balancing exercise should be undertaken in all cases of this kind, and that the concept of ‘waiver’ should be considered to be no more than a particular aspect of this exercise. The ‘rules’ governing waiver should, too, yield to the results of the broader balance within a constitutional context, and should not be regarded as being self-contained or as operating in a sealed box apart from that balance.

It may be, too, that it is inappropriate, if privilege is found not to apply, to consider the application of the right to confidentiality apart from that balancing exercise. It is conceptually sounder and more elegant to allow whatever gives strength to that right to add ballast to the right that now parades under the banner of ‘privilege’. If privilege is, as Sutherland J pointed out, merely an example of the broader right to confidentiality, there is no reason for creating a partition between the two when a court examines the specific case of privilege. To do so may, indeed, lead to error: if a court resolves the ‘privilege’ argument against the client and then moves on to the ‘confidentiality’ argument it will in all likelihood, resolve the latter question without taking into account the special reasons for protection that are peculiar to the privilege issue. There should be one enquiry in which all the factors calling for protection are weighed against all the factors calling for ventilation. In this case it probably made little difference which approach was followed: the information was, already, largely in the public domain; the conduct of the client came close to constituting a waiver of the right; disclosure was not potentially harmful to the applicant’s interests; and there were powerful policy reasons for allowing further public disclosure of the information. A broader balance of all competing interests and an examination of the question of misappropriation of the information within that exercise would probably have left the outcome unchanged.

But there will be cases where the choice of approach may be crucial. The courts will, in those cases, have to make an election about, first, how they approach the blending of pre- and post-constitutional ingredients; second, whether (and if so, how) to separate questions of misappropriation and waiver; third, whether (and if so, how) to position the ‘privilege’ question within the broader right to confidentiality; and, fourth, how to load the scales in the balancing exercise already endorsed by them.

Andrew Paizes

 Dolus eventualis: Two more decisions by the Supreme Court of Appeal

The question of dolus eventualis has featured prominently in this Review since its inception. One of the cases considered—and criticised—by this writer in this regard was the highly publicised trial of Oscar Pistorius, S v Pistorius (unreported, GP case no CC113/2013, 11 September 2014); see CJR 2014 (2) 4. In that case, the facts of which are so well known they will not be repeated, the court held that the accused lacked dolus eventualis but had mens rea in the form of culpa (negligence). It accordingly found the accused not guilty of murder but guilty of culpable homicide. Its treatment of dolus eventualis was summed up as follows (per Masipa J):

‘I now deal with dolus eventualis or legal intent. The question is:
(1) Did the accused subjectively foresee that it could be the deceased behind the toilet door; and
(2) Notwithstanding the foresight did he then fire the shots, thereby reconciling himself to the possibility that it could be the deceased in the toilet? The evidence before this court does not support the state’s contention that this could be a case of dolus eventualis.

On the contrary the evidence shows that from the onset the accused believed that, at the time he fired shots into the toilet door, the deceased was in the bedroom while the intruders were in the toilet.’

And further:

‘How could the accused reasonably have foreseen that the shots fired would kill the deceased or whoever was behind the door? Clearly he did not subjectively foresee this as a possibility that he would kill the person behind the door, let alone the deceased, as he thought she was in the bedroom at the time. The version of the accused was that had he intended to kill the person behind the door he would have aimed higher at chest level. This was not contradicted.’

And finally:

‘This court also accepts that there was no intention to kill the person behind the door. It follows that the accused’s erroneous belief that his life was in danger excludes dolus.’


The matter then went to the Supreme Court of Appeal on questions of law reserved in terms of s 319 of the Criminal Procedure Act (see Director of Public Prosecutions, Gauteng v Pistorius 2016 (2) SA 317 (SCA)). One of the questions was ‘[w]hether the principles of dolus eventualis were correctly applied to the accepted facts and the conduct of the accused, including error in objecto’. The court, per Leach JA with the other members of the court in agreement, held that they were not.

Leach JA (at [28]) found the reasoning in the above passages to be ‘confusing in various respects’. First, the question ‘How could the accused reasonably have foreseen that the shots he fired would kill the deceased or whoever was behind the door?’ wrongly applied an objective rather than a subjective test: the question is not what was reasonably foreseeable when he fired the shots but whether he actually foresaw that death might occur when he did so. Second, the apparent finding that the presence of a person behind the door was not foreseeable was at odds with the subsequent finding that the accused was guilty of culpable homicide on the basis that a reasonable person in his circumstances would have foreseen the reasonable possibility that the shots fired at the toilet door might kill whoever was in the toilet. Third, the finding that the accused did not foresee that he would kill whoever was behind the door since, if he had intended to do so he would have aimed higher than he did, conflated the requirements of dolus eventualis with those of dolus directus. Fourth, the trial court’s consideration of dolus eventualis centred upon whether the accused knew that the person in the toilet cubicle was the deceased, and its conclusion that dolus eventualis had not been proved was premised upon an acceptance that, as he had thought she was in the bedroom, he did not foresee that she was the person in the toilet. This finding involves the so-called ‘error in objecto’: it is required that an accused’s intention must relate to the very person who is killed; but it is not required that the accused must know or foresee the actual identity of the victim. The question is not whether the accused foresaw that the deceased might be killed when he fired the shots at the toilet door, but whether any person, whoever he or she might be, might be killed as a result.

The only factor, then, that could disturb a finding that the accused acted with dolus eventualis in the circumstances of the case, was if he acted in putative private defence—if, that is, it was reasonably possible that, when he fired the fatal shots, he genuinely (although erroneously) believed that his life was in danger. If this were found to be so, dolus eventualis would be lacking if it was reasonably possible that the accused did not foresee the real possibility that his conduct might unlawfully cause the death of the deceased.

Leach JA found difficulty with this defence since the accused stated in his evidence that he had not intended to shoot the person he had regarded as an ‘intruder’, and this placed him ‘beyond the ambit of the defence’ (at [53]). Further, the defence of putative private defence ‘implies rational but mistaken thought’, and ‘[e]ven if the accused believed that there was someone else in the toilet, his expressed fear that such a person was a danger to his life was not the product of any rational thought’. In addition, he did not know who was behind the door or whether that person in fact constituted any threat to him. In these circumstances, ‘although he may have been anxious, it is inconceivable that a rational person could have believed he was entitled to fire at this person with a heavy-calibre firearm, without taking even that most elementary precaution of firing a warning shot’. The defence of putative private defence was, then, no bar to a finding of dolus eventualis, and the question of law was answered in favour of the State.

Most of the criticisms of the trial court’s reasoning expressed by Leach JA square with what was submitted in this Review in CJR 2014 (2) 4 (see ‘The trial of Oscar Pistorius—dolus eventualis once again’). His statement of the principles relating to dolus eventualis are, in our view, correct and in line with authority. The failure of the defence of the putative private defence may be expressed pithily thus: it was not reasonably possible, on the facts, that the accused did not foresee the real possibility that, when he fired the shots at the door, he might unlawfully cause the death of another human being. Even if he believed that the person in the toilet cubicle was an intruder, he had no way of knowing who the intruder was (he may have been a child and unarmed), what his intentions were, whether he was entering or trying to escape in a panic through the window, or whether he was armed. He certainly had no grounds to believe that this person presented an imminent danger to his life or safety, and fired ‘without having a rational or genuine fear that his life was in danger’ (at [54]).

The second of the cases decided by the Supreme Court of Appeal that involved dolus eventualis was S v Van Schalkwyk [2016] ZASCA 49 (unreported, SCA case no 680/15, 31 March 2016). The appellant in that case had struck the deceased, who was a farm worker employed by him, with an iron hay hook on the left side of his chest. The hook pierced ten centimetres into his heart and also severed his fifth rib. The deceased died as a result of his injuries. The appellant was convicted by a magistrate of murder, his conviction was upheld by the High Court, but set aside by the Supreme Court of Appeal and replaced with a conviction of culpable homicide.

The court was divided on the question whether the appellant had actually foreseen that his actions might cause the death of the deceased. The majority, per Lewis JA (Tshiqi JA and Plasket AJA concurring) held that this had not been proved beyond a reasonable doubt. The minority, Baartman AJA and Willis JA, who delivered separate judgments, disagreed.

The most striking aspect of the case is the treatment of dolus eventualis by Willis JA, who focused in particular on the second limb of the test—the so-called ‘volitional element’. He referred to the judgment of Botha AJA in  S v Dladla en andere 1980 (1) SA 1 (A), who examined the Dutch writers and cited a passage by Van Hattum, which Willis JA (at [27]) translated as follows:

‘The reasoning concerning the question of intention puts the question ... in this way: what would the perpetrator rather have intended, the realisation of that which accompanies his intended act together with that which had been intended or the abandonment of his act (and therefore the setting of his face against that which he had intended)? If one comes to the conclusion that the perpetrator was so focused on achieving that which he had intended that he would rather continue with his intended act, despite its unintended consequences, rather than set his face against it, then one deduces therefrom that the perpetrator brought into his intention even that emergent possibility. That is then dolus (eventualis).’ (Emphasis Willis JA)

In understanding the second element of dolus eventualis—and what is meant by what the cases describe as ‘nevertheless proceeding ‘‘recklessly’’’, the concept of ‘bringing into’ one’s intention an emergent possibility had to be understood in the light of the notion of ‘afzien’ or ‘setting one’s face against something, abandoning it’ (at [28]). It is the ‘failure to do so, once one has foreseen the possibility of the consequence ensuing, that is critical’, and this, said Willis JA, is what is meant by the second limb of the test.

In S v Swanepoel 1983 (1) SA 434 (A), said Willis JA, the court referred (at 456H) with approval to the views of Snyman (Strafreg) who said that, in addition to the requirement of subjective foresight, the perpetrator must ‘versoen hom met hierdie moontlikheid’. Snyman, according to Willis JA, however, ‘subtly reinterpreted a negative obligation—to refrain or abstain from doing something into a positive requirement that the perpetrator must ‘‘versoen’’ himself with the possibility of it occurring’. And, ‘apparently influenced by Swanepoel’, the (then) Appellate Division in S v Ngubane 1985 (3) SA 677 (A) began using terminology like ‘taking a conscious risk’, ‘consenting’, ‘reconciling’, ‘taking into the bargain’ in addition to ‘nevertheless persisting in his conduct’ in order to describe the ‘volitional element of dolus eventualis.

The word ‘versoen’, said Willis JA, translates into English as ‘be reconciled with’. But, he points out, ‘[s]omething is, however, lost in translation in the process’. ‘To be reconciled’, he explains, ‘has connotations of mature and considered intellectual and moral reflection, an introspection and self-examination, often over a period of time’ (at [31]). And this, he added, ‘is not what is required before a conviction based on dolus eventualis can ensue’. In his view ‘nuances of translation’ could explain some of the difficulties associated with the term ‘be reconciled with’. The ‘ordinary, everyday idiomatic expressions in the English language such as ‘‘do not flirt with death’’, ‘‘do not court death’’, ‘‘do not play with death’’ and ‘‘do not dance with death’’’ captured better, in his opinion, what the law demands ‘rather than an abstract conceptualisation as to what it means to be ‘‘reconciled with’’ the possibility of death occurring’.

Willis JA considered that, in line with what was said in S v Dougherty 2003 (4) SA 229 (W), ‘the law requires that the prohibited act must have been committed dolo malo, that is with a bad, evil or wicked intention’ (at [33]). Accordingly, a ‘value judgment’ had to be made concerning the volitional element ‘as to whether or not the accused should ‘‘afzien’’ at the critical moment’. He found it helpful to refer to another article by the present writer in this Review (‘Dolus eventualis revisited: S v Humphreys 2013 (2) SACR 1 (SCA)’, CJR 2013 (1) 6) in which reference is made to an article by Professor Roger Whiting ((1988) 3 SACJ 440) ‘to underscore the point that the type of activity involved may be critical in determining whether dolus eventualis was present and that, for example, even though the foresight of the possibility of death and a person’s being reconciled thereto may be present in everyday activities such as driving or mining, deaths that result from such activities ordinarily do not result in a conviction of murder’ (at [34]).

Dolus eventualis, Willis JA explained, is a ‘tainted intention’. He stressed, relying on what this writer said in an earlier article (see (1988) 105 SALJ 636), that ‘when all is said and done, a moral judgment has to be formed to determine whether dolus eventualis is present’. He pointed out that in the later article by this writer (in the CJR) it was argued ‘that factors such as callousness and the purpose of exposing the victim to the risk of death all weigh in the equation to determine whether dolus eventualis was present’.

How significant is this judgment by Willis JA and what impact will it have on the articulation of the test for dolus eventualis in our criminal law? It is true that it is a minority dissenting judgment, but it must be kept in mind that he was not at odds on this point with the majority who, having held that the appellant had not foreseen the possibility of his conduct causing the death of the deceased, had no cause to examine the second leg (the ‘volitional element’) of the test for dolus eventualis. Willis JA was, on this aspect of that test, not necessarily swimming against the judicial current. Moreover, the judgment is, in part, consonant with the full bench decision of the High Court in S v Maarohanye & another 2015 (1) SACR 337 (GJ). This judgment is discussed in CJR 2015 (1) 14. The court (Mlambo JP, Maluleke J and Pretorius J), following what was said in the CJR article ‘Dolus eventualis revisited’, accepted that ‘[d]olus eventualis ... is not amenable to containment within a simple formula, the facts of the matter having a lot to do with the ultimate conclusion’ (at [21]).

To the extent that the judgment of Willis JA may contribute to the liberation of dolus eventualis from its straitjacket of mechanical formulation, it is to be welcomed. It is a salutary reminder that fault or blameworthiness (of which dolus eventualis is but one manifestation) requires a suitably tainted state of mind. It also recognises that any attempt by the courts to identify a universal solvent to describe what state of mind in all cases attracts the dolus eventualis label would, no matter how carefully that test is crafted, necessarily be imperfect.

It is true that if you commit an act that has no social utility and that is inherently dangerous and prima facie unlawful, such as hitting another person on the head with a heavy object, you will have dolus eventualis in respect of that person’s death if you go ahead with the act after you have foreseen the possibility of causing that result. But it is not true if you run a huge mining operation (which is inherently dangerous, has social utility and is not prima facie unlawful) with a similar state of mind. If it were, then you would be guilty of attempted murder every time a miner went down your mine and all mining operations would have to close. In addition, driving a car would attract a similar treatment, and transport would cease to function. In short, economic reality and common sense dictate that you cannot ordinarily be regarded as having mens rea in the form of dolus eventualis if you cause another’s death in a driving accident—even if you drive recklessly and even if you foresee the possibility that another person may be killed as a result of your conduct. Mining and traffic accidents are properly and suitably the province of another form of fault designed for this purpose: culpa or negligence.

In both Van Schalkwyk and Maarohanye the courts have endorsed the various factors identified by Whiting and by the present writer as being relevant and influential in making a non-formulaic determination of dolus eventualis. These, it is hoped, will be relied on further by the courts in developing a more flexible and nuanced approach.

What of Willis JA’s comments on the volitional element of dolus eventualis? It seems that the learned judge came very close to the arguments raised by Whiting and by this writer: that the second limb of the test adds nothing to the first since, once an accused goes ahead and commits the act that constitutes the conduct element of the actus reus, foreseeing that this act might cause the death of the victim, he will necessarily have failed to ‘afzien’ or ‘set his face against’ that act. Had he not done so, he would have desisted from performing it. The second limb, Whiting and I maintain, is merely descriptive of what must have been his state of mind when he performed the act. It has no additional value and cannot usefully be considered a separate condition for determining dolus eventualis.

Willis JA seemed implicitly to accept that this was so. In his conclusion (at [35]), after finding on the facts that the inference was irresistible ‘that when the accused was about to strike the deceased with a hay hook, he foresaw the possibility that death might ensue even though that may not have been what he wanted to happen’, he goes on to say:

‘He should have stopped himself there and then. He did not do so. He flirted with death. He did not ‘‘afzien’’ from his intended act. Having gone ahead, despite having foreseen such a well-known risk and of which he, as a farmer, must have been acutely conscious, the accused is confronted with a moral judgment of the community that is one of deep opprobrium. He is therefore guilty of murder.’

If what Whiting and I submit (and what Willis JA seems to accept) is correct, no murder trial where dolus eventualis is relied on by the prosecution should turn on whether or not the second ‘leg’ of the test is satisfied. The recent decision of the Supreme Court of Appeal in S v Humphreys (supra) was a case which did. I argued in the CJR article that the correct result was reached in that case, but for the wrong reason. The appellant was correctly found not guilty of murder, not because he did not take the risk of death into the bargain, but because the activity in which he was involved—driving a motor vehicle—was not one in respect of which the foresight of causing death could appropriately be viewed as attracting liability for murder by engaging the conventional test for dolus eventualis.

If the more flexible approach to blameworthiness endorsed by Willis JA finds its way into the judicial mainstream, and it is accepted, at the same time, that the ‘volitional’ element has no analytical teeth, a more realistic and sharper conceptual framework for tackling a wide range of circumstances will emerge, and our criminal law will be the richer for it.

One might expect that the courts will not easily give up their attachment to the simplistic formula they have used for so long. The formula, after all, works quite well in most of the cases, and the approach supported in this Review may seem a little messy and complicated. But it must be remembered that our analytical tools are no more than a means to the greater end of achieving just results. When they no longer do so, they must be reviewed and either modified or replaced. What, after all, is there to allow us to assume that a very complicated question—one that requires us to ascertain whether a person’s state of mind was sufficiently tainted to warrant a conclusion that he or she was ‘blameworthy’ or at fault—should be amenable to a simple answer by using a pithy, almost mechanical, one-size-fits-all formula?

Andrew Paizes

 The Criminal Matters Amendment Act 18 of 2015

The above Act (hereafter the ‘Act’) was assented to by the President on 15 December 2015. See GG 39522 of 15 December 2015. The Act came into operation on 1 June 2016 (see Proclamation 33 in GG 40010 of 24 May 2016).

Background to and objects of the Act

In paragraph 1 of the memorandum which accompanied the Bill [B 20B–2015] which preceded the Act, it was pointed out that there is an unacceptably high level of crime in respect of essential infrastructure providing basic services to the public. These crimes have had an adverse effect on the South African economy. They have also posed a risk to public safety, electricity supplies and the provision of water, communications and transportation. In the memorandum it was also pointed out that ‘essential infrastructure-related offences are becoming increasingly more organised and are often committed by armed and dangerous criminal groups’. In the preamble to the Act reference is made to the harmful consequences to the livelihood and daily economic activity of the public ‘if basic services cannot be provided due to loss, damage or disruption caused by essential infrastructure-related offences ...’

In addressing the problems as identified above, the Act has created offences relating to essential infrastructure. The Act also contains certain special bail procedures and sentencing provisions as regards these offences.

Definitions: s 1

The most important definition in the Act is the one governing ‘essential infrastructure’. In terms of s 1 these two words refer to ‘any installation, structure, facility or system, whether publicly or privately owned’, the loss or damage of, or the tampering with which ‘... may interfere with the provision or distribution of a basic service to the public ...’ The words ‘basic service’ are, in turn, defined as a service rendered by the public or private sector and which pertains ‘to energy, transport, water, sanitation and communication, the interference with which may prejudice the livelihood, well-being, daily operations or economic activity of the public ...’ (s 1).

The word ‘tamper’ as used in the Act has been given a broad meaning and includes altering, cutting, disturbing, interrupting, manipulating, obstructing, removing or uprooting. ‘Tampering’ can be ‘by any means, method or device ...’ (s 1). In the Afrikaans text of the Act the word ‘peuter’ is used for ‘tamper’, and ‘gepeuter’ is used for ‘tampering’.

Offences created by s 3(1)

The core offence in the Act is created by s 3(1)(a). This section provides that any person who ‘unlawfully and intentionally ... tampers with, damages or destroys essential infrastructure ... and who knows or ought reasonably to have known or suspected that it is essential infrastructure, is guilty of an offence ...’. In terms of s 3(1)(b) any person who intentionally colludes with or assists another person ‘in the commission, performance or carrying out of an activity’ as described in s 3(1)(a), is guilty of an offence.

Section 3(2) determines that for purposes of the offences created by s 3(1)(a) and (b),

‘... a person ought reasonably to have known or suspected a fact if the conclusions that he or she ought to have reached are those which would have been reached by a reasonably diligent and vigilant person having both—
(a)    the general knowledge, skill, training and experience that may reasonably be expected of a person in his or her position; and
(b)    the general knowledge, skill, training and experience that he or she in fact has.’

Section 3(2), it seems, introduces an objective standard for purposes of deciding whether there was awareness on the part of the offender that ‘essential infrastructure’ was involved in the offence; however, this test is also toned down by the criteria stated in s 3(2)(a) and (b). These criteria require that the reasonable person be placed in the same circumstances as the accused.

The sentence identified for contravention of s 3(1)(a) or (b) is imprisonment not exceeding 30 years or, in the case of a corporate body successfully prosecuted on the basis of s 332(2) of the Criminal Procedure Act, a fine not exceeding R100 million.

Bail in respect of essential infrastructure-related offences: s 2

In terms of the above section police bail and prosecutorial bail (as provided for in ss 59 and 59A of the Criminal Procedure Act) may not be granted to persons who are in custody in respect of ‘any offence involving ferrous or non-ferrous metal which formed part of essential infrastructure’ (s 2(a)) or an offence referred to in s 3 of the Act (s 2(b)). In respect of these two categories, the only option open to the arrested person would be to apply for bail in court as provided for in s 60 of the Criminal Procedure Act.

Amendment of Schedule 5 to the Criminal Procedure Act: s 4

The above Schedule has been amended by s 4 of the Act. This section inserted certain essential infrastructure-related offences in the Schedule so that an accused charged with these offences may only be released on bail if he were to adduce evidence which satisfied the court that the interests of justice permit his or her release. See s 60(11)(b) of the Criminal Procedure Act. See also the discussion of s 60 in Commentary, sv Section 60(11)(b): Interests of justice and strength (or weakness) of State’s case.

Amendments to s 51 of the Criminal Law Amendment Act 105 of 1997: ss 5 to 8

Section 5 of the Act amends s 51 of Act 105 of 1997 (the so-called ‘discretionary minimum sentence legislation’) by inserting a new Part V in Schedule 2 of Act 105 of 1997, which makes provision for prescribed minimum sentences ranging from 3 to 7 years’ imprisonment.
Part II and Part IV of Schedule 2 and Schedule 2 itself have also been amended by ss 6, 7 and 8 of the Act in order to ensure that essential infrastructure-related crimes are also brought within the ambit of discretionary minimum sentence legislation.

Amendment of Schedule 1 to the Prevention of Organised Crime Act 121 of 1998 (‘POCA’): s 9

The above Schedule to POCA has been amended by s 9 of the Act by adding as item 33B the two essential infrastructure offences created by s 3 of the Act.

Remarks in conclusion

It is clear that the Act is based on stricter crime control considerations and measures: tighter substantive law provisions, limitation of certain bail procedures and stiffer sentences. Essential infrastructure crime (like ‘cable theft’) damages the national economy and disrupts methods of communication and transportation. However, the fact that the Act is in place to combat all these risks is far from sufficient. The reality is that the ultimate success of the Act will depend upon proper policing and effective police investigations. It is the certainty of getting caught that eradicates crime of this nature.

The Judicial Matters Amendment Act 24 of 2015

The above Act (hereafter ‘the Act’) came into operation on 8 January 2016, except for sections 5, 12, 13–14 and 16–19 which must still come into operation on a date to be fixed by the President by proclamation. See GG 39587 of 8 January 2016 and s 23 of the Act. Only ss 2 and 20 are of direct importance to our criminal justice system.

Repeal of s 384 of the Criminal Procedure Act 56 of 1955: s 2

Section 384 of Act 56 of 1955 has been repealed by s 2 of the Act because it was considered obsolete in the criminal justice context. Section 384—which dealt with the binding over of persons to keep the peace—was one of two sections not repealed when the Criminal Procedure Act 51 of 1977 came into operation on 22 July 1977. See GG 5656 of 15 July 1977.

The only section in Act 56 of 1955 still ‘standing’, is s 319(3) which is cited in full in Commentary at xiii. This section deals with conflicting statements under oath and supplements common-law perjury.

Expungement of certain criminal records in terms of the Child Justice Act 75 of 2008: s 20

Section 98 of the Child Justice Act has been amended by s 20 of the Act. Section 20 added s 98(4). Section 98(4) provides that—despite the provisions of s 4 of the Child Justice Act—any child who, before the commencement of the Child Justice Act, had been convicted of offences identified in s 98(4)(a) and (b) ‘may apply for the expungement of his or her criminal record’ as provided for in s 87 of the Child Justice Act. See further the discussion of s 271A in Commentary, sv Expungement of records and the provisions of s 87 of the Child Justice Act 75 of 2008.

 (a) Criminal Law

Attempts: When does the end of the beginning change to the beginning of the end?

S v Silo (unreported, WCC case no A59/15, 22 March 2016)

In this case the court had to consider whether the appellant had been correctly convicted of attempted sexual penetration in contravention of s 55 of the Criminal Law (Sexual Offences and Related Matters) Amendment Act 32 of 2007. He had pushed the complainant into her apartment; closed the door behind him; pushed her further into the bedroom and onto the bed; told her that he wanted to have sexual intercourse with her; instructed her to take off her gown and panties, which she did; and proceeded to assault her. After wrestling with her, he choked her and went to the kitchen to fetch a knife. She escaped by jumping out of her bedroom window onto the ground outside the block.

The question facing the court was whether the conduct of the appellant had gone beyond mere acts of preparation to commit the offence, and had reached the stage described in the cases as ‘the commencement of the consummation’. As Watermeyer CJ pointed out in R v Schoombie 1945 AD 541, there is a fine line between the ‘end of the beginning and the beginning of the end of a crime or of defining in exact terms what is meant by its consummation’. As long as the accused’s acts ‘have reached such a stage that it can properly be inferred that his mind was finally made up to carry through his evil purpose he deserves to be punished because, from a moral point of view, the evil character of his acts and from a social point of view the potentiality of harm in them are the same, whether such interruption takes place soon thereafter or later.’

Henney J considered (at [18]) that a court, in applying these principles, had to follow ‘a pragmatic and common sense approach given the circumstances of the case at hand’. He was of the view that the appellant’s conduct had gone beyond mere acts of preparation and constituted a completed attempt. It was not necessary, for an attempted rape to take place, for the State to prove any of the following: that the perpetrator had forced himself on the victim by lying on top of her; that he had taken off his clothes; that he had opened her legs; that she had bruises on her upper legs; or that there was DNA evidence matching that of the perpetrator.

In R v B 1958 (1) SA 199 (A) it was held that the assault of a complainant constitutes an attempted rape once the perpetrator resolves to have intercourse with the victim, even if he expects to do so only at a later stage (see, too, S v W 1976 (1) SA 1 (A)). An assault on a victim would, then, said Henney J (at [25]), constitute attempted rape ‘if it is clear that the perpetrator inflicted such assault with the intention to rape in order to restrain or overcome the resistance of a victim’. In this case much more had happened. It was clear, from the conduct of the appellant, that he had commenced the consummation of the crime. 


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

Review of a decision to discontinue a prosecution: Legality and rationality

Democratic Alliance v Acting National Director of Public Prosecutions & others (unreported, GP case no 19577/2009, 29 April 2016) was an application to review and set aside the decision of the Acting National Director of Public Prosecutions (‘ANDPP’), taken on 6 April 2009, to discontinue the criminal prosecution of the third respondent (the current President) on account of an alleged abuse of process involving the second respondent’s undue interference with the service of the indictment for political reasons at a time when the second respondent was head of the former Directorate of Special Operations. The application was based on the Constitution and the principle of legality. It was accepted that a decision to discontinue a prosecution is reviewable on the grounds of legality and rationality; and the applicant and the respondents’ counsel, by agreement, essentially based their arguments on the ground of irrationality (at [49]).

The full bench identified several instances of ‘irrationality’. At [88] it was found, for example, that the ANDPP had failed

‘... to explain how the information he had heard on the tape could be said to have affected, compromised or tainted the envisaged trial process and the merits of the intended prosecution. In fact, in his media address, he concedes that the alleged conduct of Mr McCarthy [the Head of the former Directorate of Special Operations] had not affected the merits of the charges against Mr Zuma. There was thus no rational connection between the need to protect the integrity of the [National Prosecuting Authority] and the decision to discontinue the prosecution against Mr Zuma.’

The full bench also noted that the ANDPP had dismissed—without reasons—the prosecution team’s recommendation or advice that even if the allegations of political interference by McCarthy were true, the decision to discontinue the prosecution had to be a judicial one (at [65]). At [71] the full bench also said: ‘It is thus our view that [the ANDPP], by not referring the complaint of abuse of process and the related allegations against McCarthy to court, rendered his decision irrational’.
Reference was also made to the fact that the ANDPP did not consult the DPP concerned who, after all, was the one who had signed the indictment in respect of a case in which he had authorised the prosecution (at [87]). The extent to which the ANDPP failed ‘to source the views’ of certain senior members of the National Prosecuting Authority involved in the matter was also ‘irrational’ (at [83]).

At [90] the full bench also observed and concluded that the ANDPP

‘... in his own words on 1 April 2009 stated that he felt angry and betrayed. It is the view of this Court that his feelings of anger and betrayal caused him to act impulsively and irrationally, considering the factors as stated preceding paragraphs. He did not allow himself time to consider the question whether the very decision he was about to take, could be regarded by other people facing similar charges throughout South Africa, as a breach of the principles of equality before the law or that it would be an abuse of process to discontinue charges against people of high profile or standing in the community.’

Ledwaba DJP (Pretorius and Mothle JJ concurring) came to the conclusion that the decision concerned was ‘irrational and should be reviewed and set aside’ (at [94]). The full bench made the following order: ‘The decision of the first respondent, dated 1 April 2009, to discontinue the prosecution of the case against the third respondent in accordance with the indictment served on him on 28 December 2007 is reviewed and set aside’ (at [97.2]).

For an overview of other cases dealing with the review of a decision to discontinue prosecution, see Chapter 1 in Commentary, sv Review of prosecuting authority’s decision to withdraw charges, and the validity of a mandatory interdict to prosecute.

ss 57 and 57A: Admission of guilt and payment of a fine

Section 57 of the Criminal Procedure Act provides for the payment of a so-called ‘admission of guilt fine’ prior to any court appearance, whereas s 57A deals with the payment of such fines after court appearance but before an accused has pleaded. See s 57A(1) and S v Mafukidzi 2015 JDR 0466 (GP). In this case the review court was requested to set aside plea proceedings so that the payment of an admission of guilt fine after a plea could be considered regular. The review court refused to do so. It was stated that only an irregularity could be corrected on review—and the only irregularity was the acceptance of the fine in breach of the provisions of  s57A(1). At [7.3] Bertelsmann J (Preller J concurring) made the following order: ‘The trial proceedings that commenced with the accused’s plea of not guilty must be concluded in accordance with the provisions of the Criminal Procedure Act ...’.

S v Du Plessis (unreported, GP case no A418/16, 28 January 2016) dealt with a different issue concerning  s57A. This matter was sent on special review by a senior magistrate who also submitted documents to the effect that the accused had at no stage been informed of the consequences of paying an admission of guilt fine (at [1]–[2]).

The accused in Du Plessis had to appear in court on a charge of negligent driving in contravention of s 63(1) of the National Road Traffic Act 93 of 1996. Her first court appearance was on 27 August 2012. She was not required to plead as her attorney was in the process of seeking a withdrawal of the charge. The matter was remanded to 3 October 2012 and on that day the prosecution was at this late stage prepared to determine an admission of guilt fine as permitted in s 57A(1), that is, after the accused’s court appearance but prior to plea.
In her affidavit which served before the review court, the accused said she paid the admission of guilt fine not because she thought she was guilty, but because she was scared to go through a trial as she had never before been involved in a criminal case and she wanted to get the matter out of the way. She also thought that she was really paying a traffic fine (‘verkeersboete’).

It was more than two years later that she discovered—when she applied for a new job—that her payment of the section-57A(1) admission of guilt fine had resulted in a criminal record, namely a previous conviction for negligent driving. In her affidavit she claimed that neither the prosecutor nor her attorney had informed her that by paying the fine she would acquire a criminal record. Her attorney confirmed her claim in an affidavit (at [6]).

At [7]–[9] Louw J (Tolmay J concurring) pointed out that a magistrate had as far back as 30 May 2013 confirmed that the accused’s payment of the admission of guilt fine was in accordance with justice as provided for in s 57A(4) as read with s 57(7). However, it was only long after this date that the accused became aware of the fact that her ignorance had resulted in a criminal record.

Louw J also pointed out that in respect of s 57 fines (payment without court appearance) there have been several cases in which the payment of an admission of guilt fine was set aside precisely because at the time of the payment of the fine the accused was not aware of the serious consequence of paying an admission of guilt fine.

At [10] reference was made to cases such as S v Claassen 2012 JDR 2524 (FB); S v Gilgannon 2013 JDR 2121 (GSJ) and S v Tong 2013 (1) SACR 346 (WCC).

In Du Plessis it was accepted that the accused never knew and was indeed never informed that payment would result in a previous conviction. This being the case, held Louw J, there was no reason why the accused in Du Plessis, which involved s 57A, should be treated differently from the accused in the three cases above which involved s 57. At [11]–[12] it was pointed out that the ‘ratio’ is the same and that the payment of the admission of guilt fine in Du Plessis had to be set aside in terms of s 304(4) as being not in accordance with justice.

The decision in Du Plessis should alert prosecutors to the need to ensure that all their section-57A(1) notices should not only stipulate that an admission of guilt fine can be paid, but should also contain a notice to the effect that payment of an admission of guilt fine would result in a conviction for criminal record purposes. This much has already happened in respect of section-57 notices. In S v Houtzamer 2015 JDR 0424 (WCC) at [19] Rogers J pointed out that as a result of the decision in Tong (supra), a section-57 written notice

‘... now incorporates an annexure, also signed by the accused, expressly warning that if he pays an admission of guilt fine he will be deemed to have been convicted and sentenced and will thus have a criminal record and that by admitting guilt he will be waiving his right to be sentenced only upon proof beyond reasonable doubt in a trial in which he would be entitled to confront his accuser, call witnesses and have legal representation.’

See also S v Parsons 2013 (1) SACR 38 (WCC) which is analysed in the discussion of s 57 in Commentary, sv Duty of peace officer to inform accused of rights and consequences.

s 60(11)(a): Bail and the relevance of conditions of detention in determining exceptional circumstances

Section 60(11)(a) of the Criminal Procedure Act provides that where an accused is charged with an offence referred to in Schedule 6 to this Act, the court shall order that the accused be detained in custody until dealt with in accordance with the law, unless the accused—after having been given a reasonable opportunity to do so—‘adduces evidence which satisfies the court that exceptional circumstances exist which in the interest of justice permit his or her release’. For a survey of case law dealing with ‘exceptional circumstances’, see the discussion of s 60 in Commentary, sv Section 60(11)(a) and the meaning of ‘exceptional circumstances’. Section 60(11)(a) has been held to be constitutional even though it places a formal burden on the accused to prove ‘exceptional circumstances’. See S v Dlamini; S v Dladla & others; S v Joubert; S v Schietekat 1999 (2) SACR 51 (CC).

In S v Panayiotou (unreported, ECG case no CA&R 06/2015, 28 July 2015) the appellant—who was in custody on charges listed in Schedule 6—advanced several grounds upon which it was submitted that the magistrate had misdirected herself in finding that he had failed to prove exceptional circumstances warranting his release as provided for in s 60(11)(a). One of the arguments was that the magistrate had ignored evidence to the effect that the appellant was being held in St Albans prison in conditions which were in breach of his right to dignity.

In argument on appeal the magistrate was criticised for a comment she made during the bail proceedings, namely that ‘prisons are never intended to be a holiday away from home’. It was submitted on behalf of the appellant that this was a ‘callous’ statement, displaying a disdain for the constitutional rights of the appellant. At [36] Goosen J observed that the magistrate’s remark was ‘certainly unfortunate ... [and] ... suggests that the concerns about conditions in prison are inconsequential’. However, Goosen J was also satisfied that upon a careful analysis of the magistrate’s reasoning regarding the appellant’s reliance on his conditions of detention, it was clear beyond doubt that the magistrate proceeded from the premise that ‘conditions of detention may give rise to violations of the rights to health, dignity and safety of inmates’.

At [37] it was noted that the magistrate specifically took judicial notice of the overcrowding in prisons and its effect. She had also considered the length of time that the appellant would have to spend in prison pending trial if bail were refused (at [38]).

In argument it was submitted that the magistrate’s failure to refer to S v Jacobs 2011 (1) SACR 490 (ECP) in her judgment meant that she attributed insufficient weight to the ‘deplorable conditions’ in St Albans. The decision in Jacobs was relied on by the appellant in his bail application and on appeal. In Jacobs Roberson AJ (as she then was) also took judicial notice of the overcrowding in South African prisons and held that in view of the violation of his dignity and threat to his health and safety, the appellant had to be released on conditions that did not include the payment of a sum of money (at [18]–[20]). The bail amount initially set by the magistrate in Jacobs, was nominal (R800 and later R500) but could not be afforded by the appellant, who was willing to abide by the court’s conditions. And it would appear that the crux of the decision in Jacobs was that release on bail on conditions which do not involve payment of bail money is a means of securing a situation where an accused is not subjected to the potential risks which result from overcrowding. See further Jacobs at [16] as well as the discussion of s 63A in Commentary, sv Introduction.
To return to Panayiotou: Goosen J concluded (correctly, it is submitted) that ‘the facts of Jacobs are wholly distinguishable from the present matter’ (at [38]) and that ‘Jacobs in any event does not constitute authority for the proposition that deplorable conditions of detention necessarily favour an accused person’s release on bail’ (at [40]). Goosen J concluded that the magistrate did consider the principle in Jacobs and the fact that this case was not mentioned in her judgment was ‘of no consequence’ (at [40]): the magistrate did consider conditions of detention as a factor to be weighed up when determining the presence or absence of exceptional circumstances.
At [40] Goosen J pointed out that release on bail is not the remedy for the failures of the Department of Correctional Services to secure proper conditions of detention. In this regard reference was made to what Comrie AJA said in S v Van Wyk 2005 (1) SACR 41 (SCA) at [9]. In this case it was held that the remedy of the accused detained in conditions infringing dignity and threatening health and personal safety would be to challenge the constitutional validity of his detention, or to seek a court order compelling the authorities to comply with the law. Reference can also be made to S v Mpofana 1998 (1) SACR 40 (Tk) where Mbenenge AJ (as he then was) pointed out that a bail application or renewed bail application was not an appropriate remedy for an accused alleging detention in a congested cell and refusal by the prison authorities to permit him to consult a medical practitioner of his choice.

In Panayiotou the appellant also alleged that his conditions of detention were such that he was being kept in ‘solitary confinement’ and that ‘this was a form of punishment imposed on a person who is presumed innocent until the contrary is proved at trial’ (at [41]). It would appear that his allegation that he was being kept in ‘solitary confinement’ was based on the fact that he was afforded one hour of exercise per day. Goosen J had several difficulties with this ‘solitary confinement’ argument and disposed of it as follows (at [42]):

‘The appellant’s case in his founding affidavit dealt at considerable length with the conditions that pertain in the communal cells where he was initially detained. However at the time of the bail application he had already been moved to a single cell because of concerns about his safety. He asserted that this amounts to solitary confinement. However no evidence was presented to substantiate this assertion. There is for instance no evidence that he has no contact with other inmates; that he has no communication with other inmates or prison officials; there is no evidence regarding restrictions on visits of family or legal representatives. The assertion that accommodation in a single cell therefore amounts to ‘‘solitary confinement’’ is unsubstantiated and amounts, in my view, to an emotive argument.’

Turning, once again, to S v Van Wyk (supra) Goosen J concluded that even if it were to be accepted that the exercise period fell short of constitutional and international standards, the appellant’s remedy ‘was not release on bail but an appropriate challenge to the Department of Correctional Services’ (at [42]).

s 73: Legal representative appearing for two accused who have a conflict of interest

In S v Tjale (unreported, GP case no A 265/15, 26 January 2016) the applicant alleged that several irregularities had occurred at his regional court trial where he and a co-accused were charged with rape, that is, contravening s 3 of the Criminal Law (Sexual Offences and Related Matters) Amendment Act 32 of 2007. The applicant was convicted and sentenced to life imprisonment in terms of s 51(1) of the Criminal Law Amendment Act 105 of 1977.
At the trial the applicant and his co-accused were represented by the same legal representative from Legal Aid South Africa. This, argued the applicant, was a serious irregularity because the legal representative must have been aware of the existence of a serious conflict of interest between the applicant and his co-accused. The applicant alleged that in his ‘warning statement to the police’ he had stated that he did have sexual intercourse with the complainant but that this intercourse took place because his co-accused had threatened him with a firearm (at [4]). However, the defence of the co-accused was entirely inconsistent with the defence of duress raised by the applicant: the co-accused’s defence was ‘a complete denial’ (at [4]).

The review court did not have the benefit of access to the ‘warning statement’ relied on by the applicant. It had been lost. However, the review court stated that it appeared from the trial record that the legal representative was in possession of the police docket containing the warning statement and that she was in possession thereof when, during the course of the trial and whilst the complainant was still under cross-examination, she drew up a section-112(2) statement in support of the applicant’s plea of guilty which replaced his plea of not guilty tendered at the beginning of the trial (at [7]). When the section-112(2) statement was handed in, the applicant refused to confirm the contents thereof and claimed ‘that it was not in accordance with the initial statement that he had made to the police and that he therefore disputed the contents of the statement’ (at [5]). At this stage the legal representative wanted to withdraw on account of the applicant’s denial of the section-112(2) statement. However, the presiding magistrate granted her an opportunity to consult with the applicant. When the trial resumed the applicant, at the request of the trial court, confirmed and signed the statement. This was after the legal representative had informed the court that she had explained to the applicant how she had drafted the section-112(2) statement ‘and that they went through the warning statement that he gave to the police’ (at [5]). In his affidavit submitted to the review court, the applicant stated that he did not know that his plea of guilty was inconsistent with his defence of duress. He also explained that although he signed and confirmed the statement, he was under the erroneous impression that the court was aware of his defence (at [6]).
Having regard to all the circumstances, the review court concluded (at [12]) that

‘the fact that the applicant never had an opportunity to put his version to the trial court resulted in the court a quo never considering his version of having been threatened at gunpoint to have sexual intercourse with the complainant. On the face of it, had these facts been conveyed to the court, the presiding magistrate may well have decided that the applicant’s version discloses a possible valid defence to the charge. Consequently, the fact that he was not afforded an opportunity to place this version before the court resulted in him not receiving a fair trial.’

The review court also noted the ‘disconcerting fact’ that according to the pre-sentence report the complainant herself had informed the probation officer that the co-accused had threatened the applicant at gunpoint to rape her (at [13]). At [14] Basson J (De Vos J concurring) stated as follows:

‘If the content of the pre-sentencing report is considered (although I am mindful of the probative value of the report) it does appear that it affords some credence to the applicant’s version that he was forced to have sexual intercourse with the complainant. At the very least, the presiding magistrate should, at that stage of the proceedings, have stopped all proceedings and should forthwith have had the record remitted for a special review to this court. This was not done.’

The applicant’s conviction and sentence were set aside; and the matter was referred to the Director of Public Prosecutions for a decision whether to prosecute the applicant afresh (at [16]).

All the difficulties that arose in Tjale because one lawyer defended two accused who had a conflict of interest, were avoidable. Consider the following:

(a)    The legal representative—quite contrary to the principles of legal professional privilege—attempted to do the impossible and impermissible, namely to defend two accused who had mutually exclusive defences. This created a clear conflict of interest. See the discussion of s 73 in Commentary, sv Withdrawal by legal representative: Conflict of interests between two or more accused represented by one lawyer. When did the lawyer become aware of the conflict? One is left with the impression that a thorough pre-trial interview with each of the accused would, prior to trial, have brought to the surface the serious conflict of interest that existed between the applicant and his co-accused in Tjale. In such an instance the legal representative would have been compelled to withdraw by terminating her services in respect of both the applicant and his co-accused. See S v Mathe 1996 (1) SACR 456 (N). The risk that information protected by legal professional privilege may be used against one of the accused interviewed, warrants the call for a complete withdrawal. See Van Niekerk, Van der Merwe & Van Wyk Privilegies in die Bewysreg (1984) 223.
(b)    A prosecutor is required to protect an accused’s constitutional fair trial right. See the discussion in Chapter 1 in Commentary, sv The prosecutor and the constitutional fair trial right of the accused. Clearly, the prosecutor in Tjale should have realised that, on the basis of the applicant’s warning statement in the docket, there was a very real risk of a conflict of interest should one attorney appear for both accused. It has been accepted that the prosecution may, where circumstances so require, suggest to the defence lawyer that the latter might have to consider withdrawal on account of a conflict of interest between his or her clients. See S v Naidoo 1974 (3) SA 706 (A). It would appear though that no such suggestion was made to the defence attorney in Tjale.
(c)    In S v Ngculu [2015] ZASCA 184 (unreported, SCA case no 438/2015, 30 November 2015) the trial judge had some time prior to the commencement of the trial asked defence counsel appearing for all seven accused if there was any conflict of interest between any of his clients. A judicial query of this nature makes good sense and must be encouraged whenever a defence lawyer appears on behalf of more than one accused. In Tjale no such question was put to the defence attorney; had it been put, the defence attorney (or the prosecutor or preferably both) would probably have been alerted to the problems that lay ahead which ultimately caused an irregularity vitiating the proceedings.
(d)    It is perfectly understandable that Legal Aid South Africa might at times wish to save costs and resources by instructing one lawyer to appear on behalf of two or more accused. However, it is submitted that where this route is followed, the defence lawyer concerned should be required to inform Legal Aid South Africa, prior to trial and in writing, whether or not a conflict of interest exists, or whether it is probable that a conflict might develop during the course of the trial. Of course, it is not always possible to predict whether a conflict will arise, but the requirement as suggested would at least alert defence lawyers to the need to make a careful assessment of the possible risk of a conflict.

s 73: The regularity of trial proceedings where a co-accused is represented by someone without the right of appearance

In S v Mpunga & others (unreported, GP case no A44/2016, 28 January 2016) three accused stood trial in the regional court on charges of contravening certain sections of the Prevention of Organised Crime Act 121 of 1998 (‘POCA’). During the course of the trial and after several witnesses had testified, it came to light that the representative of accused 2 and 3 was ‘not an admitted attorney’—a matter that was confirmed by the Law Society of the Northern Provinces (at [4]). The presiding magistrate correctly took the view that this was an irregularity warranting interference by the High Court. The matter was accordingly sent on special review as provided for in s 304 of the Criminal Procedure Act (at [1] and [4]).

On review Raulinga J (Tolmay J concurring) confirmed that ‘it is in the public interest that the defence in a criminal trial be undertaken by a person who has been admitted to practise ...’ (at [9]). In this regard reference was made to S v Mkhise; S v Mosia; S v Jones; S v Le Roux 1988 (2) SA 868 (A). See also the discussion of s 73 in Commentary, sv Section 73(3): The assistance of any other person.

However, in Mpunga the review court also had to consider the position of accused 1 who at all relevant times was indeed represented by a qualified lawyer. Taking its cue from decisions such as S v Mhlanga & others (unreported, GP case no A314/10, 29 April 2010) and S v Gwantshu & another 1995 (2) SACR 384 (E), the review court in Mpunga concluded that the proceedings had to be set aside in their entirety despite the fact that accused 1 had proper representation. In Mhlanga Southwood J, after having pointed out that it was not always easy or possible to disentangle the facts and determine the effect the unqualified person had on the trial of the other accused, observed as follows (at [9]): ‘Since the object of the rule is to safeguard the integrity of the proceedings, it seems to me to be essential that the entire proceedings be set aside.’ In this respect Southwood J relied on Gwantshu where Mullins J (Lang AJ concurring) decided to set aside the entire proceedings as far as both accused were concerned. The decision in Gwantshu was taken without perusing the record of the part-heard matter, regardless of the fact that one accused was represented by a qualified lawyer and irrespective of the wishes of the prosecution and the views of the trial court (at 386a–d).

But Gwantshu is no longer the solid authority it was considered to be in Mhlanga and Mpunga. Some four months prior to the decision in Mpunga, Stretch J (Cossie AJ concurring) concluded in S v Swapi & others (unreported, ECB case no 14/14, 1 September 2015) that Gwantshu was wrongly decided. At [22] in Swapi Stretch J took the view that an approach more nuanced than the one in Gwantshu was required: in trials of co-accused the proceedings in respect of an accused who was represented by a qualified attorney need not necessarily be set aside, particularly not (a) where the trial record ‘does not call for such a course of conduct’; (b) where all the accused, the prosecutor and the presiding judicial officer do not consider ‘such an approach necessary, convenient or in the interests of justice’; (c) where it appears to be in the interests of justice to commence afresh against the affected accused only; (d) where ‘a separation of trials with appropriate measures is unlikely to prejudice the accused or the administration of justice’.

For an analysis of Swapi as compared to Gwantshu, see the discussion of s 73 in Commentary, sv Representation of one co-accused by someone without the right of appearance: Impact on trial proceedings. For a further analysis of Swapi in the context of a separation of trials, see the discussion of s 157 in Commentary, sv Section 157(2): Separation of trials under the sub-heading Separation of trials where one of accused is represented by someone without right of appearance.

s 112(1)(b): The undefended accused’s plea of guilty and questioning by the court

The purpose of the judicial questioning as provided for in s 112(1)(b) of the Criminal Procedure Act is to test the validity of the accused’s plea of guilty. See the cases referred to in the discussion of s 112 in Commentary, sv Section 112(1)(b): Purpose of questioning.

In S v Mokhati (unreported, FB case no 158/2015, 18 November 2015) an uneducated woman with a low income from manual labour had pleaded guilty to dealing in 16,4 kilograms of dagga in contravention of the Drugs and Drug Trafficking Act 140 of 1992. She had no legal representation. After questioning by the presiding officer as provided for in s 112(1)(b), the accused was convicted; and after an inadequate inquiry concerning her personal circumstances and ability to pay a fine (at [6.e]), she was sentenced to R3 000 or 12 months’ imprisonment and a further 18 months’ imprisonment wholly suspended (at [2]).

On automatic review Opperman AJ (Jordaan J concurring) pointed out that the definition of ‘dealing’ in Act 140 of 1992 was not explained to the undefended accused. An explanation of this nature was necessary because ‘it cannot be presumed that an accused knows that to carry dagga from one location to another implies dealing in drugs in terms of the law’ (at [6.c]). However, of more concern to the review court was the following question put by the magistrate to the accused: ‘Do you also confirm that, do you know that dagga is an undesirable dependence producing substance?’ Here, too, the court pointed out that the magistrate had erred in assuming that the accused knew what ‘an undesirable dependence producing substance’ was (at [6.d]). One can also add the following argument in support of the review court’s finding that the question was irregular: the question—even though clumsily phrased—was also in the form of a leading question. Leading questions should in principle be avoided by the court in the course of its section-112(1)(b) questioning of an accused who has pleaded guilty. See S v Mahlasela 2005 (1) SACR 269 (N) at 271j–272a. In S v Phundula; S v Mazibuko; S v Niewoudt 1978 (4) SA 855 (T) it was noted that unrepresented uneducated accused persons who pleaded guilty were inclined to give an affirmative answers to leading questions which covered the elements of the alleged crime. This much, it would appear, also happened in Mokhati.

The conviction and sentence in Mokhati were set aside. The decision in Mokhati is supported and is in line with cases decided shortly after the Criminal Procedure Act came into operation. See S v Phikwa 1978 (1) SA 397 (E) where it was held that judicial questioning in terms of s 112(1)(b) should avoid the use of unnecessary legal terminology (at 398). Mokhati should also be read with the recent decision in S v Ndau (unreported, GP case no A292/15, 6 May 2015) where Bertelsmann J said that it was unfair to confront an undefended and barely literate accused with legal phrases and statutory definitions not couched in ordinary language—and then to expect an informed response. In Ndau the unacceptable question was: ‘Nyaope is an undesirable dependence producing substance?’

There is another aspect of the lower court proceedings in Mokhati which requires comment. The accused—who had no legal representation—was on the day after her arrest convicted of and sentenced for a serious crime. The presiding magistrate explained to the review court that the accused’s right to legal representation was explained ‘before the case went on record’ but not ‘again for [the] record’ (at [5]). What happened in Mokhati clearly fell short of the following principles established in our case law:

(a) The record must indicate precisely what was conveyed to the accused and what her responses were.
     See S v Sibiya 2004 (2) SACR 82 (W) 90b–c; S v Sikhipha 2006 (2) SACR 439 (SCA) at [9]–[10].
(b) A court should encourage an accused to make use of legal aid where the charge is a serious one. 
     See S v Mofokeng 2013 (1) SACR 143 (FB) at [17.10].
(c) A court must be satisfied that an accused’s choice to conduct her own case, is an informed one.  
     See S v Solomons 2004 (1) SACR 137 (C) at 141ef).

s 162: Witnesses sworn in by interpreter and not by the presiding judicial officer

S v Pilane 2016 (1) SACR 247 (NWM)

All three state witnesses in S v Pilane had been sworn in by the interpreter and not by the presiding officer. This was held by Hendricks J (Djaje J concurring) to be improper. The provisions of s 162, the court insisted, were peremptory, as it was required that the oath ‘shall be administered by the presiding judicial officer or, in the case of a superior court, by the presiding judge or the registrar of the court’.

If the oath is not administered in the prescribed way, the ‘witnesses are not properly sworn in and their evidence is therefore inadmissible’ (at [6]). It had been argued, on the strength of certain remarks by Leveson J in S v Orphanou & others 1990 (2) SACR 429 (W) at 431h–j, that the provisions of s 162 should be interpreted more loosely. In that case Leveson J said that, under the maxim subsecuta observatio in contemporanea expositio, it was permissible to look to the usual practice of the persons who use a statute in order to see how that statute is interpreted. Leveson J considered ‘that in this Division the Judges interpret the section as referring to clerks and consider the phrase ‘‘Registrar of the Court’’ as sufficiently wide to include any court official whose duties are not so exacting that they cannot be appointed by the Judges for the limited purpose of administering the oath to witnesses from time to time’.
Hendricks J found himself in respectful disagreement with this contention. The use of the word ‘shall’ was a clear indication, in the judge’s view, that the provisions were intended to be peremptory. The court was, moreover, bound by the decisions in S v Raghubar 2013 (1) SACR 398 (SCA) and S v Matshivha 2014 (1) SACR 29 (SCA) (discussed in Commentary in the notes to s 162) to give s 162 this interpretation. The court agreed, too, with what the Supreme Court of Appeal said was the effect of non-compliance with s 162: that the ‘testimony of a witness, who has not been placed under oath properly, has not made a proper affirmation or has not been properly admonished to speak the truth as provided for in the Act, lacks the status and character of evidence and is admissible’ (in Matshivha at [10]; in Pilane at [8]).

ss 167 and 186: When should a court subpoena or recall a witness, and should the parties be given an opportunity to address the court?

S v Masooa [2016] 2 All SA 201 (GJ)

In this case Spilg J considered at some length the role, ambit and practical application of ss 167 and 186. Section 186 gives the court the power at any stage of criminal proceedings to subpoena any person as a witness and, further, makes it mandatory to do so ‘if the evidence of such witness appears to the court essential to the just decision of the case’. Section 167 gives the court the power to recall and re-examine any person, including the accused, already examined at the proceedings and, again makes it mandatory to do so if his or her evidence ‘appears to the court essential to the just decision of the case’.

Spilg J accepted that these sections give a criminal court a more inquisitorial role than that enjoyed by a civil court. He endorsed the view expressed by Curlewis JA in R v Hepworth 1928 AD 265 at 277 that a criminal trial is not a game and a judicial office is not merely a figurehead: he or she has not only to direct and control the proceedings according to the recognised rules of procedure, but must see that justice is done. The present case raised three concerns. First, calling for extra evidence required the exercise of great caution since it may be construed as descending into the arena. Second, the court had to be satisfied that the further evidence would probably, not just possibly, affect the outcome. And, third, since it was well recognised that the provisions of s 186 are ‘in tension with the obligation of the State to prove its case beyond reasonable doubt and with the presumption of innocence’, it was necessary for a court to adopt absolute impartiality in the matter. If further evidence is directed the court must ensure that strict neutrality is maintained in the way questions are asked, and the accused must also be afforded an opportunity to call further evidence in rebuttal should any potentially adverse evidence be produced. Spilg J raised the question of the constitutional soundness of s 186 but, since the issue was not raised, did not address it.

A question that was considered was whether the court had to afford the parties an opportunity to address it on whether the court should call witnesses in terms of its powers under s 186. Spilg J referred to S v Karolia 2006 (2) SACR 75 (SCA) where it was held to be unnecessary to do so, but pointed out that Heher JA (who delivered a minority dissent in Karolia but did not question this particular point) had said, in S v Gabaatlholwe 2003 (1) SACR 313 (SCA), where he spoke for the court, that the views of the parties should always be established before a court decides to call a witness. In his opinion (at [7]), ‘[t]he parties will often possess insights into the contribution which a witness could make not apparent to the Judge or magistrate and their views should always be canvassed before the decision is taken’. He added that the ‘best indication to the trial court of the importance that a party attaches to calling a witness is the assiduity which that party applies to ensuring that the witness is available to it’.

In Masooa the court accepted that, since everything turned on the workings and controls of a motorcycle, it was essential that the judge and assessors had an equal working knowledge of those issues, and that calling for expert evidence in that regard would be permissible under the mandatory or, at least the discretionary part of s 186. Without this evidence, said Spilg J, there was a real concern that assumptions might be made and perceptions not addressed. It was important for every member of the court to be placed on an equal footing in this regard and for any preconceptions to be addressed. It was held further that a key witness, to whom the accused had failed to put details of his version of events, should be recalled under s 167, since this was essential for the just decision of this case. Spilg J (at [79]) pointed to two decisions of the Supreme Court of Appeal which demonstrated that that court had ‘favourably considered the calling of a number of witnesses by the trial court’: S v Karolia (supra) and S v Gerbers 1997 (2) SACR 601 (SCA).

s 168: Right to have trial postponed for legal representation cannot be abused

S v Dibakoane 2016 (1) SACR 532 (GP)

Section 168 gives a court the power to adjourn the proceedings at a criminal trial ‘if the court deems it necessary or expedient’ to do so. Such an adjournment will normally be in order to afford an accused an opportunity to obtain legal representation. In Dibakoane the court refused to postpone the proceedings for this purpose. The accused then refused to conduct the trial on his own, remained passive throughout the proceedings, refused to cross-examine State witnesses in spite of being advised of this right repeatedly, and declined to testify or lead any witnesses in his own defence.

These circumstances would ordinarily induce a court of appeal to conclude that the appellant had not received a fair trial and that he had wrongly been denied his right to legal representation. This was, however, not so in Dibakoane.

In that case there had, previously, been a number of postponements for the appellant to obtain legal representation. When the matter eventually proceeded, his objection was that he had insufficient time to familiarise himself with the docket and that he did not understand the statements which were written in Afrikaans. These were interpreted and explained to him. He then again ‘changed the goalposts’ by maintaining again that he could not proceed without legal representation. As Ismail J (with whom Swartz AJ agreed) pointed out, it was evident from the record that the appellant ‘was not desirous that the matter should proceed and he sought to frustrate the commencement of the trial at all costs’ (at [15]). The magistrate, who showed a great deal of patience and restraint, sought to accommodate the appellant further by postponing the matter until the next day to allow the appellant to study the statements which had been interpreted for him. On that day he declined to plead to the changes in the absence of legal representation. A plea of not guilty was entered, and the trial continued.

The court of appeal held that it could not be said that the appellant had had an unfair trial. Although the right to legal representation was a fundamental right, Ismail J pointed out that the right to a fair trial meant that there should be fairness to both sides, the accused and the State. Although legal representation was a fundamental right, one that was perhaps, of all the rights in the Bill of Rights, closest to being an absolute right, it nevertheless had its limitations. It could not be used as a stratagem to delay the trial or frustrate its commencement. And it could not be used to allow matters to be postponed indefinitely in order for an accused to raise funds to obtain representation.

The appellant in this case, said the court (at [23]), ‘adopted a stance similar to Custer’s last stand at the Alamo’. He was not interested in securing a lawyer appointed by legal aid and was prepared only to secure a lawyer personally instructed by him. Finance was the problem, but this was not a justifiable reason for postponing a trial indefinitely.

There was ‘no one other to blame than himself for his predicament’, since he was offered legal aid but declined it. He was, in the end, the ‘cause of his own undoing’ for, as Ismail J put it, ‘you can take a horse to water but cannot make it drink’ (at [23]). The trial was a fair one and no misdirection had taken place.

ss 190 and 210: Previous consistent statements and sexual complaints: Complainant not testifying

S v Mathikinca 2016 (1) SACR 240 (WCC)

The appellant in Mathikinca had been convicted of contravening s 3 of the Criminal Law (Sexual Offences and Related Matters) Amendment Act 32 of 2007 based largely on a report made by the complainant, a 4-year-old girl, to her mother. The complainant was not, however, called as a witness at the trial. The court of appeal found that the report should not have been admitted in evidence.

It was, said Fourie J (Van Staden AJ concurring), trite law that the fact of a complaint and its terms are admissible in proceedings relating to sexual offences as establishing consistency in the complainant’s evidence, and therefore supporting her credibility (see S v Hammond 2004 (2) SACR 303 (SCA) at [17]). If, however, the complainant gives no testimony at all, neither the terms of the complaint nor the fact that it was made can ordinarily be admitted. Fourie J at [10] cited Zeffertt and Paizes The South African Law of Evidence 2 ed (2009) at 452 to explain this principle: the complaint is inadmissible ‘precisely because it would be absurd to regard a statement as being consistent with something that does not exist’ (see, too, R v Kgaladi 1943 AD 255 at 261).

Do the provisions of s 58 of the Criminal Law (Sexual Offences and Related Matters) Amendment Act assist the state in circumstances such as this? That section provides that ‘[e]vidence relating to previous consistent statements by a complainant shall be admissible in criminal proceedings involving the alleged commission of a sexual offence’ (emphasis added), and is silent about the probative purpose of the evidence: does it go only to consistency or may it be used to assert the truth of its contents? Fourie J accepted the proposition of Zeffertt and Paizes (at 452) that the section in effect restates the common-law position, with the result that sexual complaints were admissible only to show consistency.

Can the provisions of s 3 of the Law of Evidence Amendment Act 45 of 1988 be invoked to allow evidence of this kind to be received as an exception to the hearsay rule, to prove the truth of what the complainant said in her report? There is, in principle, no reason why they should not, provided that all the conditions and requirements of s 3 are satisfied. In Mathikinca Fourie J (at [12]) considered the provisions of s 3. However, since the State had not attempted to lay any basis for invoking this section, and since it could not be said that the defence had specifically agreed to the introduction of the complaint in terms of s 3(a), the State could not rely upon the evidence of the complaint. The remaining evidence, which was circumstantial, was insufficient to support the conviction, which was accordingly set aside.

ss 217 and 218: Statements made in the course of a pointing out after accused denied legal assistance

S v Mabaso 2016 (1) SACR 617 (SCA)

The evidence for the State in this case, in which the appellant had been convicted of murder, did not have much going for it. The appellant had agreed, said the police witnesses, to making a formal pointing out. The circumstances surrounding this event were rather suspicious, and the appellant alleged that he had not acted freely and voluntarily. During the pointing out, he made certain statements which amounted to a confession. The pointing out evidence was ruled admissible by the trial court, but it was clear to the Supreme Court of Appeal that the reception of this evidence was problematic.

First, it was clear that the appellant had expressed a wish to have immediate legal representation and that this request was refused. He was then asked if he wished to continue with the pointing out and his affirmative answer was qualified by an express request to speak to his sister-in-law, who was a captain in the South African Police Service, clearly for the purpose of obtaining legal representation. This request was honoured, but only after the pointing out had taken place.

This irregularity, said Fourie AJA (with whom Leach and Zondi JJA agreed), ‘amounted to a flagrant disregard of the appellant’s constitutional right to legal representation’ (at [10]). The judge stressed (at [12]) that although s 218(2) on the face of it entitles the prosecution to adduce evidence of the pointing out by an accused notwithstanding that the pointing out forms part of an inadmissible confession, ‘our courts have often warned that s 218(2) does not authorise the production of a confession in the guise of a pointing-out’ (see S v Mbele 1981 (2) SA 738 (A) at 743C; S v Magwaza 1985 (3) SA 29 (A) at 36 and Commentary in the notes on s 218).

The State had correctly conceded ‘that the circumstances giving rise to the pointing-out, as well as the manner in which [the policeman] questioned the appellant and obtained the damning answer from him, constituted a confession being elicited from him’. The issue was, then, ‘whether evidence of either the pointing-out itself, without regard being had to the appellant’s answer to the question he was asked, or such answer, was admissible’. The court took into account the fact that the appellant, after speaking to his sister-in-law, refused to make a formal confession the following day when he was taken to another senior police office to do so. The inference was, said Fourie AJA, ‘irresistible that this was due to the advice she had given him and that, if he had seen her before the pointing-out, he would have remained silent or not done the pointing-out at all’. Accordingly, it could ‘hardly be said that the admission into evidence of the confession at the pointing-out, made only after the appellant had been denied legal assistance and questioned ... was not detrimental to the administration of justice’. As a result ‘both the pointing-out and the confession probably fell to be excluded under s 35(5) of the Constitution’ (at [14]).

But no final decision had to be taken on that issue since the evidence suffered from yet another fatal deficiency. The police officer conducting the process had, since keeping the written notes of the pointing out, been assisted by an interpreter, who recorded in English the ipsissima verba of the appellant, who spoke in isiZulu. It appeared that these notes had not been read back to the appellant. The appellant denied making the alleged incriminating statements and he at no stage confirmed the correctness of the police officer’s pointing out notes. As a result those notes constituted no more than inadmissible hearsay statements and had to be excluded on this basis alone.

s 252A: Undercover operations and the Constitution

S v Singh & others [2016] ZASCA 37 (unreported, SCA case no 862/2015, 24 March 2016)

The admissibility of evidence relating to trapping and undercover operations in criminal proceedings is governed, in the main, by s 252A of the Criminal Procedure Act. In Singh, however, no reliance was placed on that section by the appellants, who based their arguments against receiving the evidence squarely on s 35(5) of the Constitution. The evidence in question involved an undercover operation conducted by the police which had as its objective the infiltration and detection of the criminal activities of a syndicate responsible for several hijackings and armed robberies of very large trucks on the highway between Johannesburg and Durban. Conventional investigation techniques had not succeeded in identifying and apprehending those at the head of this syndicate, and several incidents of serious violence had already taken place prior to the operation, which used both an ‘in place informer’ and a ‘trap agent’, the latter being employed to transport the stolen goods.

As a result of the operation and the involvement of the trap agent in six incidents, the appellants were arrested and indicted on twenty counts covering a wide range of offences, including attempted murder and robbery with aggravating circumstances. The trap did not initiate any of the criminal activities and his main role was to convey the wheels stolen from the hijacked trucks to a certain address in return for payment. He had, at all times during the six incidents, activated both video cameras and a tracking device.

The trial court held the evidence to be admissible and rejected claims that it fell foul of the requirements of s 252A of the Criminal Procedure Act. In the appeal the appellants disavowed any reliance on s 252A and argued for exclusion on the strength of only s 35(5) of the Constitution, which provides that ‘[e]vidence obtained in a manner that violates any right in the Bill of Rights must be excluded if the admission of that evidence would render the trial unfair or otherwise be detrimental to the administration of justice’. Reliance was placed specifically on the latter part of that section, it being argued that the alleged violations were detrimental to the interests of justice.

What distinguishes this case from many other is that the appellants contended that the violations affected not their own rights but, rather, those of the public at large. They argued that the conduct of the state was detrimental to the administration of justice in that the state ‘undertook the operation whilst aware that members of the public had already been exposed to serious violence, and that it continued with the operation whilst aware of a real possibility of further exposure to such violence’ (at [15]).

Tshiqi JA, who delivered the judgment of the court, pointed out that the inquiry as to whether the admission of evidence would be detrimental to the administration of justice centred around the public interest and that, it being a purely legal question, it did not necessitate any questions being asked about the incidence and quantum of proof required to discharge the onus of proof. What was needed was a ‘value judgment’ (see S v Pillay & others 2004 (2) SACR 419 (SCA) at 447H).

Tshiqi JA invoked the balancing of interests endorsed in cases such as Key v Attorney-General, Cape Provincial Division 1996 (4) SA 187 (CC) at [13], S v Mthembu 2008 (2) SACR 407 (SCA) at [26], S v Tandwa 2008 (1) SACR 613 (SCA) at [118] and S v Mphala 1998 (1) SACR 654 (W) at 657G–H. In resolving the delicate balance between the respect for the Bill of Rights by the police and the respect for the judicial process by the man in the street, Tshiqi JA made specific mention of the list of factors identified by Zeffertt and Paizes (The South African Law of Evidence 2 ed (2009) at 747–757) as considerations which the cases show may be taken into account in determining whether the reception of evidence is detrimental to the administration of justice. These are: (i) the bona fides of the investigation; (ii) the nature and seriousness of the violation of the accused’s rights; (iii) considerations of urgency and public safety; (iv) the availability of alternative, lawful means of obtaining the evidence in question; (v) the deterrent function of the courts in excluding improperly obtained evidence; (vi) the nature of the evidence; and (vii) the fact that the evidence would inevitably have been discovered even if improper means had not been employed. But these factors, Tshiqi JA noted (at [18], were ‘merely guidelines’ and the list was ‘not exhaustive’, since every case depended on its own facts.

The court held that the attack by the appellants on the operation could ‘simply be dismissed on the basis that it was meant to protect the very public alleged to have been exposed to violence’ (at [19]). The crimes committed by the syndicate were becoming increasingly violent; patrols were ineffective; the very economy of the country was threatened by the criminal activity in question; and it was necessary for the undercover operation to continue as long as it did in order to detect and apprehend the master mind of the syndicate. The trap did not initiate any of the offences and played no role beyond what had been planned already by members of the syndicate. He merely followed orders. There was thus ‘no close ... connection between the rights violation and the criminal acts of the syndicate that exposed the members of the public to serious acts of violence’.

There was also no suggestion that the operation was conducted ‘in a flagrant disregard of the constitutional rights of possible victims or other members of the public’ (at [21]). On the contrary, the evidence showed that the authorities were ‘very cautious’. The trap agent performed his tasks in terms of guidelines approved by the Director of Public Prosecutions and acted under the supervision of the Director and the investigation team. He had to report his activities to that team and the team was, in turn, obliged to comply with the guidelines set by the office of the DPP.

The court stressed that public opinion was a relevant consideration in determining whether a violation was detrimental to the administration of justice (see S v Tandwa at 649). And, said Tshiqi JA (at [22]), ‘the public would baulk at the idea that the law enforcement agencies failed to take bona fide measures aimed at effective detection of such an organised crime syndicate because of the fear that there may be danger to the public, specifically in the present circumstances where the crimes would have taken place irrespective of the operation’. The public reaction would, indeed, have been one of ‘shock, fury and outrage’ (see S v Ngcobo 1998 (10) BCLR 1248 (N) at 1254G).

The challenge to the admissibility of the evidence was, accordingly, unsuccessful.

 (ii) Sentencing

s 276B: The fixing of a non-parole period as part of a sentence of imprisonment

Section 276B(1)(a) of the Criminal Procedure Act provides that where an accused is sentenced to imprisonment for two years or longer, the court may—as part of the sentence—determine a period during which the accused person shall not be placed on parole. In S v Mtintso (unreported, GP case no A1038/2013, 21 April 2015) it was confirmed that a sentencing court misdirects itself when a non-parole period is fixed in circumstances where there is ‘nothing ... to justify a departure from the general rule that it is for the correctional and parole authorities and not the court to determine when, if at all, a prisoner should be released on parole’ (at [8]). See further the discussion of s 276B in Commentary, sv Case law: Aggravating factors and the requirement that there must be exceptional circumstances.
In terms of s 276B(1)(b) the non-parole period ‘may not exceed two thirds of the term of imprisonment imposed or 25 years, whichever is the shorter’. In S v Selli [2015] ZASCA 173 (unreported, SCA case no 220/15, 26 November 2015) a sentence imposed by a regional court magistrate was set aside because a non-parole period of four-fifths of an effective 15 years’ imprisonment fixed by the magistrate was in breach of s 276B(1)(b).

The Supreme Court of Appeal has on more than one occasion identified the fixing of a non-parole period (as provided for in s 276B) as an increase in penalty which cannot be applied to crimes committed prior to the coming into operation of s 276B. See S v Zono [2014] ZASCA 188 (unreported, SCA case no 20182/2014, 27 November 2014) at [3]. See also the discussion of s 276B in Commentary, sv Section 276B: No retrospective operation. It is therefore important to note that s 276B came into operation only on 1 October 2004. See GG 26808 of 1 October 2004. Before this date there was no legislative provision in terms of which courts could stipulate a non-parole period. In the absence of such a provision some sentencing courts nevertheless resorted to the doubtful practice of ‘recommending’ (as opposed to ‘fixing’) non-parole periods. Indeed, appeals where such recommendations were made still crop up.

In S v Mohammed (unreported, GP case no A340/15, 22 January 2016) the trial judge, in sentencing the accused on 26 September 1997 for crimes committed in 1996, had imposed an effective sentence of 40 years’ imprisonment. However, the trial judge had also recommended that the appellant had to serve 30 years ‘before being eligible for parole’ (at [2]).

On appeal Legodi J (Prinsloo and Raulinga JJ concurring) noted that ‘parole was within the discretion of the executive’ (at [38]) and that the Supreme Court of Appeal has been critical of non-parole periods (at [40]). See also the discussion of s 276B in Commentary, sv The judiciary and the executive.

The full bench in Mohammed also pointed out that in making the recommendation which he did, the trial judge was fully aware that the officials of the Department of Correctional Services are better placed than a court to determine when a particular prisoner could be placed on parole and that these officials would—despite the recommendation—carry the ultimate responsibility for determining release on parole (at [41]). But the full bench thought that the mere recommendation itself could cause problems (at [39]):

‘In the present case, the appellant will be completing half of his 40 years imprisonment in 2017. The recommendation for non-parole period before the appellant had served 30 years imprisonment, could present a problem or uncertainty to the prison authority which might feel obliged to consider the appellant’s release on parole only in 2027 when the appellant shall have served 30 years of his term of imprisonment.’

At [41] it was also pointed out that the non-parole period as recommended, was ‘too long’ and had ‘the potential to interfere with the Department’s discretion to integrate much earlier into the community a well-behaved prisoner’. The full bench accordingly set aside the recommendation concerned and specifically ordered that the appellant was ‘entitled to be considered for parole in accordance with applicable legislative framework’ (at [43.2]).

In S v Hendricks (unreported, WCC case no A420/14, 18 February 2015) the trial judge, having imposed an effective sentence of 44 years’ imprisonment on the appellant for crimes committed in 1998, also strongly recommended to the Commissioner of Correctional Services that the appellant should not be considered eligible for parole until such time as he had served at least 25 years of his sentence. Here, too, a full bench considered the matter and came to the conclusion that the recommendation of the trial judge could not stand. At [20] Gamble J (Blignault and Mantame JJ concurring) stated as follows:

‘It is therefore important, in my view, that when imposing sentence judicial officers should avoid consideration of the fact that the accused may serve a lesser sentence than that imposed by the court because of the leniency of, or due to policy considerations applied by, the Department of Correctional Services. The sentence imposed is the sentence which the court believes is just in the circumstances due regard being had to the general principles applicable to sentencing. Consideration of the possibility of early parole is an impermissible incursion into the functions of the Executive and will constitute a misdirection. Insofar as [the trial judge] appears to have been of the view that the appellant was an offender deserving of spending at least 25 years in prison, he was obliged ... to impose that effective period of imprisonment.’

s 280(2): Order that a later sentence of imprisonment should run concurrently with an earlier one

Section 280(2) of the Criminal Procedure Act provides that sentences of imprisonment ‘shall commence the one after the expiration, setting aside or remission of the other ... unless the court directs that such sentences of imprisonment shall run concurrently’. An order that some sentences, or parts thereof, should run concurrently, is to ensure that the cumulative effect of two or more sentences of imprisonment does not result in excessive punishment. See further the discussion of s 280 in Commentary, sv Concurrent sentences.

In Thabiso v Minister of Justice & others (unreported, GJ case no 06542/2015, 1 April 2016) a magistrate had ordered on 4 April 2016 that a sentence of 10 years’ imprisonment imposed by him (hereafter the ‘later sentence’) had to run concurrently with a sentence the accused was already serving (hereafter the ‘first sentence’). The first sentence was a sentence of 15 years’ imprisonment imposed by another court on 11 May 2004. But on that very day the accused had escaped from custody. He only started serving this first sentence on 3 October 2007 when he was rearrested. On 14 April 2014—when he had served 6 years and 6 months of the first sentence—the later sentence of 10 years’ imprisonment for robbery committed while he was on the run, was imposed. ‘At issue’, said Satchwell J at [10], ‘is whether or not the 10 years is encapsulated within the remaining 81/2 years of the first sentence or whether it runs from 2014 parallel with the remaining portion of the first sentence but continues beyond the first sentence which terminates earlier than the later sentence’.

At [11] it was noted that the word ‘concurrent’ is not defined in the Criminal Procedure Act. Satchwell J suggested, nevertheless, that ‘where one entity runs ‘‘concurrently’’ with another there is ... a parallel relationship running in tandem’ which meant that where prison sentences run concurrently ‘a shorter sentence runs at the same time as the longer sentence, is encapsulated within and is usually bounded by the beginning and the ending of the longer sentence’ (at [13]). But the facts in Thabiso were such that the later sentence turned out to be the longer sentence and could not be encapsulated within and bounded by the termination date of the earlier sentence which had 8 years and 6 months to run. No relevant South African cases could be found.

At [15] Satchwell J refused to find that the later sentence of 10 years was to terminate upon expiry of the remaining 8 years and 6 months of the first sentence. Her reasons were as follows:

First, the sentencing magistrate was fully aware that the accused had only 8 years and 6 months of his first sentence to serve: ‘One cannot assume that a judicial officer was ordering an absurd result ... that 10 years should fit into 81/2 years’ (at [16]).

Second, there is no provision in the Criminal Procedure Act to the effect that a later sentence, if ordered to run concurrently with a first sentence, must be entirely subsumed within the first sentence (at [17]).
Third, although the word ‘concurrent’ indicates that the first sentence and later sentence ‘run in parallel while they operate at the same time’, there is nothing to imply that the later (and longer) sentence cannot continue on its own upon termination of the first sentence (at [18]).

The decision in Thabiso is supported. The magistrate’s order of concurrency was to the benefit of the accused and was permissible and necessary to avoid the unduly harsh punishment that would have followed if the two sentences were to run consecutively. The fact that a small portion of the later sentence remained to be served upon termination of the earlier sentence, is an inevitable consequence of the bigger and overall benefit of the order of concurrency. At [19] Satchwell J explained that

‘the applicant would serve 15 years of the first sentence and thereafter the remaining 11/2 years of the 10 year sentence instead of 15 years of the first sentence and thereafter 10 years of the later sentence. The impact of the subsequent sentence has certainly been ameliorated by the learned magistrate.’
 (iii) Appeal and Review

Appeal: Can two judges sitting as a review court grant leave to appeal to the Supreme Court of Appeal?

The above question was addressed in S & another v De Villiers [2016] ZASCA 38 (unreported, SCA case no 20732/14, 24 March 2016) which concerned an appeal to the Supreme Court of Appeal against the dismissal of a review application brought by the appellant in the High Court where two judges presided. The High Court had granted leave to appeal to the Supreme Court of Appeal. The appellant’s attorneys doubted whether the High Court concerned had the power to grant such leave. After some correspondence between the registrar and the appellant’s attorneys, the appellant also filed a petition to the Supreme Court of Appeal for leave to appeal. As a result of the appellant’s two-pronged approach, the parties were requested to address as a preliminary aspect whether the High Court had ‘the requisite power to grant leave to appeal to this court, in view of the provisions contained in s 16(1)(b) of the Superior Courts Act [10 of 2013]’.

Majiedt JA (Fourie and Baartman AJJA concurring) noted at [3] that a division of the High Court which sits on review with two judges presiding, is a court of first instance as contemplated in s 16(1)(a) of the Superior Courts Act; and it was held that leave had therefore properly been granted in terms of s 16(1)(a)(ii) of this Act. At [3] it was also said that s 16 was applicable because ‘[t]he review before us is regulated by ... rule 53 [of the Uniform Rules of Court] ... [and] is not regulated by the Criminal Procedure Act ... or by any other criminal procedural law as envisaged in s 1 of the [Superior Courts] Act’.

It was accordingly ruled that the matter was properly before the court ‘on appeal’ and that the appellant’s petition was ‘unnecessary and should be regarded as superfluous’ (at [3]).
For a further Supreme Court of Appeal decision which also concerned ss 1 and 16 of the Superior Courts Act, see Director of Public Prosecutions, Gauteng v Mphaphama 2016 (1) SACR 495 (SCA) which is discussed elsewhere in this edition of Criminal Justice Review, sv s 316B: Can the State appeal to the Supreme Court of Appeal against a sentence imposed by the High Court sitting as a court of appeal?

s 316B: Can the State appeal to the Supreme Court of Appeal against a sentence imposed by the High Court sitting as a court of appeal?

In Director of Public Prosecutions, Gauteng v Mphaphama 2016 (1) SACR 495 (SCA) the Supreme Court of Appeal was required to address the above issue. The Director of Public Prosecutions (‘DPP’), as representative of the State, sought special leave to appeal against a sentence imposed by the Gauteng Local Division sitting as a court of appeal. The High Court had reduced to 20 years a sentence of life imprisonment imposed by the regional court. The Supreme Court of Appeal, however, directed that the DPP should first argue whether the decision of the High Court was indeed appealable. For purposes of this directive, the attention of the parties was drawn to the decision in DPP Western Cape v Kock 2016 (1) SACR 539 (SCA).
In Mphaphama Willis JA (Majiedt JA and Baartman AJA concurring) pointed out at [8] that the DPP could not rely on s 316B of the Criminal Procedure Act because this section is confined to the situation where the sentence against which the DPP seeks to appeal, was imposed by a superior court sitting as a court of first instance. Reference was made to Director of Public Prosecutions v Olivier 2006 (1) SACR 380 (SCA) at [15]—a case that was followed in Kock (supra) and referred to with approval in S v Nabolisa 2013 (2) SACR 221 (CC) at [81]. See also the discussion of s 316B in Commentary, sv General.

It was, furthermore, argued by counsel for the DPP that the matter was appealable as a question of law in terms of s 311(1) of the Criminal Procedure Act. This argument also failed on account of an earlier Supreme Court of Appeal decision in S v Mosterd 1991 (2) SACR 636 (T) where it was held that sentence can never be a question of law decided in favour of a convicted person. At [11] in Mphaphama Willis JA stated: ‘Certainly, when it comes to the exercise of a judicial discretion in favour of a convicted person in regard to sentence, that cannot be a question of law decided in his or her favour’. See also the discussion of Mosterd in the notes on s 311 in Commentary.

However, counsel for the DPP also sought to rely on s 16(1)(b) of the Superior Courts Act 10 of 2013, which stipulates that an ‘appeal against any decision of a Division on appeal to it, lies to the Supreme Court of Appeal’. This argument also led to a cul-de-sac. In Kock (supra) it was held that the general provision in s 16(1)(b) of the Superior Courts Act must be read in conjunction with the definition of ‘appeal’ in s 1 of the same Act. According to this definition the word ‘appeal’ as used in Chapter 5 of the Act ‘does not include an appeal in a matter regulated in terms of the Criminal Procedure Act ... or in terms of any other criminal procedural law ...’. This ‘definition of an appeal in the Superior Courts Act’, said Willis JA at [12], ‘precludes our coming to the assistance of the DPP ... [and] ... the Criminal Procedure Act does not allow the DPP a right of appeal from the High Court, where that court has sat as a court of appeal’. The appeal was accordingly struck from the roll.

It should be noted that in Mphaphama the Supreme Court of Appeal was more than somewhat frustrated by the fact that jurisdictional and procedural rules had prevented it from addressing the merits of the matter in the course of an appeal: ‘While the approach of the High Court in this matter is to be strongly deprecated, our hands are tied. This court’s jurisdiction is circumscribed by the Constitution and legislation’ (at [12]). The ‘approach’ of the High Court was indeed unacceptable: it had taken the view that even though a rape victim under the age of 12 was indeed in law incapable of giving consent, factual circumstances indicating consent or cooperation constituted ‘substantial and compelling circumstances’ justifying a departure from the prescribed minimum sentence legislation of life imprisonment in terms of the Criminal Law Amendment Act 105 of 1997.

 Table of Cases

Click here to see the full Table of Cases of the Criminal Justice Review No 1 of 2016


Click here for a printable .pdf version of this Review


Criminal Procedure Legislative Guide 2e
Criminal Procedure Act 51 of 1977 & Regulations 15e
Cases and Materials on Criminal Law
Newborn Baby, The


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