A bi-annual update complementing the

Commentary on the Criminal Procedure Act


NO 2 OF 2017


Andrew Paizes, Author (Editor)

Steph van der Merwe (Author)


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



One of the feature articles in this edition explores the requirement that there be ‘new facts’ before a bail applicant is entitled to launch a renewed bail application after the first application was unsuccessful. The article considers the meaning of ‘new facts’, stresses the right of the applicant to a reasonable opportunity to present those facts, and endorses the view expressed in a recent case that a court should not lightly deny an applicant for bail the opportunity to present new facts by adducing evidence. There are also suggestions on how problems in this area may be avoided and on how s 60(14) of the Criminal Procedure Act, which concerns access to the police docket in bail proceedings, should be applied.

The other feature article considers some of the many pitfalls that await the unwary in the application of the doctrine of common purpose. One danger, in particular, is singled out for attention: the need to keep apart the two distinct forms of common purpose, one based on agreement or mandate, the other on active association. It is argued that these two forms apply to quite different sets of circumstances, have different conditions for their application, and must not be invoked when those circumstances and conditions are not present. When they are applied in the wrong context, as they were, it is submitted, in the recent decision of the Constitutional Court in S v Makhubela & another, there is the danger that asking the wrong questions may lead to the wrong result.

Other issues dealt with in this edition include:

  • the meaning of ‘joint possession’ in relation to a prohibited object or substance (such as a firearm) and the unsuitability of the doctrine of common purpose in such cases;
  • the meaning of a ‘reasonable suspicion’ in s 40(1)(b) of the Criminal Procedure Act relating to an arrest without a warrant, and the need to keep distinct, when applying that section, the period before a first court appearance and the period after that event;
  • prosecutorial independence and integrity and the rationality of a decision by the prosecuting authority not to prosecute;
  • the need to respect the rights of women in their arrest and detention;
  • the weight to be given to the exculpatory aspects of a statement made by an ac-cused in terms of s 115 of the Act;
  • the distinction between the competence of a witness to testify and his or her ability to understand the nature and import of the oath;
  • the importance of using the judicial, and not the scientific measure of proof in assessing the evidence of an expert witness;
  • whether a s 204-witness may be cross-examined, in a trial of another person, on what the witness said in a bail affidavit in circumstances where no warning was given in terms of s 60(11B)(c) of the Act;
  • the role of ‘advanced age’ in sentencing, in particular in respect of a sentence of life imprisonment;
  • sentencing a convicted offender who maintains innocence and, as a result, never expresses remorse;
  • the effect of defective charges and jurisdictional rules on an appeal court’s power to intervene on sentence;
  • the validity of non-parole periods in sentencing; and
  • the inappropriateness of applying the forfeiture provisions in s 50 of POCA as an intervention in a commercial dispute.

Andrew Paizes





Why do we so often get common purpose wrong?

‘New facts’ for purposes of a renewed bail application: Principles, issues and procedures


The Criminal Procedure Amendment Act 4 of 2017

Judicial Matters Amendment Act 8 of 2017


(a) Criminal Law

Unlawful possession of automatic firearms—meaning of ‘joint possession’ and whether doctrine of common purpose may be applicable

(b) Criminal Procedure and Evidence

  (i) Pre-sentence

Judicial review of the decision not to prosecute: Prosecutorial integrity and rationality  

s 40: Arrest—when is a suspicion ‘reasonable’? Detention—distinction between period before first court appearance and period after this event

Arrest and the treatment of women: Need to respect women’s rights 

s 60(11B)(c): Cross-examination of a s 204-witness on the basis of his bail affidavit  

ss 115, 151, 203: Weight given to exculpatory parts of a s 115 statement; effect of accused’s failure to testify

ss 162 and 164: Admonishing a witness who does not understand the nature and import of the oath or affirmation, and the need to determine competence  

ss 151 and 212: Circumstantial evidence, the onus of proof and expert evidence—confusion between the scientific and judicial measures of proof

       (ii) Sentencing

Sentencing: The ‘advanced age’ of the convicted offender and life imprisonment as prescribed sentence

Sentencing an offender who maintains innocence: Rehabilitation, remorse and mercy  

Sentencing jurisdiction and minimum sentence legislation  

s 276B(1): Invalid non-parole periods and sentencing jurisdiction

       (iii) Forfeiture and confiscation

Forfeiture of property: s 50(1) of POCA—Act not intended to allow for intervention in a commercial dispute


Why do we so often get common purpose wrong?

It is surprising how often things go wrong when the doctrine of common purpose is ap-plied. It is surprising because the doctrine has, as its core, a very simple idea. That idea is that, when two or more people engage in a criminal enterprise together, responsibility in law for the act or series of acts that is or are performed by one of the group (the immediate party) may, in certain circumstances, be attributed to each of the other members (the remote parties) of that group.

When one looks at the elements of criminal liability, then, it is an idea that finds expression within the context of the conduct element of the actus reus. It tells us that the conduct of one of the parties, the immediate party, is, for the purposes of the criminal law, to be regarded, in addition, as the conduct of each of the other members of the group (the remote parties) when the necessary requirements are met. If, then, A, B, C and D form a common purpose to kill X, and if the requirements are met, the act of A in, say, stabbing X, will be treated by the criminal law as being the act of B, the act of C and the act of D in addition to being the act of A himself.

Before an articulation of what these conditions are, it is useful and instructive to consider briefly what this simple idea does not entail. It says nothing, to begin with, about causation. The requirement, in consequence crimes such as murder where there must be a causal link between the conduct element and the prohibited result, the death of the victim, remains in place and is not dislodged by the operation of the fiction that treats the act of A as the act of B or any other member of the group. It is still required in the case of, say, B that there be a causal nexus between the conduct element and the prohibited result, between conduct for which he is in law responsible and the death of X. It is just that a fictive element has been introduced which deems A’s act of stabbing X to be B’s own act. The causal requirement, in other words, has not gone away. It remains, but gives the illusion of having disappeared because B may be convicted even though there may be no causal link between conduct physically carried out by him and X’s death. He cannot, however, be convicted if there is no causal link between the attributed conduct and that result, so it is clear that the causal requirement remains.

The illusion of the disappearing causal element has led the courts into error in consider-ing the constitutional validity of the doctrine. In S v Thebus 2003 (6) SA 505 (CC), where the appellants attacked the doctrine ‘principally on the ground that it does not require a causal connection between their actions and the crimes of which they were convicted’, the Constitutional Court accepted (at [34]) that the ‘doctrine of common purpose dispenses with the causal requirement’, but upheld the doctrine since the requirement of a causal nexus was ‘not a definitional element of every crime’ so that, ‘under the common law, the mere exclusion of causation as a requirement of liability is not fatal to the criminal norm’ (at [37]–[38]).

I do not propose to consider, in this brief article, whether the doctrine should be viewed as passing constitutional muster. But I would point out that the real issue, in making such a determination, is whether and in what circumstances it is acceptable to attribute responsibility to an accused for conduct that is not his or her own. It is not whether dispensing with the causal nexus is justified, because that requirement is not dispensed with at all. It is, in effect, re-directed as an inquiry between attributed (as opposed to physical) conduct and the prohibited result. The focus of the Constitutional Court was thus, in my view, misdirected.

But causation is not the only aspect of criminal liability that the doctrine is wrongly understood to involve. Another is fault (or mens rea). Once the conduct of A is imputed to B, the question of course arises as to whether B had mens rea in respect of causing X’s death in relation to that act. In the context of murder, the question is whether B intended X to die as a result of that act. Legal intention or dolus eventualis will, of course, suffice, so that it will be enough if the prosecution could prove that B foresaw the real possibility that X might die as a result of the imputed act. This requirement is not affected by the operation of the doctrine apart from the fact that it pivots around a conduct element that is the result of a legal fiction which replaces actual conduct with imputed conduct. But the activation of that fiction may create problems that may easily lead to error if one is not vigilant. One such problem may be that the remote party has conduct attributed to him that he is not even aware of and does not, as a result, foresee. How, then, can he foresee the causing of death by that act?

Consider this example: A and B set out to rob X in circumstances where they accept that X may be shot and killed should he resist the robbery. B is heavily armed and it is he who effects the robbery while A waits out of sight in an idling car, ready to allow the pair to make a quick escape after the robbery. X offers fierce resistance and B considers it necessary to shoot him before X can kill B. The gun, however, jams, whereupon B hits him very hard on the head with the butt of the firearm. X dies as a result of this blow. Three things are clear in respect of A’s liability for the murder of X. First, the causal act of striking X on the head with the firearm will be imputed to A, to whom it was a matter of indifference whether B overcame X’s resistance by shooting him or striking him on the head. Second, this act would not ordinarily have been foreseen by A, so that it could not be said that he foresaw the causing of death by the very act that caused it. This would, in the ordinary course, be fatal to a conviction for murder. Third, however, it is clear that A would almost certainly have foreseen the possibility of the blow causing X’s death were it not for his remoteness from the scene. This fact, together with the fact that the causal act is, as has been pointed out, imputed to A so that it is regarded in law as being his own act, makes the case for liability unanswerable. The important lesson is that the fictive quality of the doctrine of common purpose forces us to introduce a corresponding fictive inquiry to prevent us from getting things wrong once more, one that considers what the remote party would have foreseen had he been aware of the act we are attributing to him. See, further, my article entitled ‘‘‘Mistake as to the causal sequence’’ and ‘‘mistake as to the causal act’’: Exploring the relation between mens rea and the causal element of the actus reus’ (1993) 110 SALJ 493.

A further source of confusion is that common purpose may take one of two forms. The first deals with a ‘mandate’ or agreement, in terms of which the parties form an actual agreement—whether express or implied—to embark on the criminal enterprise. The result is that an act performed by any one of them will be attributed to each of the others provided that the act falls within the borders of the mandate or agreement. The mandate will usually arise as a result of an express agreement but its ambit will include all acts incidental to the conduct expressly agreed upon which the parties must have accepted as being impliedly within its compass. Thus, if the agreement is to rob a heavily guarded store by using firearms known to each of the participants to contain live ammunition, it would ordinarily be impliedly accepted that the mandate includes the fatal shooting of any guard or other person who violently seeks to obstruct this purpose.

The second form of common purpose is described as arising out of ‘active association’. It applies ordinarily within the context of mob violence, where the parties do not necessarily know each other and have not been shown to have been operating as a result of any prior mandate or agreement, tacit or express. It arises, in this context, where it has been established that one of them has performed an act which has led to the prohibited result (usually the death of another) and it attributes this act to others who have, before the performance of that act, performed acts of their own which establish both that they were associating actively with the causal act and that they intended to make common cause with the person performing that act. The conditions that have to be proved by the prosecution when relying on this form of common purpose were set out in S v Mgedezi & others 1989 (1) SA 687 (A) at 705–6. A person who has been charged with murder arising out of an incident of mob violence leading to the death of another will be liable, said Botha JA, only if these prerequisites are satisfied:

‘In the first place, he must have been present at the scene where the violence was being committed. Secondly, he must have been aware of the assault on the [victims]. Thirdly, he must have intended to make common cause with those who were actually perpetrating the assault. Fourthly, he must have manifested his sharing of a common purpose with the perpetrators of the assault by himself performing some act of association with the conduct of the others. Fifthly, he must have had the requisite mens rea; so, in respect of the killing of the deceased, he must have intended them to be killed, or he must have foreseen the possibility of their being killed and performed his own act of association with recklessness as to whether or not death was to ensue.’

The two forms of common purpose, then, operate in quite different sets of circumstances and are governed by fundamentally different requirements. The ‘mandate’ form is predicated upon actual agreement between the parties to the common purpose. This presupposes some communication between them which sets the boundaries relating to the kind of act that is envisaged and how far the parties are prepared to go in giving effect to their agreed purpose. Once carried out by the immediate party, any act falling within the compass of what has been expressly or impliedly agreed upon will be attributed to each of the remote parties. The ‘active association’ form is fundamentally different. It is concerned with what the remote party actually does at the scene of the crime in associating himself with that conduct of the immediate party which is causally related to the death of the victim. He must, then, actually intend to make common cause with the actor or actors engaged in that conduct, and manifest his sharing of the common purpose by himself performing some act of association with that conduct.

Much has been said and written about this form of common purpose and the require-ments for its engagement, but it is worth drawing attention to two propositions for the purpose of this article. The first is that the conduct of the remote party by which he manifests his sharing of the common purpose to kill, must have been performed before the act of the immediate party that caused death was committed (see S v Motaung 1990 (4) SA 485 (A)). This is because our law rightly does not recognise responsibility for the conduct of others on the strength of the ratification of that conduct. Such responsibility can arise only in respect of conduct yet to be carried out, not conduct which has already been performed. The second is that our courts have made it quite clear that relatively innocuous or equivocal acts on the part of the remote party will not be regarded as constituting the kind of conduct that manifests the sharing of a common purpose with the immediate party. The ‘act of association’ would have to be ‘significant and not a limited participation removed from the actual execution of the crime’, and there ‘must be a close proximity in fact between the conduct considered to be active association and the result’ (see S v Dewnath [2014] ZASCA 57 (unreported, SCA case no 269/13, 17 April 2014) at [15]; S v Toya-Lee Van Wyk [2013] ZASCA 47 (unreported, SCA case no 575/11, 28 March 2013); and S v Gubuza (unreported, WCC case no A511/2013, 4 March 2014)). The warning issued by the Supreme Court of Appeal in Van Wyk (at [16]) that ‘care needs to be taken to avoid lightly inferring an association with a group activity from the mere presence of the person’ was heeded by the court in Gubuza to the extent that the holding of a screwdriver at the scene of the robbery by the accused was held to be insufficient, in the circumstances, to allow the court to infer that he had associated himself with the robbery committed by three others since it was, in the circumstances, reasonably possible that he had used the screwdriver to threaten his victims for the purpose of perpetrating the separate crime of rape, which had also taken place at the scene.

It is clear from this description of the two forms of common purpose that it is very important for a court applying the doctrine to identify correctly the form that best fits the facts. One notices, however, a worrying trend in which the courts, for some reason, choose to apply the active association requirements to circumstances for which that form of common purpose was clearly not designed. The result is an attempt to force those circumstances into a resistant and ill-fitting straitjacket, and a failure to address the questions that should have been asked had the court invoked the correct form of the doctrine.

This practice is of concern and there is an urgent need for our highest courts to put matters right. It is a pity, then, that this did not happen when the Constitutional Court had its most recent brush with common purpose in S v Makhubela & another 2017 (2) SACR 665 (CC).

The deceased in that case, a warrant officer, had been shot three times in his home by a group of men who had planned to steal his motor vehicle. He died later of his injuries. The applicants and their five co-accused were arrested and charged with, inter alia, murder and robbery with aggravating circumstances. They were convicted by the trial court. The full court dismissed their appeals, and the matter found its way to the Constitutional Court. Various extra-curial statements by some of the accused implicating the others were correctly held to have been wrongly received by the trial court, since it is now accepted that an admission, as well as a confession, is admissible only against its maker (see S v Litako & others 2014 (2) SACR 431 (SCA) and S v Mhlongo; S v Nkosi 2015 (2) SACR 323 (CC); see, too, Commentary in the notes to s 219 sv Is an admission (as opposed to a confession) by an accused admissible against a co-accused?). This left, in the main, the self-incriminating statements made by each applicant as well as their oral testimony. From this evidence it appeared that the two applicants had, in the company of their co-accused, been driven to the deceased’s home. The other members of the party were carrying weapons, and it was clear that they were intending to steal a vehicle. The co-accused alighted from the vehicle they had driven to the deceased’s home and went into his house. They later came back running and carrying an extra firearm. Makhubela stated that he remained in the car with, it seems, Matjeke, and added that he heard shots being fired. His co-accused ‘returned and started arguing about why the other had shot the deceased and the response was that the deceased had a firearm and would have retaliated’. Further, ‘they left the scene together’ (at [34]). The evidence showed, in addition, that the two applicants spent the afternoon prior to the shooting together, went to the same tavern with their co-accused after travelling with them in the same car, and drove to the scene of the crime with them in the full knowledge that they were carrying firearms.

There was no evidence that the applicants were at any stage coerced to travel and to remain with the group. Nor was there evidence that they had inquired from their co-accused what their intentions were when they parked the vehicle at a distance from the scene of the fatal shooting, or raised questions when they became aware that they were carrying firearms or when they left the vehicle with the weapons and headed to the deceased’s house. There was no evidence that they attempted to distance themselves from their co-accused or that they questioned their actions when they returned after the gunshots were fired. They did not flee or dissociate themselves in any way from their co-accused and did not even ask why the vehicle had to be driven at a very high speed from the scene or where the extra firearm had been procured. All this ‘suggest[ed];, said the court, that ‘they had an understanding with their co-accused to participate in criminal activity’ and made it ‘reasonable to infer that [they], far from being caught up unawares in illicit conduct, had an intention to commit a crime with their co-accused’ (at [40]; emphasis added).

No one would quarrel with any of these findings or pronouncements. All that was left for the court to do, one would have thought, was to inquire whether the ‘suggestion’ that the applicants had an ‘understanding with their co-accused to participate in criminal activity’ could be taken to the next step, in which these crucial questions would have to be answered in the affirmative before dismissing their appeals: had it been proved, beyond a reasonable doubt, that there was an agreement between the applicants and their co-accused that included the theft of the deceased’s vehicle by means of an act of force (for the purpose of holding them liable for robbery); and did the agreement, expressly or impliedly, include the possibility that it might be necessary to inflict on the deceased a serious and potentially fatal injury (for the purpose of holding them liable for murder)? Affirmative answers to these questions would undoubtedly have left intact the convictions for robbery and murder, since it is clear that mens rea in the form of dolus eventualis could, in that event, easily have been established in view of the court’s finding that the applicants must have ‘foreseen the possibility of the criminal result [of murder] ensuing’ (at [44]). Although the question should, strictly speaking, have been whether the applicants had foreseen the possibility that one of their co-accused might cause the death of the deceased as a result of any one of the range of acts to the commission of which they had expressly or impliedly agreed, it is clear that the answer would still have been ‘yes’ if it is accepted that the act that did cause death was one that fell within the ambit of their agreement.

So, then, did the causal act (shooting the deceased) fall within the four corners of the ‘mandate’ given to the perpetrator of that act? The court, unfortunately, did not address this crucial question at all. It failed to recognise that the case displayed none of the features calling for the application of the principles relating to ‘active association’. Instead of addressing the requirements for liability under the mandate leg of the doctrine, the court proceeded, without even acknowledging the existence of that form of common purpose, to examine the requirements for active association set out in Mgedezi. It found these to have been satisfied. A cursory consideration of these requirements would, however, reveal that they could have no meaningful purchase in the circumstances of the case. To say that the applicants were ‘present at the scene of the crime’ when they were sitting in a vehicle some distance away from the house where the fatal shooting took place, reveals a failure to appreciate the true juristic nature and purpose of ‘active association’. Whereas, in the case of a ‘mandate’, it is agreement that attracts responsibility for the conduct of another, active association concerns situations where no agreement has been shown to have been formed, and where the basis for the responsibility of the remote party derives from doing certain acts. If a mob is bent on stoning a person to death, the people who join in the attack will not necessarily have formed any agreement with those already so engaged. They must be shown to have done something significant, while at the scene and while the mob was involved in throwing stones at the victim, in order to attract responsibility for the act that does cause death. ‘Presence at the scene’ here really does mean that they must be at the specific place where the attack is being carried out and while it is being carried out. If they are not, they do not have a proper opportunity to appreciate the nature, development and direction of what the mob is doing. A person some distance away from the mob’s actions cannot have his ‘finger on the pulse’ of the ebbs and flows of the mob’s conduct, and any act of his cannot ordinarily, in any meaningful sense, be seen as an ‘act of association’.

An agreement between X and Y to do something binds one to the acts of the other in a very specific and descriptive way. Conduct, on the other hand, tends to be more equivocal and less precise than words, so that a strong link and a high degree of connection and immediacy is required to convince a court that the remote party did enough to warrant the conclusion that he has accepted and incurred responsibility for the act of the immediate party.

The fallacious reasoning of the Constitutional Court was compounded in its application of the remaining requirements of active association. Given that these requirements were applied in a context for which they were not designed, this is not surprising. To say, as the court did (at [43]), that the ‘applicants manifested their sharing of a common purpose with the perpetrators  . . . by performing an act of association with the conduct of the others in the form of travelling with them to and away from the scene of the crime’ is erroneous on two counts. First, the travelling to the scene of the crime is insufficiently proximate in time or place to meet the objections mentioned above: it certainly did not afford the remote party the necessary opportunity to keep his finger on the pulse of what was happening once the attack began in the deceased’s home. Their conduct in driving to the scene was, then, too early to be able to constitute an ‘act of association’. The act of driving from the scene, on the other hand, was too late to constitute such an act. The fatal shot had already been fired by then, and, as the court held in Motaung (supra), our law does not recognise common purpose by ratification.

Would the same result have been reached if the court had correctly viewed the facts through the lens of a mandate or agreement? We will never know, as the right questions were not asked and the evidence for and against such an agreement was not set out in the judgment or properly evaluated. There is certainly enough, on the facts as they have been set out by the court, to suggest that an agreement was reached between the applicants and their co-accused—whether expressly or by implication. This agreement would, almost certainly, have included the intimidation of the deceased in order to steal his vehicle and, if necessary, the use of force. Whether that force included the possible firing of shots at the deceased with a view to causing serious injury and, possibly, death, is less clear. A rigorous examination of all the facts would be necessary before such a finding could be made. It is true that the court found, in the context of mens rea, that dolus eventualis was present because the applicants ‘must have ‘‘foreseen the possibility of the criminal result [of murder] ensuing’’’, because they knew that their co-accused were carrying firearms, ‘which they must have known would be used if the plan went awry’ (at [44]). If this realisation was shared by the applicants and the perpetrators, it may be permissible to infer, as the only reasonable inference, that the possibility of a fatal shooting was at least an implied term of the agreement between the occupants of the car.

In short, it is probable that the same result would have been reached by the court even if the correct test had been applied. But this will not always be the case, and this fact cannot be used to justify the erroneous reasoning in S v Makhubela & another.

Andrew Paizes

New facts’ for purposes of a renewed bail application: Principles, issues and procedures


An accused has a fundamental and constitutionally protected right to apply for bail. However, it is an ‘abuse of  . . . proceedings’ to allow an unsuccessful bail applicant ‘to repeat the same application for bail based on the same facts week after week’. See S v Vermaas 1996 (1) SACR 528 (T) at 531e. A court may also not, in the absence of new facts justifying release on bail, set aside its own earlier refusal of bail. See S v Waldeck 2006 (2) SACR 120 (NC) at [53].

The requirement that new facts must be advanced or presented for purposes of a renewed bail application is also reflected in s 65(2) of the Criminal Procedure Act 51 of 1977. In terms of this section an appeal shall not lie in respect of new facts which arose or were discovered after the refusal of bail, unless such new facts were first placed before the bail magistrate concerned and such magistrate had given a decision against the bail applicant ‘on such new facts’. The effect of s 65(2) is that it merely precludes the right of appeal on new facts not before the bail court. See further the discussion of s 65 in Commentary, sv Section 65(2) and (3).

The meaning of ‘new facts’: General principles

There is no definition of ‘new facts’ in Act 51 of 1977. Case law, however, provides at least five guidelines or principles which are of assistance:

(a)    New facts are facts that came to light after refusal of bail, and obviously also include circumstances which have changed since the unsuccessful bail application was lodged. A detention period of almost three years between the first and the renewed bail application amounts to changed circumstances constituting a ‘new fact’ (S v Moussa 2015 (3) NR 800 (HC) at [7]); and the passage of considerable time coupled with the State’s failure to make progress with the investigation of the case, can also qualify as a new fact (S v Hitschmann 2007 (2) SACR 110 (ZH) at 113b).

(b)    New facts must be ‘sufficiently different in character’ from the facts presented at the earlier unsuccessful bail application (S v Mohammed 1999 (2) SACR 507 (C) at 512b) and ‘must not constitute simply a reshuffling of old evidence’ (S v Petersen 2008 (2) SACR 355 (C) at [57]).

(c)    The alleged new fact or facts must be ‘relevant for purposes of the new bail application’: see S v Petersen (supra) at [57]. This means that there must at least be some advance indication that the new facts, if received, would on their own—or in conjunction with all the facts placed before the court in the earlier unsuccessful bail application—assist the court in considering release on bail afresh: see S v Mohammed (supra) at 511h–512a.

(d)    In determining whether facts are new or not, a court is inevitably required to have due regard to the evidence presented or information received at the earlier unsuccessful application: see S v Vermaas (supra) at 531e–g. In S v Mpofana 1998 (1) SACR 40 (Tk) at 44g–45a Mbenenge AJ explained that ‘whilst the new application is not merely an extension of the initial one, the court which entertains the new application should come to a conclusion after considering whether, viewed in the light of the facts that were placed before court in the initial application, there are new facts warranting the granting of the bail application’.

(e)    In a situation where evidence was known and available to a bail applicant but not presented by him at the time of his earlier application, such evidence cannot for purposes of a renewed bail application be relied upon as ‘new facts’: see S v Le Roux en andere 1995 (2) SACR 613 (W) at 622a. In Le Roux at 622b it was explained that in the absence of such a rule, there could be an abuse of process (‘misbruik van hofprosedure’) leading to unnecessary and repeated bail applica-tions. An accused should not be permitted to seek bail on several successive occasions by relying on the piecemeal (‘broksgewyse’) presentation of evidence. It has been suggested that the rule is an absolute one and should be applied regardless of the bail applicant’s reasons for not adducing the evidence at the unsuccessful application: see generally S v Petersen (supra) at [58]. However, it is submitted that the rule should be applied with caution. It can hardly find application where the probable reason for the applicant’s failure to present the impugned evidence at the first bail application can be attributed to the applicant’s bona fide misinterpretation of the probative value of the evidence in relation to factual and legal issues concerning bail. The right to liberty pending the outcome of a trial or final appeal should not be frustrated by an inflexible rule. A bail court should be willing to examine and consider the reasons why relevant and available facts known to the bail applicant were not relied on in the initial application.

The right to a reasonable opportunity to present new facts

The recent decision in S v Nwabunwanne 2017 (2) SACR 124 (NCK) emphasises the importance of the above-mentioned right and also confirms that a court ‘should not lightly’ deny a bail applicant the opportunity to present new facts by adducing evidence (at [25]). The appellant and a co-accused were in custody on charges of dealing in cocaine and methamphetamine. Their bail application failed. The bail magistrate concluded that ‘neither of them was a suitable candidate for release on bail’ (at [9]). One of the main reasons for refusing bail was that there appeared to be ‘a very strong case’ against the appellant and his co-accused (at [9]). In this regard the bail magistrate had relied on the investigating officer’s testimony to the effect that video footage identifying the appellant as perpetrator, was part of the prosecution’s case (at [21]).

Some five months after the conclusion of the initial bail application, the appellant once again approached the magistrate with a further bail application. For purposes of this second application it was alleged that since the earlier refusal of bail, new facts had come to light. The bail magistrate ruled that in this second bail application the appellant would be allowed to present evidence only if counsel for the appellant could, by way of address and argument, first persuade the bail court that the alleged new facts did indeed constitute ‘new facts’ (at [20]).

In his address on the new facts, counsel for the appellant informed the bail magistrate that the investigating officer’s evidence in the first bail application ‘appeared to be false and  . . . had misled’ the court in that the video footage referred to by the investigating officer did not link the appellant to any of the charges against him. Indeed, it was submitted that the person identified in the relevant video was not the appellant. Photographs depicting the appellant entering a salon also did not link the appellant to any criminal activity supporting the charges against him. Counsel for the appellant also placed on record that the facts contradicting the evidence of the investigating office came to the attention of the appellant only after the unsuccessful bail application, that is, when the contents of the case docket were disclosed to the appellant’s attorney for purposes of trial preparation. In responding to the arguments of the appellant’s counsel, the prosecutor did not dispute counsel’s averments pertaining to the video footage and photographs (at [22]).

Having heard the arguments and submissions of both parties, the bail magistrate concluded that the appellant ‘had not convinced her of new facts’ (at [23]). The appellant was accordingly denied the opportunity to adduce evidence for purposes of introducing new facts at the second bail application. The bail magistrate’s finding, held Erasmus AJ at [27], was wrong. The appellant should have been granted an opportunity to adduce evidence in respect of the ‘alleged new facts’ (at [27]). Indeed, in her judgment the magistrate had not even addressed the issues pertaining to the video footage and photographs. She merely noted that matters concerning credibility could and would be tested during the course of the criminal trial (at [23]). This approach was a misdirection because it ignored the true nature and essential purpose of ‘new facts’ in the context of a renewed bail application. Erasmus AJ stated as follows (at [24]–[25], emphasis added):

‘New facts can and should be put before a magistrate by adducing oral evidence or submitting a document stating facts which are common cause. The purpose of adducing new facts is not to address problems encountered in the previous application, but should be facts discovered after the bail application. The facts relied on by the appellant in this instance were discovered after the initial application.  . . . An accused should not lightly be denied the opportunity to present such facts by means of adducing evidence.’

It was also evident that the bail magistrate had ignored the potential probative value of the ‘new facts’ which counsel for the appellant sought to introduce. The new facts were relevant and, if received, would have assisted the bail magistrate in reassessing her earlier refusal of bail—a refusal which was to a large extent based on the fact that the State had a strong case against the appellant. In this regard Erasmus AJ observed as follows (at [25]):

‘The submissions by the appellant’s counsel, at least prima facie, indicated that the evidence, presented on behalf of the respondent during the initial bail application, might be compromised and that the state’s case might not be as strong as the magistrate assumed it to be. The respondent did not dispute what had been conveyed on behalf of the appellant in respect of the photos, the video and the audio footage.’

In setting aside the magistrate’s decision in the second bail application, Erasmus AJ made the specific order that the appellant should, in the course of his further bail application, be afforded the opportunity to lead evidence of ‘the alleged new facts that had come to light and/or any new facts that had subsequently come to light’ (at [29].3). At [29].4 it was also specifically ordered that the State should—in response to any further evidence led by the appellant—be given an opportunity to lead evidence. The final order was that the appellant had to remain in custody pending the finalisation of his second bail application before the magistrate (at [29].5).

Avoiding problems concerning new facts: The proper application of s 60(14) of the Act

It is evident from the discussion of S v Nwabunwanne (supra) that the bail appellant was at his initial and unsuccessful bail application at a considerable disadvantage on account of the fact that at that stage he had no access to the contents of the police docket, more specifically the video footage referred to by the investigating officer. The position in which the bail appellant found himself should be understood in the context of s 60(14) of the Act. This section provides that no accused shall for purposes of bail proceedings have access ‘to any information, record or document relating to the offence in question, which is contained in, or forms part of, a police docket,  . . . unless the prosecutor otherwise directs’. See the discussion of s 60 in Commentary, sv Section 60(14). In S v Nwabunwanne (supra) the appellant’s attorney was only at some stage after the initial and unsuccessful bail application given access to the contents of the police docket. This access was for trial purposes. Indeed, s 60(14) contains a proviso to the effect that its provisions ‘shall not be construed as denying an accused access to any information, record or document to which he  . . . may be entitled for purposes of his  . . . trial’. This proviso is in line with the decision of the Constitutional Court in Shabalala & others v Attorney-General of Transvaal & another 1995 (2) SACR 761 (CC). In this case it was held that the State’s pre-constitutional ‘blanket docket privilege’ was inconsistent with the Constitution. See in this regard the discussion of s 201 in Commentary, sv The decision of the Constitutional Court in Shabalala under the main heading The ‘docket privilege’ in particular and the constitutional challenge.

However, must it be accepted that the appellant in S v Nwabunwanne (supra) was at his initial and unsuccessful bail application simply the unfortunate casualty of the provisions of s 60(14)? Surely not. It is argued below that in the first bail application in Nwabunwanne, the parties as well as the bail magistrate had failed to apply s 60(14) correctly and had overlooked relevant case law and other principles and statutory rules governing bail.

What happened at the first bail application in S v Nwabunwanne (supra) should be considered in the light of the decision of the Supreme Court of Appeal in S v Green & another 2006 (1) SACR 603 (SCA). In Green the attorney for the bail applicants had applied to the bail magistrate for access to the police docket (at [23]). The bail magistrate refused access, basing his decision on the provisions of s 60(14). Farlam JA (Heher JA and Cachalia AJA concurring) took a different view and relied on s 60(3) as a provision which, in the circumstances of the case, had to be given preference over the provisions of s 60(14). Section 60(3) provides that where a bail court ‘is of the opinion that it does not have reliable or sufficient information or evidence at its disposal or that it lacks certain important information’ to reach a decision on bail, the bail court ‘shall order that such information or evidence be placed before the court’. Having noted that a bail court is required to be more ‘proactive’ than a trial court, Farlam JA concluded as follows in Green (at [23], emphasis added):

‘On a proper consideration of the case on which the State relied, any reasonable court must have concluded that it lacked reliable and important information necessary to reach a decision, notwithstanding that such information was apparently readily available. In such circumstances the court has no discretion but to invoke s 60(3). In my view, the magistrate should, instead of refusing bail without more, have ordered the State to grant the defence access to the video tapes and any statements made by the police fingerprint experts, linking the fingerprints of either of the appellants with the crime, with the decision on whether or not to grant bail to be made thereafter.’

It is submitted that in Nwabunwanne (supra) the bail magistrate should, at the first bail application, have resorted to an order as provided for in s 60(3) and made in Green (supra). In Nwabunwanne there was a clear conflict of fact: the bail appellant denied involvement in the offences charged, whereas the investigating officer testified that there was video footage incriminating the appellant. The answer to this factual dispute was important for bail purposes in that it related to the strength of the prosecution’s case, a factor the State had relied upon in opposing bail. The absence of the video footage meant that the bail magistrate lacked ‘certain important information’ as envisaged in s 60(3). The factual dispute could for bail purposes readily be resolved by production of the footage. The passive attitude of the bail magistrate really complicated matters in that when the footage was eventually released for trial purposes, the appellant was forced to bring a second bail application on the basis of alleged ‘new facts’ discovered after the initial and unsuccessful bail application—the ‘new facts’ being the fact the person in the relevant video footage was not the appellant and the fact that the available photographs did not link the appellant to the offences as charged. Of course, the irony of the matter is that the relevant video footage existed at the time of the first bail application but was—in the absence of a s 60(3)-order by the bail court—inaccessible to the bail appellant on account of the provisions of s 60(14). It is submitted that in Nwabunwanne the appellant’s legal representative at the first bail hearing should, on the basis of the decision in Green, have asked the bail magistrate to consider making a s 60(3)-order.

It can also be argued that the prosecutor who opposed the first bail application in Nwabunwanne should, of his own accord, have granted the appellant’s legal representative access to the video footage referred to by the investigating officer. It was in the course of the latter’s testimony that it became clear that the identity of the perpetrator was in dispute and that this had a direct bearing on the bail issue, namely the strength of the State’s case. Why should a s 60(3)-order be necessary if s 60(14) itself gives the prosecutor a discretion to allow a bail applicant access to information in the police docket? In Nwabunwanne there were indeed compelling circumstances calling for disclosure in order to ensure a fair bail hearing. Access to the footage was crucial to the bail appellant’s case for release on bail.

Almost two decades ago the Constitutional Court—in confirming the constitutional validity of s 60(14)—was careful to point out that s 60(14) should not be read as if it permitted a ‘flat refusal’ by the prosecutor to disclose any information relevant to the charges against a bail applicant: see S v Dlamini; S v Dladla & others; S v Joubert; S v Schietekat 1999 (2) SACR 51 (CC) at [84]. The prosecutor has a discretion, but it is ‘not an unfettered discretion’: see S v Josephs 2001 (1) SACR 659 (C) at 664c–d. Protection of the right to a fair bail hearing may also require disclosure: see generally S v Mauk 1999 (2) SACR 479 (W) at 489c and 490b–c.

On the strength of the cases referred to in the previous paragraph, it would appear that there was no acceptable reason for the prosecutor’s non-disclosure of the video footage at the first bail hearing in Nwabunwanne. It might be that the prosecutor—like the bail appellant’s legal representative and the bail magistrate—had simply overlooked the decision in Green. It is also possible that the prosecutor in Nwabunwanne had at the time of the investigating officer’s testimony not yet seen the video footage and would therefore not have picked up the discrepancy between the investigating officer’s testimony and the fact that the person in the relevant video footage was not the appellant. But if this was indeed the case, the prosecutor should at some stage prior to the completion of the first bail proceedings have taken the trouble to view the video footage concerned. If the conflict between the investigating officer’s evidence and the video footage was detected at that stage, the prosecutor would no doubt have had to disclose the footage. A bail applicant’s constitutional right to a fair bail hearing may not be frustrated by a prosecutor’s suppression of evidence favouring release on bail. See Schwikkard & Van der Merwe Principles of Evidence 4 ed (2016) 192 for a discussion of s 60(14) and the ethical duty of the prosecutor.

Remarks in conclusion

At the beginning of this article it was pointed out that the presentation of ‘new facts’ is for good reason an important requirement for purposes of a renewed bail application. The case law makes it clear that a bail applicant should be given a reasonable opportunity to present such facts.

Attention was also drawn to the fact that a prosecutor should, at the first bail application, disclose facts that favour release on bail but are not known to the bail applicant on account of s 60(14). It is unacceptable to leave the matter to the accused on the basis that later, when he has access to the contents of the police docket for purposes of his trial, he can bring a second bail application based on ‘new facts’ discovered by him after his unsuccessful bail application. It is fundamentally unfair, and somewhat absurd, that facts favouring release on bail and known to the prosecutor at the initial bail application must somehow be held in abeyance until such time as they become known to the bail applicant, who can present these facts as ‘new facts’ for purposes of a renewed bail application.

Steph van der Merwe



The Criminal Procedure Amendment Act 4 of 2017

The above Act came into operation on 29 June 2017. See GG 40946 of 29 June 2017.

Act 4 of 2017 makes various amendments to ss 77, 78 and 79 in Chapter 13 of the Criminal Procedure Act. This chapter is headed ‘Accused: Capacity to Understand Pro-ceedings: Mental Illness and Criminal Responsibility’. For a discussion of the need for and aims of the amendments to ss 77, 78 and 79, see the previous edition of Criminal Justice Review, sv ‘(B) LEGISLATION’ under the sub-heading ‘Criminal Procedure Amendment Bill [B2-12017]’. The amendments to ss 77, 78 and 79 are also noted and discussed by James Grant in Revision Service 59, which updates Commentary and which will be published in January 2018.

Judicial Matters Amendment Act 8 of 2017

The above Act came into operation on 2 August 2017. See GG 41018 of 2 August 2017. This Act amended ss 18 and 184 of the Criminal Procedure Act. It also inserted a new s 194A. For the background to and the contents and impact of these amendments and insertion, see the discussion in 2016 (2) Criminal Justice Review, sv ‘(B) LEGISLATION’ under the sub-heading ‘Judicial Matters Amendment Bill [B14–2016]’. The amended ss 18 and 184 and the inserted s 194A are also noted and discussed in Revision Service 59 which updates Commentary and which will be published in January 2018.

 (a) Criminal Law

Unlawful possession of automatic firearms—meaning of ‘joint possession’ and whether doctrine of common purpose may be applicable

S v Ramoba 2017 (2) SACR 353 (SCA)

One of the offences of which the appellant in Ramoba had been convicted by the trial court was the unlawful possession of automatic and semi-automatic firearms. His appeal to the High Court was unsuccessful, and his further appeal to the Supreme Court of Appeal against his conviction (three counts) for this offence led the court to consider the ‘possession’ element of the offence and the possible engagement of the doctrine of common purpose in relation to that element.

Mbha JA (with whom the other judges agreed) considered that the principles of joint possession in relation to the offence of unlawful possession of firearms in instances of robbery committed, as in this case, by a group of people were trite and were aptly set out by Marais J in S v Nkosi 1998 (1) SACR 284 (W) at 287b–c. Marais J, in examining the mental aspect of possession in this context—the intention or animus to render physical possession of the guns by some, the possession of the group as a whole—held that two conditions had to be satisfied: first, the group must have had the intention to exercise possession of the guns through the actual detentor or detentors; and, second, the actual detentor or detentors must have had the intention to hold the guns on behalf of the group.

The High Court in Ramoba had relied on the decision in S v Khambule 2001 (1) SACR 501 (SCA) where it was held (at 503e–f) that ‘[t]here was no reason why in appropriate situations and if the principle of common purpose was applied, the common intention to possess the firearms jointly could not be inferred’. It was held further, on the facts of that case, that the only inference that could be drawn from the proven fact of common purpose was that there was joint possession of firearms used in the commission of the robbery. Mbha JA, however, considered the proposition set out in Khambule to be incorrect. That case, he said, was correctly criticised in S v Mbuli 2003 (1) SACR 97 (SCA) at 115a–g, where Nugent JA stated that, while he agreed that ‘there is no reason in principle why a common intention to possess firearms jointly could not be established by inference’, he could ‘not agree with the further suggestion that a mere intention on the part of the group to use the weapons for the benefit of them all [would] suffice for a conviction’ in respect of the unlawful joint possession of firearms. ‘Mere knowledge by the others that [one of their own] was in possession of a hand grenade, and even acquiescence by them in its use for fulfilling their common purpose to commit robbery’ was, said Nugent JA, ‘not sufficient to make them joint possessors’ of that grenade.

In Ramoba the investigating officer had found a pistol stuck ‘and presumably hidden away’ between the two front seats of a vehicle stolen by the robbers in the course of the robbery. There was no evidence showing who put the pistol in the vehicle and no evidence showing whether or not the appellant was aware of its presence in that vehicle. There were accordingly ‘no facts from which it [could] be inferred that the appellant had the intention to possess the  . . . pistol through the actual detentor thereof, who [was] in any case unknown, and whether or not the person who put it inside the [vehicle] intended holding it on behalf of the group, including the appellant’ (at [15]). The conviction on this particular count—one of three involving unlawful possession of a firearm—could not be upheld. In respect of the other two counts, there was ample evidence to establish that the firearms in question which were used in the robbery were ‘clearly possessed by the robbers for themselves and for each other’ (at [19]), and the appeal against his conviction on these counts was unsuccessful.

The decisions in Mbuli and Ramoba are important for the light they shed on the distinct requirements for ‘joint possession’ and ‘common purpose’. As Nugent JA put it in Mbuli at 114–15, ‘[c]ommon purpose and joint possession both require that the parties concerned share a common state of mind but the nature of that state of mind will differ in each case’. He was entirely correct in rejecting the suggestion in Khambule ‘that a mere intention on the part of the group to use the weapons for the benefit of all of them will suffice for a conviction’.

The truth is that the doctrine of common purpose lends itself to use in crimes that are known in the criminal law as ‘consequence crimes’, where a causal nexus is required between the conduct of the accused and a specific, prohibited result. Since the doctrine has the effect of imputing to one party (the remote party) the conduct of another (the immediate party), the doctrine has great efficacy in helping the prosecution to prove its case when it is unable to establish a causal nexus between the accused’s own conduct and the result in question. It does not lend itself with any great efficacy to crimes known as ‘circumstance crimes’, where it is the existence of a certain state of affairs involving the accused that constitutes the conduct element of the actus reus. Unlawful possession of a firearm is such a crime. Whereas it makes sense to think of a group of wrongdoers agreeing to rob a bank or to kill another person, it makes much less sense to imagine them agreeing to being in possession of a firearm.

Even if it did make sense to impute the conduct of ‘being in possession’ to others in the group who were not the detentors of the firearm, how could we make sense of the fact that ‘possession’, as the courts have explained, has the mental element involving the intention to possess? The doctrine of common purpose makes no provision for the attribution of states of mind, only conduct.

There is, of course, nothing to prevent a court, in applying the appropriate rules governing circumstantial evidence and inferential reasoning, from drawing various inferences, in a proper case, from the fact that the parties formed a common purpose, say, to rob. These inferences may or may not be helpful in its determination of whether one or both of the conditions for ‘joint possession’ set out in Nkosi have been satisfied. But it is crucial to understand that this will not necessarily be so, since the inquiries into ‘common purpose’ and ‘joint possession’ are distinct and operate in different planes.

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

Judicial review of the decision not to prosecute: Prosecutorial integrity and rationality

Zuma v Democratic Alliance & others; Acting National Director of Public Prosecutions & another v Democratic Alliance & another [2017] ZASCA 146 (unreported, SCA case no 771/2016, 1170/2016, 13 October 2017)

The appeals in the above consolidated matter (hereafter the ‘SCA case’) were dismissed by Navsa ADP, writing for a unanimous full bench. The SCA case dealt with appeals against the decision of the High Court in Democratic Alliance v Acting National Director of Public Prosecutions & others 2016 (2) SACR 1 (GP). This case (hereafter the ‘High Court case’) is discussed in Chapter 1 in Commentary, sv Review of prosecuting authority’s decision to withdraw charges, and the validity of a mandatory interdict to prosecute.

In the SCA case Navsa ADP made several observations regarding prosecutorial integrity and rationality. The present note focuses on these observations. However, it is also necessary to make very brief reference to the High Court case where prosecutorial integrity and rationality were first addressed. In the High Court case the Acting National Director of Public Prosecutions (the ‘ANDPP’) had argued that his decision on 6 April 2009 not to prosecute the current President was rational in that non-prosecution was necessary to protect the integrity of the National Prosecuting Authority (the ‘NPA’). The ANDPP alleged that the then Head of the former Directorate of Special Operations (the ‘DSO’) had abused the prosecutorial process for political reasons by trying to manipulate the timing of the service of the indictment. However, the ANDPP had in a media address confirmed that the alleged misconduct of the Head of the DSO had not affected the merits of the charges against the President. In a joint judgment the High Court (Ledwaba DJP and Pretorius and Mothle JJ) concluded that in these circumstances ‘[t]here was thus no rational connection between the need to protect the integrity of the NPA and the decision to discontinue the prosecution  . . .’ (at [88]).

In the SCA case it was contended on behalf of the NPA that the High Court should have found that the overall conduct of the Head of the DSO was such that it clearly evidenced ‘the manipulation of the prosecutorial process for political ends’, which had compelled the ANDPP ‘to discontinue the prosecution’ (at [52]). The stated aim of the ANDPP in discontinuing the prosecution ‘was to preserve the integrity of the NPA and to promote its independence’ (at [83], emphasis added). It is submitted that the ANDPP’s idea that non-prosecution was necessary to ‘promote’ the independence of the NPA is rather quaint and certainly more than whimsical. The decision to discontinue prosecution can as a matter of principle not be taken in order to promote prosecutorial independence. It can, however, be taken because of prosecutorial independence. Prosecutorial independence is a constitutional guarantee. See the cases referred to and discussed in Chapter 1 of Commentary, sv Professional independence and sv The discretion to prosecute. The fact of the matter is that reliance on prosecutorial independence and integrity means nothing more and nothing less than a duty to prosecute where there are reasonable prospects of success and if there is no compelling reason to decline to prosecute.

Was protection of the integrity of the NPA a compelling reason? Navsa ADP took the view that it is ‘inimical to the preservation of the integrity of the NPA’ if a decision to discontinue prosecution is taken ‘because of a non-discernible negative effect of the timing of the service of an indictment on the integrity of the investigation of the case and on the prosecution itself’ (at [84]). Issues concerning abuse of process in relation to a specific prosecution should in principle be decided by a trial court and not by the prosecutorial authority by way of an ‘extra-judicial pronouncement’, as the ANDPP had done (at [94](viii)). 

However, the NPA argued that where the prosecution itself believes that there has been an abuse of process, discontinuing becomes a prosecutorial decision because it cannot be required of the prosecution to prepare a case on the basis that a court should at a later stage decide whether a stay of prosecution is justified. Navsa ADP rejected this argument (at [91]):

‘I disagree. It is incumbent on prosecutors to disclose to a court any fact which, in their view, may impact negatively on the prosecution and in favour of the accused. This is in line with constitutional values and the provisions of the NPA Act. It is in the interest of the NPA, accused persons and the public’s confidence in the administration of justice, that decisions concerning allegations of abuse of process be made by a trial court.’

The Supreme Court of Appeal concluded, like the High Court, that there was no rational connection or link between the decision to terminate the prosecution and preservation of the integrity of the NPA (at [84]). Navsa ADP also found as follows (at [94](x)–(xi), emphasis added):

‘[The ANDPP’s] stated purpose of preserving the integrity of the NPA and advancing the cause of justice, can hardly be said to have been achieved. The opposite is true. Discontinuing a prosecution in respect of which the merits are admittedly good and in respect of which there is heightened public interest because of the breadth and nature of the charges and the person at the centre of it holds the highest public office, can hardly redound to the NPA’s credit or advance the course of justice or promote the integrity of the NPA. Regrettably, the picture that emerges is one of [the ANDPP and a DNDPP] straining to find justification for the termination of the prosecution.  . . . Thus the conclusion of the court below, that the decision to terminate the prosecution was irrational, cannot be faulted.

Of further importance is the fact that the ANDPP had, in the course of the process to terminate the prosecution, taken certain steps which reflected adversely on the integrity of the NPA. This is rather ironic given the fact that the NPA had argued that non-prosecution was necessary to protect the integrity of the NPA. Consideration of the steps taken was relevant because a rationality review includes an assessment of the process that was followed in reaching the impugned decision.

Navsa ADP identified several processes and decisions which had preceded the ultimate decision to discontinue prosecution and which contributed to the finding that rationality was absent. Only three examples need be mentioned.

The first example is the ANDPP’s exclusion of the prosecution team from the process that led to the decision not to prosecute, especially the exclusion from the final deliberations that took place on 1 April 2009. This, said Navsa ADP at [89], ‘was in itself irrational’. The exclusion of ‘the senior litigators steeped in the case’ (at [89]) also appeared ‘to have been deliberate’ (at [94](vii)).

The second example is the fact that in his supplementary affidavit the ANDPP explained that he had been untruthful when he told a senior member of the prosecution team that he, and he alone, had decided to delay the service of the indictment, whereas it was in fact the Head of the DSO who had made the decision (at [85] and [94](vi)). At [85] Navsa ADP observed crisply and accurately: ‘If anything affects the integrity of the NPA, it is an ANDPP lying to a senior prosecutor. The admitted deception compellingly affects the credibility of [the ANDPP’s] motivation for discontinuing the prosecution.’

The third example relates to the passive and submissive attitude of the NPA regarding the interception and recordings of the telephone conversations referred to by the ANDPP and which were in possession of the legal team of the President. The NPA, said Navsa ADP at [63], ‘allowed itself to be cowed into submission by the threat of the use of the recordings, the legality of the possession of which is doubtful’. There was, it would appear, no real effort by the NPA to assert its independence and constitutional duty.

In dismissing the appeals with costs, Navsa ADP remarked that it was difficult to comprehend why the present regime at the NPA thought that the decision to discontinue the prosecution could be defended (at [94](xiii)).

In Democratic Alliance v President of the Republic of South Africa & others 2013 (1) SA 248 (CC) at [13] the Constitutional Court said that ‘an effective criminal justice system  . . . is vital to our democracy’. In this regard the NPA has a pivotal role to play, especially when it comes to the decision whether to prosecute or not. In National Society for the Prevention of Cruelty to Animals v Minister of Justice and Constitutional Development & another 2016 (1) SACR 308 (SCA) at [24] Saldulker JA stated that ‘all decisions by the prosecuting authority to prosecute or not to prosecute must be taken impartially without fear, favour or prejudice [and] must adhere to prosecuting policy and directives’. Indeed, prosecutorial integrity also demands that this be so.

s 40: Arrest—when is a suspicion ‘reasonable’? Detention—distinction between period before first court appearance and period after this event

Minister of Safety and Security v Magagula [2017] ZASCA 103 (unreported, SCA case no 991/2016, 6 September 2017)

In terms of s 40(1)(b) of the Criminal Procedure Act, a peace office may without warrant arrest any person ‘whom he reasonably suspects of having committed an offence referred to in Schedule 1, other than the offence of escaping from lawful custody’. One of the issues raised in Magagula was whether the arresting officer did in fact entertain a reasonable suspicion as required in this provision.

Lamont AJA (with whom Lewis, Petse and Swain JJA and Fourie AJA agreed) held that the suspicion regarding the respondent was, in the circumstances of the case, a reasonable one. The court referred to the meaning of ‘suspicion’ set out in Shabaan Bin Hussein & others v Chong Fook Kam & another [1969] 3 All ER 1627 as, in its ordinary sense, ‘a state of conjecture or surmise where proof is lacking; I suspect but I cannot prove’. Suspicion, Lamont AJA added, ‘arises at or near the starting point of an investigation of which the obtaining of prima facie proof is the end’: Powell NO & others v Van der Merwe NO & others 2005 (5) SA 62 (SCA) at [36]; Woji v Minister of Police 2015 (1) SACR 409 (SCA).

A suspicion will be reasonably held ‘if, on an objective approach, the arresting officer has reasonable grounds for his suspicion’ (at [10]): see Duncan v Minister of Law and Order 1986 (2) SA 805 (A) at 814. Once the required suspicion exists, the arresting officer is ‘vested with a discretion to arrest, which he must exercise rationally’ (emphasis added; see Minister of Safety and Security v Sekhoto & another 2011 (5) SA 367 (SCA)).

In the present case, the following factors pointed to the conclusion that the suspicion formed by the arresting office was a reasonable one: he obtained cogent evidence which was, on the fact of it, acceptable; this evidence was corroborated; he personally obtained information from the investigating officer as well as from another suspect, a fellow arrestee of the respondent; the totality of the evidence indicated that the respondent had committed the offence (a fatal shooting); and the suspect was able to and did point out the person who had committed the crime as being the respondent. An argument that the suspect had given conflicting evidence to another police office was rejected as irrelevant because it was—even if one assumed the truth of the allegation—unknown to the arresting officer.

The respondent argued that he had been unlawfully detained as well as arrested. Lamont AJA rejected this claim as well. Once an arrest has been effected, the authority to detain that is inherent in the power to arrest, is exhausted. The authority to detain the suspect further is then within the discretion of the court: Sekhoto (supra) at [42]; Minister of Safety and Security v Tyokwana 2015 (1) SACR 597 (SCA). The two periods of detention—one until first appearance in court; the other from first appearance until ultimate release—must, said Lamont AJA, be considered separately. In this case the respondent’s case for unlawful detention in respect of the first period was dependent upon the appellant failing to establish that his arrest was lawful. Since the arrest was lawful, his detention for the period ending on the day of his first appearance in court was not unlawful.

The respondent’s detention in respect of the second period was dependent on the lawfulness or otherwise of the magistrate’s orders. There was no evidence that the magistrate had behaved in an unlawful manner and, in any event, the magistrate was not a servant of the appellant, so no liability could ensue in the present action. The evidence established, further, that the detention of the respondent at this stage was not at the instance of the appellant’s servant.

The appellant was accordingly held to have justified both the arrest and the detention of the respondent up until the day of his first appearance in court.

Arrest and the treatment of women: Need to respect women’s rights

Mathe v Minister of Police 2017 (2) SACR 211 (GJ)

This case was an action for damages for an unlawful arrest and detention. The defendant admitted the unlawfulness of the arrest, and the court had only to determine the quantum of damages to be awarded to the plaintiff.

The plaintiff, along with two other women, were waiting for transport, seated on chairs at a filling station at 2.00 am in the morning, when an unmarked police vehicle arrived. On these scanty facts, the police surmised that the women were prostitutes and arrested them after a very short exchange. There was nothing to suggest that the police had any lawful reason for arresting them. The police clearly ‘abused the power entrusted to them’ (at [32]). They did not even take the basic step of identifying themselves to the women prior to questioning them. They simply bundled them into the vehicle and did not inform them of the offence for which they were being arrested. The plaintiff’s constitutional rights were explained to her only on the following morning.

The plaintiff was put in a cell with four other people. The cell was filthy, had one non-functioning toilet and a tap with no water, had dirty blankets on the floor, and had an unbearable smell. She was not allowed to make a telephone call.

Opperman J subjected the conduct of the police to scathing criticism. They were in breach of their obligations set out in s 205 of the Constitution to prevent and combat crime, and subjected the women in question to treatment that was in breach of the right to equality. Men in their situation, the judge added, would not have been treated in this fashion. The police, in doing the opposite of t s 205 required them to do, ‘added unnecessarily the infinite quotient of women’s humiliation and distress in the history of our society’ (at [35]). This could not be ‘treated lightly by a court enjoined to apply the Constitution’.

The plaintiff, said Opperman J, was ‘subjected to prejudices which [were] exclusively based on gender’ (at [36]). The ‘grinding down of women’s rights erode[d] the rights of the community as a whole’. In respect of the youths who spoke cruelly to the plaintiff in the wake of the police actions, Opperman J pictured them being ‘encouraged by the fact that it was the police who instigated her fall from grace’. They should, instead, ‘be seeing our police being considerate and respectful of the women in our communities, in the finest traditions of all South African cultures’. The judge referred to R & others v Minister of Police (unreported, GP case no A315/2015, 21 April 2016), where it was said that cases of this kind had a ‘public interest element’, since their ‘impact is not limited to the individuals but extends to the community of which they form part’.

In view of the inhumane conditions to which she was subjected, the loss of her employment as a result of her arrest and detention, and the stigma following it, damages in the amount of R120 000 were awarded to the plaintiff.

s 60(11B)(c) and s 204: Cross-examination of a s 204-witness on the basis of his bail affidavit

S v Miya & others 2017 (2) SACR 461 (GJ)

Miya (supra) is a rather odd case. At the trial of four accused, counsel for accused 1 (hereafter ‘defence counsel’) wanted to cross-examine a s 204-witness (hereafter ‘G’) on the basis of an affidavit made by G in an earlier bail application in respect of another case (hereafter the ‘Sandton case’). In that case G was standing trial together with accused 2 in the present trial. It was common cause that at G’s earlier bail application, the bail court had not warned G, as required by s 60(11B)(c) of the Criminal Procedure Act, that anything he said at the bail hearing ‘may be used against him  . . . at his  . . . trial and such evidence becomes admissible in any subsequent proceedings’ (at [24]). In the absence of this warning, a bail applicant’s evidence—in Miya it happened to be an affidavit—would in principle be inadmissible at the bail applicant’s trial. For a detailed discussion of this matter, see the analysis of s 60(11B)(c) in Commentary, sv Section 60(11B)(c) and the admissibility of bail evidence under the main heading Section 60(11B)(c). See also Miya at [22] and [24]. The issue in Miya, said Msimeki J at [35], was whether G could be ‘cross-examined on the statement that he made during the bail hearing if he was not properly warned’.

In Miya the prosecutor objected to defence counsel’s proposed use of G’s bail affidavit in cross-examining him. It was argued that the defence had to satisfy the court that G’s bail affidavit was admissible in the sense that it could be used for purposes of cross-examination (at [3]). At [14] the prosecutor also submitted that s 60(11B)(c) should be given a ‘wide interpretation’: the bail affidavit would be inadmissible not only at G’s trial in respect of the pending Sandton case, but also in respect of ‘any subsequent proceedings’, like the situation where G subsequently happens to be a State witness, as in the present matter. According to the prosecutor, the issue was not whether G ‘is a witness or an accused, as the protection is derived from s 60(11B)(c)’ (at [30]). Section 60(11B)(c), so the argument ran, distinguishes between the bail applicant’s later trial and ‘any subsequent proceedings’. Msimeki accepted this distinction (at [29], [34] and [38]). At [40] it was held:

‘The legislature, if its intention was to restrict the applicability or admissibility of evidence to the trial to which the bail record relates, in my view, would not have added the words ‘‘in any subsequent proceedings’’. [G], in my view, enjoys the protection, even though he is a witness.’

The protection referred to in the above passage presumably relates to disclosure of incriminating information or evidence obtained in breach of G’s privilege against self-incrimination. But did he need this protection? As a s 204-witness, G was required to answer ‘any question put to him, whether by the prosecution, the accused or the court, notwithstanding that the reply thereto may incriminate him’ with regard to offences specified by the prosecution: see s 204(1)(b). For an analysis of the details and application of s 204(1), see the discussion of s 204 in Commentary, sv Practical application of s 204(1).

Defence counsel submitted that s 60(11B)(c) was not applicable to G in his capacity as a State witness. Section 60(11B)(c) was meant to ‘cover and protect’ G in his capacity as an accused at his trial in the Sandton case and in respect of which he had applied for bail without the required warning by the bail court (at [14]). Msimeki J rejected this argument on the basis that s 60(11B)(c) ‘does not distinguish or discriminate between a witness and an accused covered by it’ (at [30]). It was also held that G was ‘by reason of the protection he enjoys in the Sandton case  . . . also equally protected in this case, even though he is a witness. He was not warned in the bail proceedings in the Sandton case which is still to be heard and concluded’ (at [31]).

It is respectfully submitted that the words ‘at any subsequent proceedings’ in s 60(11B)(c) were not meant to cover a situation where a bail applicant becomes a s 204-witness in another case where the constitutional fair-trial rights of other accused are at risk. To put the matter differently: use of G’s bail affidavit for purposes of cross-examination in Miya was a matter that Msimeki J should have addressed in the context of the fair-trial rights of the four accused who were on trial. The presiding judge at G’s trial in respect of the pending Sandton case would in turn be required to take into account G’s fair-trial right in order to decide any admissibility issues concerning G’s bail affidavit. In S v Dlamini; S v Dladla & others; S v Joubert; S v Schietekat 1999 (2) SACR 51 (CC) Kriegler J said (at [99], emphasis added):

‘Provided trial courts remain alert to their duty to exclude evidence that would impair the fairness of the proceedings before them, there can be no risk that evidence unfairly elicited at bail hearings could be used to undermine accused persons’ rights to be tried fairly. It follows that there is no inevitable conflict between s 60(11B)(c)  . . . and any provision of the Constitution. Subsection (11B)(c) must, of course, be used subject to the accused’s right to a fair trial and the corresponding obligation on the judicial officer presiding at the trial to exclude evidence, the admission of which would render the trial unfair.’

It was pointed out in Miya at [36] that counsel for accused 2 also argued that G could be cross-examined with reference to the contents of his bail affidavit. He relied on S v Aimes & another 1998 (1) SACR 343 (C) which was decided before s 60(11B)(c) came into operation. In Aimes there were two accused. Desai J held that admission of accused 1’s bail evidence—obtained in breach of his right to silence—would violate his fair-trial right. However, it was also true that the exclusion of accused 1’s bail evidence would in the specific circumstances of the case have infringed the fair-trial right of accused 2. Desai J accordingly ruled that the bail evidence of accused 1 was admissible for a limited purpose: it could be used by accused 2 for purposes of cross-examining accused 1, provided that counsel for accused 2 did not seek to introduce the transcript of the bail evidence as ‘being a statement of the truth of its contents to be used against accused no 1’ (at 351c).

In Miya Msimeki J at [36] did not follow the principle relied upon in Aimes: ‘The position’, said the judge, ‘is now settled because the section is clear.’ It is respectfully submitted that s 60(11B)(c) is not that clear, and that a court of appeal would in all probability, on account of fair-trial rights, not follow the decision in Miya.

ss 115, 151, 203: Weight given to exculpatory parts of a s 115 statement; effect of accused’s failure to testify

Director of Public Prosecutions, Gauteng Division, Pretoria v Heunis 2017 (2) SACR 603 (SCA)

What weight should a court accord to the exculpatory aspects of a statement made by an accused in terms of s 115 of the Criminal Procedure Act which are not repeated in evidence? This question, said Bosielo JA, who delivered the judgment of the court in Heunis, ‘has long engaged our courts and spawned many judgments’ (at [14]). The ‘correct approach’ he considered, was adopted in R v Valachia & another 1945 AD 826 as follows:

‘Naturally, the fact that the statement is not made under oath, and is not subject to cross-examination, detracts very much from the weight to be given to those portions of the statement favourable to its author as compared with the weight which would be given to them if he had made them under oath, but he is entitled to have then taken into consideration, to be accepted or rejected according to the Court’s view of their cogency.’

This ‘salutary approach’, said Bosielo JA, has been followed consistently and was re-stated in S v Cloete 1994 (1) SACR 420 (A) at 428b–c and e–g, where the court added that it could ‘think of no other reason why a court should be entitled to have regard to the incriminating parts of such a statement while ignoring the exculpatory ones’. The court in Cloete went further, recognising that an accused might try to abuse the procedure allowed by s 115, but warned that a court ‘should ensure that such an attempt does not succeed by refusing to attach any value to statements which are purely self-serving, and, generally, by determining what weight to accord to the statement as a whole and to its separate parts’.

Further clarification was provided in S v December 1995 (1) SACR 438 (A) at 444b–e, where the court, after warning that exculpatory statements contained in a confession which were ‘not supported by credible evidence’ could not be taken for the truth, added that they might nevertheless ‘serve to alert a court to a possibility of events or circumstances not otherwise revealed by the evidence’. If that possibility is a reasonable one, said the court, the accused, even if he repudiates the statement, is ‘entitled to have his conduct and state of mind assessed in the light thereof’.

In Heunis the respondent had been charged with murder but found guilty by the trial court of culpable homicide. He did not testify, but the trial court placed weight on his s 115 statement, in which he claimed that he had shot the deceased accidentally when they were both seated in the front of a vehicle and she had tried to convince him not to kill himself, causing the firearm to discharge when she attempted to displace his hand from the weapon.

Bosielo JA, after considering the evidence as a whole, including the s 115 statement, was of the view that the trial court did not apply the law as set out in the above cases. The clear and unchallenged evidence of a ballistic expert was to the effect that the trajectory of the bullet was such that it was not possible for the deceased to have pushed down the firearm held by the respondent in the position described by him in his statement. In view of the actual trajectory and the respondent’s failure to testify, the only reasonable inference was that the respondent intended to shoot the deceased. Absent any explanation by the respondent, this had to amount to a direct intention to kill.

The ‘damning evidence’ called for an answer from the respondent. No answer was forthcoming, and although he was exercising his constitutional rights under s 35(3)(h) ‘to remain silent and not to testify during the proceedings’, his failure to testify had to be ‘taken into account against him’ (at [19]) (see S v Boesak 2000 (1) SACR 633 (SCA) and see the discussion of the cases following Boesak in Commentary in the notes to s 203 sv Constitutional implications). The result of the respondent’s failure to testify was that the State’s strong evidence became ‘conclusive proof of his guilt beyond reasonable doubt’ (at [20]). The conviction for culpable homicide was accordingly replaced by one for murder.

s 162 and s 164: Admonishing a witness who does not understand the nature and import of the oath or affirmation, and the need to determine competence

S v Haarhoff & another [2017] 4 All SA 446 (ECG)

It has been clearly articulated by the Supreme Court of Appeal that, if it is found that a witness does not understand the nature and import of the oath, it is necessary for the court, before it admonishes the witness to tell the truth in terms of s 164(1), to establish whether the witness is able to distinguish between truth and lies: see, for instance, S v Matshivha 2014 (1) SACR 29 (SCA). This makes sense, because a witness who cannot make such a distinction is not a competent witness, and competence is a requirement that applies to all witnesses in all cases (see s 192 and the notes on that section in Commentary).

This distinction is a crucial one, but it is clear, on a reading of the cases, that the courts, on occasion, either conflate the two inquiries (into competence, on the one hand, and into whether the witness can take the oath or affirmation, on the other) or, worse, fail properly to give effect to the former when addressing the latter (see the discussion in Commentary in the notes to s 164). Sometimes, as the court in Matshivha observed, it is not clear from the questions put to the witness in an inquiry by the court whether the purpose of the questions is to establish his or her competence as a witness or the ability to understand the nature and purport of the oath. Such imprecision and sloppiness may lead to the conclusion, as it did in Matshivha, that the court has not properly complied with its duties under ss 162 and 164.

The treatment of the witness by the majority of the court in S v Haarhoff & another [2017] 4 All SA 446 (ECG) may be such an instance. The complainant in that case, in which the appellant had been convicted of rape, was 23 years of age, but with a mental age of just 10 and an IQ of only 70, which ‘placed her on the border line between mild mental retardation and border line intellectual functioning’. A clinical psychologist who examined her described her as ‘having a below-average intellectual functioning, but not mentally retarded’. The psychologist testified that in her expert opinion the complainant was able to testify in court and had ‘a basic understanding of what it meant to tell the truth and what it meant to tell a lie’ (at [13]). Further, in her view, she ‘had the cognitive capacity suitable to being admonished by the Court’. She understood what it meant to have sexual intercourse and the possible consequences of such intercourse, and was, in the psychologist’s view, able to express her consent or otherwise to it. The trial court admonished the complainant to tell the truth after its own investigation and after concluding that she did not understand the nature and import of the oath.

This, said the majority (Brody AJ, with Chetty J concurring), was sufficient to render the complainant’s evidence admissible, as this was, said Brody AJ, consistent with the principle established in S v Williams 2010 (1) SACR 493 (ECG). In Williams, however, the transcript demonstrated unequivocally that the court a quo was satisfied that the complainant comprehended the difference between truth and falsehood, and it was for this reason that its admonishment that she speak the truth was sufficient to render the evi-dence admissible. The transcript in Haarhoff, however, suggested strongly that the complainant had no such comprehension. The inquiry conducted by the court a quo in this case was very cursory and raised many concerns. In response to the question ‘Do you know what happens to someone who does not tell the truth?’, the complainant answered ‘No’. Further, to the question, ‘Is it good to tell lies?’, she answered ‘Yes’.

These concerns prompted Mjali J, who gave a separate, dissenting judgment, to conclude that there had not been proper compliance with s 164(1). Before admonishing the complainant to tell the truth, said Mjali J, the court had a duty to determine whether the witness was competent, a duty that could not be ‘abdicated’ but had to be carried out by the court itself. The purpose of the inquiry prior to admonishing the witness was not only to determine if the witness could understand the abstract concepts of truth and falsehood, or could give a coherent and accurate account of the events in question, but also to determine if he or she could distinguish between truth and falsity. This entailed a recognition of ‘the danger and wickedness of lying’ (see Henderson v S [1997] 1 All SA 594 (C)). The crux of the inquiry under s 164(1) was ‘to determine whether the witness [understood] her obligation to testify truthfully’ (at [144]). In this case, said Mjali J, she ‘clearly demonstrated the lack of such obligation and appreciation of the dangers of lying as she [did] not know what happens when one tells lies and [thought] it  . . . good to do so’.

Since it is a ‘precondition for admonishing a child to tell the truth that the child can comprehend what it means to tell the truth’, and since the ‘evidence of a child who does not understand what it means to tell the truth is not reliable’ (see DPP, Transvaal v Minister of Justice and Constitutional Development & others 2009 (2) SACR 130 (CC) at [166]), it would, said Mjali J, undermine the accused’s right to a fair trial if the evidence were to be admitted. In the judge’s view, then, no reliance should have been placed on the evidence of the complainant in this case. It is difficult to disagree. Although the report of the expert clinical psychologist pointed in the other direction, and although the appellants in Haarhoff did not, in the appeal, attack any deficiencies arising out of the court a quo investigation and the failure to take the oath, it is difficult to see how admonishing the witness to tell the truth in the wake of what can only be considered an unfortunate series of questions and answers could be regarded as proper practice.

A far more rigorous investigation was called for, and much more would have been required before the negative impression created by the answers set out above could have been dispelled. It would certainly be more conducive to good practice if the courts were to realise that the inquiry into competence is notionally different from that concerning the taking of the oath. It is logically anterior to the latter and, if it yields an answer that the witness is not competent, renders the latter inquiry superfluous. In the absence of a far more searching and sensitively handled inquiry, it is difficult to see how the complainant in Haarhoff could have been regarded as a competent witness.

ss 151 and 212: Circumstantial evidence, the onus of proof and expert evidence—confusion between the scientific and judicial measures of proof

S v Maqubela 2017 (2) SACR 690 (SCA)

There is a simple and obvious reason that the conviction of the appellant in Maqubela for murdering her husband could not be sustained: the expert testimony made it quite clear that it was reasonably possible that the deceased had died as a result of natural causes and had not been murdered at all. The fact that the Supreme Court of Appeal took great pains to explore such vexed areas of the law as the distinction between applying, in the case of expert witnesses, a scientific measure of proof (which is inappropriate in the context of judicial fact-finding) and a legal or judicial measure (which is not), as well as the rules of inferential reasoning set out in R v Blom 1939 AD 188 at 202–3 (and subjected to searching scrutiny and criticism in CJR 2017 (1)), is therefore a little baffling.

The trial court, said Swain JA (who delivered the court’s judgment), ‘carried out a painstaking and detailed examination of the conflicting expert evidence of Dr Mfolozi, a specialist pathologist called by the State, and Professor Saayman, a specialist pathologist called by the appellant, as to the cause of death of the deceased’. Dr Mfolozi con-ceded that, as he had not examined all of the deceased’s heart, he could not state with certainty that natural causes could be excluded and, further, that ‘it was possible that he may have missed this condition’ (at [14]). Professor Saayman, on the other hand, was of the view that an inference of ‘death by natural causes or other undetected unnatural causes’ and an inference of suffocation were ‘equally possible’.

This prompted the trial court to conclude that Professor Saayman ‘did not state definitively what might have been an operative natural cause’ and that ‘ultimately then, the cause of death cannot be determined by the medical evidence alone’, as it was, ‘in the final analysis,  . . . inconclusive’.

These remarks prompted Swain JA to conclude that the trial court had ‘unfortunately failed to appreciate the distinction’ between the scientific standard of proof, which is the ‘ascertainment of scientific certainty’ and the judicial standard, which is ‘the assessment of probability’ (at [5]) (see, in this regard, the discussion in Commentary in the notes on s 212 sv Expert evidence and the decisions in Michael & another v Linksfield Park Clinic (Pty) Ltd & another 2001 (3) SA 1188 (SCA) at [40], Dingley v The Chief Constable, Strathclyde Police 200 SC (HL) 77 at 89D–E and Oppelt v Department of Health, Western Cape 2016 (1) SA 325 (CC) at [38] and [41]). This is no doubt true. Had the trial court applied the judicial measure, said Swain JA (at [8]), which concerns ‘what the probable cause of death was’, then this assertion would follow: ‘Professor Saayman was of the view that natural causes as the cause of death was the more probable inference to be drawn because ‘‘there was a substantially greater likelihood’’ that the pathology in the deceased’s heart ‘‘could have caused his death’’ and that ‘‘the probabilities are that his heart killed him’’’ (at [8]). The ‘inadvertent application of the scientific measure of proof to the medical evidence, which produced an inconclusive answer to the cause of death, had the serious consequence that the trial court failed to recognise that the opinion of Professor Saayman that the deceased probably died of natural causes, was the correct finding, when the judicial measure of proof was applied to the medical evidence’ (at [13]).

The ‘absence of proof of a probable or certain cause of death’ was, said Swain JA (at [16]), regarded as an essential element in its finding of guilt ‘based solely upon the conduct of the appellant ‘‘showing consciousness of guilt’’’. ‘If the trial court had applied the appropriate judicial measure of proof to the evidence of Professor Saayman, it would have concluded that the deceased probably died of natural causes’ (at [16]; emphasis added). It should, then, have concluded that ‘proof of natural causes as a probable cause of death, precluded a finding of murder’ (emphasis added again).

It is respectfully submitted that the approach of the Supreme Court of Appeal is unnecessarily complicated. The question was not whether death by natural causes was probable but, merely, whether it was reasonably possible. If it was, a fatal blow would have been landed to the State’s attempt to prove murder—by the appellant or anyone else, for that matter. The failure of the expert led by the State to exclude this reasonable possibility and the seemingly clear creation of reasonable doubt by the expert led by the defence should have sufficed. The Supreme Court of Appeal’s insistence on the judicial rather than the scientific measure when assessing expert testimony cannot be faulted. And its deployment of the rules of logic set out in Blom square with the conventional treatment of circumstantial evidence. However, neither of these tools, it is submitted, was needed.


 (ii) Sentencing

Sentencing: The ‘advanced age’ of the convicted offender and life imprisonment as prescribed sentence

S v JA 2017 (2) SACR 143 (NCK)

It is trite law that the tender age or youth of the convicted offender is in principle a mitigating factor. He is almost invariably on account of his level of immaturity in terms of experience and judgment, less blameworthy than the adult offender. See, generally, the discussion under s 277 in Commentary, sv Juvenile offenders. However, where an adult offender has reached an advanced age, age may once again surface as a possible mitigating factor—albeit for different reasons. See S v Barendse 2010 (2) SACR 616 (ECG) at 619c–d. According to Terblanche A Guide to Sentencing in South Africa 3 ed (2016) at 222–3, ‘advanced age’ in this context generally means 60 years and older. This broad observation appears to be in line with our case law. Carnelley and Hoctor 2008 Obiter 268 at 270 have pointed out that reported judgments in South Africa ‘seem to regard a person as elderly from about 58, although that would depend on the offender before the court, especially since old age is often accompanied by another mitigating factor, namely illness or ill health’. Indeed, ‘advanced age’ is a wide concept. For purposes of sentencing, it must necessarily be interpreted with reference to the personal circumstances of each convicted offender. In S v Chabalala 2014 (1) SACR 458 (GP) at [9] it was said that the accused, a 65-year-old pensioner, was ‘in the twilight of his life’. See further the discussion in numbered paragraph 12 under s 277 in Commentary, sv Imposing imprisonment in the absence of minimum sentence legislation—general principles.

In the recently reported case S v JA 2017 (2) SACR 143 (NCK), a full bench had the opportunity to deal with the argument that the sentencing court should, as a mitigating factor and for purposes of minimum sentence legislation, have taken into account that the accused was a relatively old offender. The appellant in this matter had been convicted of raping his 12-year-old daughter on at least three occasions over a period of some thirty months. He was ‘approximately’ 56 years old at the time of the offences but 59 when sentencing procedures commenced (at [5]). At [41] this age was described as ‘relatively advanced’. The sentencing court concluded that there were no substantial and compelling circumstances justifying a lesser sentence than a sentence of life imprisonment as provided for in the Criminal Law Amendment Act 105 of 1997. Due to an incomplete and inaccurate charge sheet, the sentencing court had sentenced the appellant to life imprisonment for having raped a 12-year-old girl and not for having raped her more than once (at [2] and [18]–[20]).

In S v JA (supra) there were two main grounds of appeal. The first one was that the sentence of life imprisonment was ‘disproportionate’ to the appellant’s personal circum-stances and the crimes committed by him (at [17.6]). Olivier J (Kgomo JP and Erasmus AJ concurring) identified good reasons for rejecting this ground. At [25] it was said:

‘[T]he appellant, as the biological father of the complainant and the adult person in whose house she had grown up, had abused his position of trust and had in fact used it to manipulate the complainant to subject herself, without having to apply any violence. The threat to kill the complainant’s mother should also not be lost sight of.’

Olivier J also referred to two Supreme Court of Appeal cases where life imprisonment for incestuous rape of daughters by fathers was confirmed on appeal (at [42]). These two cases are S v PB 2013 (2) SACR 533 (SCA) and S v MDT 2014 (2) SACR 630 (SCA). According to Olivier J there was ‘sufficient similarity for the purpose of comparison’ be-tween these two cases and the facts in JA, to conclude that the sentence was not dis-proportionate ‘to what the appellant had done or to his personal circumstances’ (at [42]). Indeed, Olivier J was satisfied that the facts in JA were, on the whole, more serious than the facts in PB and MDT (at [45]).

The second main ground of appeal—which was really closely linked to the first main ground—was that the appellant’s ‘advanced age  . . . at the time of sentencing had militated against a sentence of life imprisonment  . . .’ (at [17.6], emphasis added). In support of this ground of appeal, counsel for the appellant referred to s 73 of the Correctional Services Act 111 of 1998 (at [30]). This section is headed ‘Length and form of sentences’. It limits and regulates, amongst other matters, the possible placement on day parole or parole of a person who ‘has been sentenced to  . . . life incarceration’. Counsel for the appellant submitted that, having regard to the provisions of s 73 of Act 111 of 1998 and the life sentence imposed by the sentencing court, ‘the appellant would become eligible for parole no sooner than the age of 74’ (see s 73(6)(b)(vi)) and ‘possibly only when reaching the age of 84’ (see s 73(6)(b)(iv)). The full bench in JA was in effect being asked to determine whether the fact that the appellant might spend the rest of his life in prison on account of the sentence of life imprisonment imposed on him at his advanced age, could be a mitigating factor constituting or contributing to, substantial and compelling circumstances justifying a lesser sentence as permitted in terms of Act 105 of 1997.

In addressing the above issue, Olivier J distinguished between the role and functions of a sentencing court and the role and functions of parole boards and officials in the Department of Correctional Services. On the different roles of the judiciary and the executive in this context and the need to affirm and maintain the broad but important doctrine of separation of powers in the sentencing process, see also the discussion of s 276B in Commentary, sv The judiciary and the executive where reference is made to S v Jimmale & another 2016 (2) SACR 691 (CC) and Supreme Court of Appeal cases such as S v Botha 2006 (2) SACR 110 (SCA), S v Mhlakaza & others 1997 (1) SACR 515 (SCA), S v Motloung 2016 (2) SACR 243 (SCA) and S v Matlala 2003 (1) SACR 80 (SCA). See also the discussion of S v Ntozini & another 2017 (2) SACR 448 (ECG) elsewhere in this edition of Criminal Justice Review.

In S v JA at [36] Olivier J confirmed that a sentencing court must see a sentence of life imprisonment ‘as exactly that—imprisonment for the rest of the natural life of the offender’. After all, this was the intention of the legislature when life imprisonment was prescribed for purposes of certain crimes and categories of crimes. And the question whether an offender sentenced to life imprisonment would indeed spend the rest of his natural life behind bars, concerns the executive and not the judiciary. The executive has the statutory power and discretion to release the sentenced offender on parole. Parole is the ‘domain of the executive’ (JA at [37]), whereas the sentencing court is required to determine the maximum period a convicted person may be kept in prison (JA at [38]). See also Motloung (supra) at [18].

At [39] in JA Olivier J made the pertinent observation that ‘[i]t is not for the sentencing court to try to work out how old an offender could be when (if at all) the executive decides to release him  . . . on parole’. At [37] reference was also made to what Howie JA said in Matlala (supra) at [7], namely that a sentencing court must impose the sentence that ‘it intends should be served and it imposes that on an assessment of all the relevant facts before it’.

In JA it was clear, at the time when sentence had to be considered, that the appellant’s ‘relatively old age’ was not accompanied by ill health or physical infirmities or impairment of mental capacity. Furthermore, he had committed the rapes over a period of time during which he had sufficient opportunity to reconsider his actions and ‘come to his senses’ (at [45]). At some stage he had even falsely promised his partner, the mother of the victim, that he would not rape their daughter again (at [45]). The appellant had also acted in a ‘calculated’ manner in that he had created opportunities to be alone with his daughter so that he could rape her (at [46]). Given these circumstances as well as the prevalence of the rape of young girls (at [47]–[49]), the advanced age of the appellant was not and could not be a factor that precluded the imposition of the sentence of life imprisonment as prescribed. At [41] Olivier J concluded that the ‘relatively advanced age’ of the appellant was ‘not  . . . a mitigating factor in the context of a prescribed sentence of life imprisonment and in considering whether there [were] substantial and compelling circumstances justifying a lesser sentence’.

Sentencing an offender who maintains innocence: Rehabilitation, remorse and mercy

The present note provides a survey of several recent cases which dealt very briefly with the possible and permissible views a sentencing court may or should take in determining an appropriate sentence in respect of a convicted offender who has pleaded not guilty, has shown no remorse and persists in maintaining his innocence. On the meaning of ‘remorse’ and issues concerning remorse in the sentencing process, see the case law referred to in the discussion of s 274 in Commentary, sv Mitigating factors: The plea of guilty and remorse and sv The plea of not guilty and remorse: Acceptance of responsibility versus absence of insight.

In S v Ngcukana (unreported, WCC case no A443/15, 18 August 2017) at [97] Rogers J pointed out that in several Supreme Court of Appeal cases the offender’s prospects of rehabilitation were taken into account despite the offender’s persistence in his innocence prior to and after conviction. At [97] Rogers J also said that ‘it would not be in keeping with our constitutional order to hold that the prospect of rehabilitation must be ignored just because the accused, as is his right, maintains his innocence’. In S v Bezuidenhout (unreported, NCK case no CA&R76/2016, 2 December 2016) at [22] Olivier J also noted that an accused has a constitutional right to put the prosecution to proof of its case and that it is accordingly a misdirection if a sentencing court were to view a plea of not guilty and ‘persistence in  . . . innocence’ as aggravating circumstances.

What does happen though, is that the convicted offender who has denied guilt and not taken the court into his confidence ‘is at a disadvantage in advancing the prospect of rehabilitation as a mitigating factor’ (per Rogers J in Ngcukana (supra) at [97]). The other difficulty is that it becomes impossible for the convicted offender to rely upon remorse as a mitigating factor. The matter was put into proper perspective by Steyn J in S v S (unreported, KZD case no AR233/05, 22 March 2017) at [16]:

‘Whether an accused professes remorse is not the test. The penitence must be sin-cere and an offender should take the court into his confidence. As can be seen from the accused’s own evidence and the experts, he considers himself not guilty. Whilst it is acknowledged that he as of right may challenge any conviction, it cannot be found, given the circumstances and facts of this case, that the accused is remorseful.’

Referring to S v Hewitt 2017 (1) SACR 309 (SCA), Steyn J also pointed out in S v S (supra) at [14] that whilst it is indeed correct that the absence of remorse is not an aggravating factor, it is also true that ‘remorse cannot be taken into account as a mitigating factor if it is not genuine and not displayed in the conduct of the accused’. The absence of remorse can also make it very difficult for a sentencing court to assess the convicted offender’s prospects of rehabilitation. See the observations and findings by Petse ADJP in S v Dyantyi 2011 (1) SACR 540 (ECG) at [26] as read with the remarks by Rogers J in Ngcukana (supra) at [97] that absence of remorse does not necessarily preclude a court from considering prospects of rehabilitation.

In S v Smith 2017 (1) SACR 520 (WCC) Rogers J had another opportunity to deal with an aspect of remorse. In this case the trial magistrate had taken the view that remorse was the ‘flipside of the coin of mercy’ and that, in the absence of remorse, she could not really consider the element of mercy in the sentence she imposed. ‘Here the magistrate’, said Rogers J at [107], ‘fell into error. Mercy is not a reward for remorse’. Referring to cases such as S v Rabie 1975 (4) SA 855 (A), S v Zinn 1969 (2) SA 537 (A) and S v Roux 1975 (3) SA 190 (A), Rogers J explained at [107] that mercy (‘or compassion or plain humanity’) is a balanced and humane way of thinking which ‘infuses the assessment’ of the triad in Zinn, and ‘is not an independent fourth element.

Smith (supra) can hardly be reconciled with S v Ndubane (unreported, GP case no A238/2016, 8 February 2017) at [35] where Madima AJ stated that ‘for mercy to be shown on the appellant, he needed to show remorse’. It is respectfully submitted that the principle stated in Smith (supra) should be followed in preference to that in Ndubane.

Sentencing jurisdiction and minimum sentence legislation

S v Ndlovu 2017 (2) SACR 305 (CC)

The applicant in Ndlovu (supra) had appeared in the regional court on a charge of rape as read with the provisions of s 51(2) of the Criminal Law Amendment Act 105 of 1997. At the commencement of the trial and after the prosecutor had put the charge to the applicant, the regional court magistrate explained to the applicant that if he were convicted as charged, he would—in terms of s 51(2) of Act 105 of 1997 and in the absence of substantial and compelling circumstances—receive a minimum sentence of 15 years’ imprisonment if he were a first offender (at [6]). In Ndlovu it was pointed out in a foot-note that at that time, upon conviction of an offence identified in s 51(2), the minimum sentence for a first offender was 10 years and not 15 (see fn 3 at [6]). However, the ultimate decision of the Constitutional Court in Ndlovu did not turn on the correctness or otherwise of the magistrate’s statement immediately after the charge was put.

In Ndlovu the regional court convicted the applicant ‘as charged’, that is, of rape as read with s 51(2) of Act 105 of 1997 (at [43]–[45] and [48]). Khampepe J, writing for a unanimous Constitutional Court, stated that this finding on the merits was ‘unambiguous’ (at [44]). But ‘in a perplexing turn of events’ on the same day as the conviction, the regional court proceeded to sentence the applicant to life imprisonment in terms of s 51(1) despite his having been charged with and convicted of rape as read with s 51(2) and identified in Part III of Schedule 2 to Act 105 of 1997 (at [7] and [44]). However, the regional court took the view that the rape involved the infliction of serious bodily harm and therefore fell within the ambit of s 51(1) which compelled the court to impose life imprisonment in the absence of substantial and compelling circumstances (at [8]). See ss 51(1) and 51(3) of Act 105 of 1997 as read with Part I of Schedule 2 to the same Act.

The applicant’s appeal to the High Court failed on the basis that the initial and repeated reference to the wrong section in Act 105 of 1997 had not caused prejudice to the applicant. The High Court was also convinced that on account of the violent and serious nature of the rape, the regional court had not erred in imposing life imprisonment (Ndlovu at [11]–[12]). The Supreme Court of Appeal confirmed the findings and conclusions of the High Court: the incorrect reference in the charge sheet had not infringed the constitutional fair-trial right of the applicant and the latter would not have conducted his case differently (at [14]–[15]). The Supreme Court of Appeal also concluded that there were no substantial and compelling circumstances justifying a departure from the prescribed minimum sentence (Ndlovu at [16]).

However, neither the Supreme Court of Appeal nor the High Court had considered the matter from the perspective of sentencing jurisdiction. The issues before the Constitutional Court were formulated as follows by Khampepe J (at [2], emphasis added):

‘The central question is whether [the applicant’s] right to a fair trial was infringed when, after he had been charged with rape read with one minimum-sentencing provision, he was sentenced pursuant to a different, harsher, minimum-sentencing provision. This matter also raises the threshold question whether the regional court had the requisite jurisdiction to sentence him to life imprisonment in the circumstances.’

The ‘threshold’ issue had to be addressed first: if the regional court had lacked jurisdiction to impose life imprisonment on the applicant as provided for in s 51(1) as read with Part I of Schedule 2, that would be the end of the matter and the sentence had to be set aside. The need to address fair-trial issues concerning sentencing, would then not arise (at [24]).

As a point of departure, Khampepe J stated that it is ‘trite that magistrates’ courts are creatures of statute and have no jurisdiction beyond that granted by the Magistrates’ Courts Act and any other  . . . statutes’ (at [41]; see also the discussion of S v Ntozini & another 2017 (2) SACR 448 (ECG) elsewhere in this edition of Criminal Justice Review). In Ndlovu the regional court’s sentencing jurisdiction in terms of s 51(2) of Act 51 of 1997 was limited to a maximum of 15 years. See fn 28 at [41] in Ndlovu. But the regional court, said Khampepe J, had imposed ‘life imprisonment under s 51(1), which it would have had the power to do only if the application of the section were triggered’ (at [41], emphasis in the original).

On the facts there was nothing that had put, or could have put, s 51(1) into operation. The State argued that in terms of s 88 of the Criminal Procedure Act the defect in the charge had been cured automatically by the evidence of the victim’s injuries. This argument was rejected because the charge ‘was complete and not defective’ (at [45]). Khampepe J also pointed out that the regional court magistrate could and should have taken steps to ensure that the applicant was prosecuted or convicted in terms of the correct provision of Act 105 of 1997 (at [56], emphasis in the original):

‘Courts are expressly empowered in terms of s 86 of the Criminal Procedure Act to order that a charge be amended. Upon realising that the charge did not accurately reflect the evidence led, it was open to the court at any time before judgment to invite the state to apply to amend the charge and to invite [the applicant] to make submissions on whether any prejudice would be occasioned by the amendment. This the magistrate failed to do. It was only after conviction, at sentencing, that she sought to invoke the correct provision. This failure is directly implicated in the finding made in this judgment.’

Khampepe J also pointed out that the prosecutor’s failure to draft an accurate charge was unacceptable. The injuries sustained by the rape victim were properly recorded in the J88 form which was available to the prosecution when it decided to prosecute. A proper charge of rape read with the provisions of s 51(1) should have been preferred; and the prosecutor’s ‘remissness  . . . should have been corrected by the court’ (at [58]).

The conclusion of the Constitutional Court was that the sentence of life imprisonment imposed by the regional court was beyond the latter’s sentencing jurisdiction. The life imprisonment had to be set aside (at [48]); and it was also found ‘unnecessary to consider the fair-trial question’ (at [47]).

The Constitutional Court decided that it was not in the interests of justice to return the matter to the trial magistrate for purposes of considering and determining a new sentence. The length of time that had passed since the applicant’s trial, ‘eroded’ the benefits of having the trial court imposing the new sentence (at [49]). Finality was necessary.

The Constitutional Court, in considering the new sentence, took the following main factors into account: first, the new sentence had to be imposed within the limitations of the regional court’s jurisdiction in terms of s 51(1) of Act 105 of 1997 (at [49]); second, the maximum sentence that could have been imposed by the regional court was 15 years’ imprisonment; third, the applicant was a first offender and therefore the minimum applicable sentence was 10 years; fourth, the circumstances of the rape were ‘especially heinous’ and the applicant had ‘viciously and mercilessly assaulted and raped’ the victim, who had to spend five days in hospital (at [50]); fifth, the seriousness of the offence was such that the minimum sentence of 10 years’ imprisonment would be ‘grossly inadequate’ (at [51]).

A new sentence of 15 years’ imprisonment was imposed by the Constitutional Court, the maximum permissible number of years that it could impose having regard to s 51(2) and the regional court’s sentencing jurisdiction (at [52] and [59]). In this respect it should be noted that s 51(2) contains a proviso to the effect that any term of imprisonment that a regional court may impose in terms of this subsection, ‘shall not exceed the minimum term of imprisonment that it must impose in terms of this subsection by more than five years’.

One gets the impression that the rapist in Ndlovu had probably deserved life imprisonment. But the fact of the matter is that the problem concerning sentencing jurisdiction was insurmountable; and the final outcome was a disappointing but inevitable one. As a parting shot Khampepe J said: ‘The failings of the prosecutor are directly to blame for the outcome of this matter’ (at [58]).

s 276B(1): Invalid non-parole periods and sentencing jurisdiction

S v Ntozini & another 2017 (2) SACR 448 (ECG)

The two accused in the above special review had pleaded guilty in the magistrate’s court. They were convicted in terms of their pleas. Accused 1 was sentenced to three years’ imprisonment. As part of this sentence, the sentencing court also ordered that he should serve his sentence at Cradock prison, that he be assessed there ‘and be enrolled for the courses offered by the said institution—eg, woodwork/plumbing etc for the duration of his sentence’ (at [3]). In the course of her judgment on sentencing the magistrate had also explained to accused 1 that in terms of her order, he would for the duration of his sentence have to attend the courses and would not be considered for parole without having acquired ‘the skills’ that she had in mind (at [17]). Accused 2 received a sentence of two years’ imprisonment. Here, too, an order similar to the one in respect of accused 1 was made, except for the fact that in respect of accused 2 the requirement was that he should enrol in ‘skills/trade courses  . . . for the duration of his sentence’.

The above sentences made no reference to s 276B(1) of the Criminal Procedure Act and did not use the term ‘non-parole period’. However, Beard AJ (Roberson J concurring) was satisfied that the magistrate had ‘effectively imposed a non-parole period’ in respect of each accused. The review court therefore had to determine whether the sentences were competent in terms of the provisions of s 276B(1) and, furthermore, whether there was compliance with certain procedural rules established in the case law relating to the interpretation and application of this section. See in this regard the discussion of s 276B in Commentary, sv Procedural matters and the requirement that the court should give reasons. In that discussion reference is made to cases like S v Jimmale & another 2016 (2) SACR 691 (CC); S v Strydom [2014] ZASCA 29 (unreported, SCA case no 20215/2014, 23 March 2015); S v Mhlongo 2016 (2) SACR 611 (SCA); S v Stander 2012 (1) SACR 537 (SCA); S v Pauls 2011 (2) SACR 417 (ECG) and S v Madolwana (unreported, ECG case no CA&R 436/12, 19 June 2013).

Section 276B(1) provides that where imprisonment for two years or longer is imposed, the sentencing court ‘may as part of the sentence, fix a period during which the [sentenced offender] shall not be placed on parole’ (s 276B(1)(a)). This period is known as the ‘non-parole period’ and ‘may not exceed two thirds of the term of imprisonment imposed or 25 years, whichever is the shorter’ (s 276B(1)(b)).

In S v Ntozini & another (supra) at [12] Beard AJ pointed out that prior to the enactment of s 276B, courts—which derive their sentencing jurisdiction from statute—had no statutory power to determine non-parole periods. The taking of decisions concerning release on parole was—in accordance with the separation of powers doctrine—viewed as an executive function within the ‘exclusive jurisdiction of the Department of Correctional Services’. It is against this background that s 276B has to be interpreted and applied. The practical effect of a sentencing court’s limited power to determine and order a non-parole period as provided for in s 276B—and thus enter a domain traditionally reserved for the executive—was explained as follows in Ntozini at [13]:

‘When making an order in terms of s 276B(1), the sentencing court, in effect, makes a ‘‘present determination’’ that the convicted person will not merit being released on parole in the future, notwithstanding that the decision as to the suitability of a prisoner to be released on parole involves a consideration of facts relevant to his conduct after the imposition of sentence. It is thus a ‘‘predictive judgment’’ as to the likely behaviour of the convicted person in the future, reached on the basis of the facts available to the sentencing court at the time of sentence.’

The above passage is an accurate summary of what was said in cases like Stander (supra), Strydom (supra) and Madolwana (supra) at [7]. In fact, in Stander (supra) at [13] it was specifically pointed out that the Correctional Services Act 111 of 1998 contains provisions regulating a sentenced offender’s parole and that the Department of Correctional Services ‘and not a sentencing court, is far better suited to make decisions about the release of a prisoner on parole and  . . . it remains desirable to respect the principle of separation of powers in this regard’. In Stander at [16] Snyders JA also noted that s 276B ‘is an unusual provision and its enactment does not put the court in any better position to make decisions about parole than it was prior to its enactment;. See also the remarks made by the Constitutional Court in Jimmale (supra) at [13].

To return to Ntozini: Beard AJ confirmed the need for a proper factual basis before a court can order a non-parole period. And a proper ‘judicial consideration’ of the facts can only be undertaken once both parties have had an opportunity to address the court on this matter (at [16]): see also Jimmale (supra) at [13]. In Ntozini the magistrate gave no indication to the prosecution and the convicted offenders that she was considering a non-parole order; and they were at no stage invited to make submissions on the issue (at [18]). This was a serious misdirection and also infringed ‘the  . . . right to a fair trial as enshrined in the Constitution’ (at [16]). Indeed, in terms of the Constitutional Court’s decision in Jimmale, a sentencing court ‘should invite and hear oral argument; on the issue (at [20]) and, where necessary, receive further evidence (at [13]).

The magistrate’s misdirection in Ntozini was sufficient ground for the court of review to delete the non-parole periods stipulated by the magistrate. However, Beard AJ pro-ceeded to point out that the magistrate’s non-parole periods also exceeded the maximum non-parole period permitted by s 276B(1)(b), namely two thirds of the term of imprisonment. See also the discussion of s 276B in Commentary, sv Section 276B(1)(b): Limitation on the length of the non-parole period. On this basis too, the non-parole periods had to be set aside (at [19]).

However, mere deletion of the non-parole periods stipulated in the magistrate’s sentences was not the end of the matter because the balance of the magistrate’s orders, which formed part of the sentence, fell outside the sentencing jurisdiction of the magistrate. At [21] Beard AJ remarked and found as follows:

‘There is no provision in the [Criminal Procedure Act] or other legislation that permits a district magistrates’ court to direct where the accused person will serve out his sentence. Nor is there any statutory provision permitting such a court to order that an accused person be enrolled in skills-transfer courses whilst serving the term of his imprisonment. These functions fall exclusively within the purview of the executive. In exceeding her jurisdiction by making these orders the magistrate has fallen foul of the separation-of-powers doctrine.’

But Beard AJ had a further and final problem with the prison sentences imposed by the magistrate: the sentences—even when trimmed of the non-parole periods and the or-ders pertaining to the specific prisons and courses—were ‘shockingly’ severe and required review court interference (at [24]–[25]). The two accused were first offenders. They were ‘relatively young’ and had admitted guilt: the 21-year-old accused 1 had pleaded guilty to housebreaking with intent to steal and theft, and the 19-year-old accused 2 had pleaded guilty to receiving stolen property (at [26]). Both accused left school upon completing grade 8 and were not in full-time employment when the crimes were committed. At [29] Beard AJ concluded that the magistrate erred in assuming that the two accused would commit further crimes simply because they were unemployed and unskilled. ‘A sentencing court’, it was said at [29], ‘simply cannot impose a sentence of direct imprisonment as a means of ensuring that accused persons acquire skills through the rehabilitation programmes run by the Department of Correctional Services’.

The sentences and orders in respect of both accused were set aside and replaced with the following: one year’s imprisonment with seven months suspended in respect of accused 1, and eight months’ imprisonment with four months suspended in respect of accused 2.


 (iii) Forfeiture and confiscation

Forfeiture of property: s 50(1) of POCA—Act not intended to allow for intervention in a commercial dispute

National Director of Public Prosecutions v Kalmar Industries SA (Pty) Ltd 2017 (2) SACR 593 (SCA)

The issue in this case was whether a lifting platform and certain tools, equipment and other items could be made the subject of a preservation order under s 38 of POCA and a forfeiture order under s 48 read with s 50 of that Act. The court a quo found that they could not, holding that the NDPP had not met the jurisdictional requirements for either order.

Section 38(2) of POCA enjoins a court to make a preservation order if there are reasonable grounds to believe that the property concerned is an ‘instrumentality of an offence’ referred to in Schedule 1 or the ‘proceeds of unlawful activities’. Section 48(1) read with s 50(1) provides for a forfeiture order to be made if the court finds, on a balance of probabilities, that the property is an instrumentality of an offence referred to in Schedule 1 or is the proceeds of unlawful activities.

The question to be asked, said Schippers AJA (who delivered the judgment of the court), was why the NDPP applied for such orders under POCA to be granted in the first place. All the documents before the NDPP pointed to a commercial dispute between two contracting parties, and not to the commission of a crime. One party claimed ownership of the platform and equipment as well as payment for work done in terms of a contract. The other party claimed that the first party did not comply with the agreement. A contention that the second party had unlawfully appropriated the property was unsustainable on the evidence.

The court a quo had proceeded on the assumption that the property had been stolen, but made it clear that theft was in dispute and had not been proved. Having regard to the commercial nature of the dispute, said Schippers AJA (at [17]), it could hardly have been contemplated that theft could be proved. The NDPP must have known, too, that there were many genuine disputes of fact which could never have been resolved on the papers in motion proceedings involving the two parties.

More fundamentally, said Schippers AJA (at [19]), the commercial dispute between the two parties ‘was far removed from the objectives of POCA’ which was enacted, inter alia, to combat organised crime, money laundering and criminal gang activities, as well as to prohibit certain acts relating to racketeering activities. The dispute in question had nothing to do with the purposes of POCA forfeiture orders, which included ‘removing incentives for crime; deterring persons from using or allowing their property to be used in crime; eliminating or incapacitating the means by which crime may be committed; and advancing the ends of justice by depriving those involved in crime of the property concerned’ (see NDPP v RO Cook Properties (Pty) Ltd; NDPP v 37 Gillespie Street Durban (Pty) Ltd & another; NDPP v Seevnarayan 2004 (2) SACR 208 (SCA) at para 18, discussed in Commentary in the notes to s 20 sv Seizure and forfeitures under the Prevention of Organised Crime Act 121 of 1998).

On this basis alone, said Schippers AJA, the appeal fell to be dismissed: after three years, the NDPP had still not decided to institute criminal proceedings arising out of the theft complaint, and there was nothing to show that the platform and equipment were either instrumentalities of crime or the proceeds of unlawful activities. In determining whether property was an ‘instrumentality of an offence’, the focus was on the role played by the property in the commission of the crime, and not on the wrongdoer. A ‘functional relation’ had to be established between the property and the crime, in that it had to ‘play a part, in a reasonably direct sense, in those acts which constitute the actual commission of the crime in question’ (see Cook Properties) supra at para 32). The very word ‘instrumentality’ itself suggested that the property had to be ‘instrumental in and not merely incidental to the commission of the offence’.

‘Instrumentalities’, said the court, are treated as a form of ‘guilty property’: it is the property itself that is proceeded against, ‘as if it were living and not inanimate’. Examples from the case law included: a houseboat with particular attractions to lure minors into falling prey to sexual offences; a ski-boat and diving equipment used to harvest perlemoen unlawfully; a house specially adapted and equipped to manufacture or conduct a trade in drugs; and a house used to sell liquor unlawfully (see the cases listed in [24] of the judgment). In this case the platform and equipment were not instrumentalities of the crime of theft, but were the very things alleged to have been stolen.

Neither were these items the ‘proceeds of unlawful activities’ as understood by POCA: apart from the fact that no criminal conduct at all had been established on a balance of probabilities, the platform and equipment could in no way be seen as property or an advantage or benefit ‘derived, received or retained as a result of crime’ (at [31]).

The application for a forfeiture order was thus without merit.


 Table of Cases

Click here to see the full Table of Cases of the Criminal Justice Review No 2 of 2017


Click here for a printable .pdf version of this Review


Criminal Procedure Handbook 12e
Principles of Criminal Law (inc. Cases & Materials on Criminal Law) (eBook)


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