South Africa: Supreme Court of Appeal

You are here:  SAFLII >> Databases >> South Africa: Supreme Court of Appeal >> 2006 >> [2006] ZASCA 52

| Noteup | LawCite

Fakie NO v CCII Systems (Pty) Ltd (653/04) [2006] ZASCA 52; 2006 (4) SA 326 (SCA) (31 March 2006)

Download original files

PDF format

RTF format

Links to summary

PDF format

RTF format

Bookmark/share this page

Bookmark and Share



THE SUPREME COURT OF APPEAL

OF SOUTH AFRICA


    1. Case no: 653/2004

    2. REPORTABLE



In the appeal between:


S A FAKIE, NO Appellant


and


CCII SYSTEMS (PTY) LTD Respondent



Before: Howie P, Farlam JA, Cameron JA, Heher JA and Cachalia AJA

Heard: Wednesday 15 February 2006

Judgment: Friday 31 March 2006


Contempt of court – constitutional characterisation – standard of proof – whether applicant in civil proceedings must prove requisites beyond reasonable doubt – Applicant failing to establish that Auditor-General acted wilfully or mala fide in failing to comply with court order – ORDER IN PARA 67


Neutral citation: Fakie v CCII Systems (Pty) Ltd [2006] SCA 54 (RSA)



JUDGMENT

_______________________________________________________


CAMERON JA:


  1. In the Pretoria High Court, De Vos J held the appellant, the Auditor-General,1 in contempt of an order of court. As a sanction, she imposed one month’s imprisonment on him, suspended on condition that he comply fully with the order within four weeks from the date of her judgment. This is an appeal with her leave against that order.

  2. The dispute has its origin in a Cabinet decision in June 1997 to buy military equipment. The purchases (the ‘strategic defence packages’) were put out to tender: they included four corvettes. The respondent (CCII), a supplier of military software and computer systems, was a (partially) unsuccessful bidder for a sub-contract in relation to the corvettes. Following widespread claims that the procurement process had been irregular, the Parliamentary Standing Committee on Public Accounts appointed the Auditor-General, the Public Protector and the National Director of Public Prosecutions (the ‘joint investigating team’) to investigate allegations of corruption. In November 2001, a joint report was presented to the President and accepted by Parliament.

  3. CCII was not satisfied with the report’s findings. It asked the Auditor-General under the Promotion of Access to Information Act 2 of 2000 (PAIA) for documentation the joint investigators considered during their investigation. This was refused. CCII then instituted proceedings in the High Court in Pretoria. On 15 November 2002, Hartzenberg J upheld the application.2 He granted CCII an order that required the Auditor-General to provide it with specified records within 40 court days. The order3 referred to these in two parts. Para 1.1 required the Auditor-General to provide –

All draft versions of the report submitted to Parliament by the Joint Investigating Team regarding the so-called Strategic Defence Packages for the procurement of armaments for the South African National Defence Force.’


  1. The second part required the Auditor-General to furnish to CCII, from documentation specified, (a) certain files to the disclosure of which he did not object under PAIA, and (b) a list of the files to the disclosure of which he did object, setting out his grounds of objection. It was clear – and correctly conceded on appeal – that when CCII instituted the present proceedings the Auditor-General had not complied with the order of Hartzenberg J. It is common cause that the Auditor-General released –

  • the files encompassed in the second part of the order (including the four categories of documents specified – audit files; contracts; minutes; and working papers), though only after CCII launched the current application; and

  • the draft reports envisaged in para 1.1 of the order, though only after De Vos J granted leave to appeal against her contempt finding on 24 November 2004.

  1. The issue before us is whether the circumstances in which the Auditor-General complied so late with Hartzenberg J’s order justify De Vos J’s finding that he was in contempt, and her consequent imposition of suspended imprisonment. That depends on the circumstances of the admitted default. But the proper approach to considering those circumstances must first be determined. This requires a consideration of the nature of this form of contempt of court, and – what was much argued before us – whether in these civil proceedings the standard of proof to be applied in determining whether the Auditor-General was in contempt is a balance of probabilities or beyond reasonable doubt.

Contempt of court

  1. It is a crime unlawfully and intentionally to disobey a court order.4 This type of contempt of court is part of a broader offence, which can take many forms, but the essence of which lies in violating the dignity, repute or authority of the court.5 The offence has in general terms received a constitutional ‘stamp of approval’,6 since the rule of law – a founding value of the Constitution – ‘requires that the dignity and authority of the courts, as well as their capacity to carry out their functions, should always be maintained’.7

  2. The form of proceeding CCII invoked appears to have been received into South African law from English law8 and is a most valuable mechanism. It permits a private litigant who has obtained a court order requiring an opponent to do or not do something (ad factum praestandum),9 to approach the court again, in the event of non-compliance, for a further order declaring the non-compliant party in contempt of court, and imposing a sanction. The sanction usually, though not invariably,10 has the object of inducing the non-complier to fulfil the terms of the previous order.

  3. In the hands of a private party, the application for committal for contempt is a peculiar amalgam,11 for it is a civil proceeding that invokes a criminal sanction or its threat. And while the litigant seeking enforcement has a manifest private interest in securing compliance, the court grants enforcement also because of the broader public interest in obedience to its orders, since disregard sullies the authority of the courts and detracts from the rule of law.

  4. The test for when disobedience of a civil order constitutes contempt has come to be stated as whether the breach was committed ‘deliberately and mala fide’.12 A deliberate disregard is not enough, since the non-complier may genuinely, albeit mistakenly, believe him- or herself entitled to act in the way claimed to constitute the contempt. In such a case good faith avoids the infraction.13 Even a refusal to comply that is objectively unreasonable may be bona fide (though unreasonableness could evidence lack of good faith).14

  5. These requirements – that the refusal to obey should be both wilful and mala fide, and that unreasonable non-compliance, provided it is bona fide, does not constitute contempt – accord with the broader definition of the crime, of which non-compliance with civil orders is a manifestation. They show that the offence is committed not by mere disregard of a court order, but by the deliberate and intentional violation of the court’s dignity, repute or authority that this evinces.15 Honest belief that non-compliance is justified or proper is incompatible with that intent.

  6. Before the decision of this court in S v Beyers,16 it was not clear whether disobedience of a civil order could lead to a public prosecution, since prosecutions were (and still are) almost unknown.17 Beyers – which involved the alleged violation of an interdict granted in civil proceedings – ended the uncertainty. The force of the issue lay in the fact that after the alleged violation of the interdict, and while Beyers’s appeal against its grant was pending, he and his opponent reached a settlement in which the latter abandoned the interdict with retrospective effect ‘as if it had never been granted’. The state decided nevertheless to press ahead, and this court held that the private abandonment did not preclude the public prosecution. Steyn CJ emphasised that while mere non-compliance did not necessarily constitute contempt, sustained disregard and flouting of a court order could be calculated to injure and diminish the authority and status of the court.18 He described the procedure in terms of which a litigant can in own interest seek punishment of an opponent for contempt of court to enforce compliance with a court order as ‘ambivalent in nature’ (van tweeslagtige aard)19: while it follows the rules of civil procedure (the ‘contempt in procedure’ of English law), it has not forfeited its criminal dimension. South African case law, he pointed out, repeatedly treats the civil infraction as a crime ‘with no indication that it is regarded as anything other than common law contempt of court’. This appears most clearly ‘from the fact that an ordinary punishment is imposed if the application succeeds’ since ‘imposition of punishment without a crime being committed, would be something repugnant to (onbestaanbaar in) our law’:

Even though enforcement of a civil obligation is the primary purpose of the punishment, it is nevertheless not imposed merely because the obligation has not been observed, but on the basis of the criminal contempt of court that is associated with it. The fact that the punishment is generally suspended on condition of compliance with the order in issue, and that the punishment is thus not enforced if the applicant should abandon his rights under the order, does not detract from this at all. Depending on the nature and seriousness of the contempt, the court would accordingly be able to suspend only a portion of the punishment, and then the abandonment of rights by the applicant would not affect the unsuspended portion.’20

  1. These observations bear directly on the main question of principle in the appeal, on which our approach to the facts it presents must depend. This is whether civil contempt can be established when reasonable doubt exists as to any of the requisites of the crime. The pre-constitutional approach to proof was that once the enforcer established that the order had been granted, and served on or brought to the alleged contemnor’s notice, an inference was drawn that non-compliance was wilful and mala fide, unless the non-complier established the contrary.21 The alleged contemnor bore the full legal burden of showing on balance of probabilities that failure to comply was not wilful and mala fide.22

  2. The question is to what extent the introduction of the Constitution supersedes this (and hence whether constitutional values might imperil the existence of an important enforcement mechanism). Mr Marcus for the Auditor-General made a wide-ranging attack on the employment of civil proceedings to establish contempt, arguing in general terms that use of application procedure was itself unconstitutional (although he conceded that the Auditor-General had himself acquiesced in the use of motion proceedings, and that their propriety could therefore not be challenged in this case). His main contention was that any onus short of the absence of reasonable doubt conflicted with the fair trial guarantees in s 35 of the Constitution. Mr Rogers for CCII urged that there was no reason to deviate from the established common law approach to civil contempt proceedings.

  3. Counsel’s differing contentions are reflected in conflicting high court decisions. In Uncedo Taxi Service Association v Maninjwa,23 Pickering J carefully evaluated the post-constitutional status of civil enforcement of contempt remedies. He concluded that the fact that contempt proceedings are brought summarily by way of notice of motion does not entail inevitable unconstitutional unfairness: the procedure infringes neither the alleged contemnor’s constitutional right to be properly informed of the charge, nor to remain silent, while the question whether the right to adequate legal representation is infringed depends on the facts of each case.24 He considered it clearly unconstitutional, however, to deprive a person of liberty on proof merely on balance of probabilities, holding that in motion proceedings the initiator must establish the offence beyond reasonable doubt. In Uncedo Taxi Service Association v Mtwa,25 Mbenenge AJ endorsed this. He found that once non-compliance and service were proved, it would be in accordance with constitutional principle to place an evidential burden on the alleged contemnor regarding whether disobedience was deliberate and mala fide: in the absence of evidence raising a reasonable doubt, those elements would be established to the requisite criminal standard. In Victoria Park Ratepayers Association v Greyvenouw CC,26 Plasket J gave enhanced voice to the constitutional considerations underlying the reasoning in the Uncedo decisions, applying the criminal onus to the matter before him.

  4. In Laubscher v Laubscher,27 De Vos J dissented from this approach. She emphasised that the initiator desires not merely to punish a respondent, but to enforce compliance with a court order. She considered that there was a striking difference between a public prosecution and a civil proceeding for contempt, since in the former the accused has to contend with the ‘giant machinery of the state’, whereas in civil proceedings the respondent faces only a similarly-resourced applicant. Deprivation of liberty on proof merely on balance of probabilities was therefore a reasonable and justifiable limitation of rights. In the present matter De Vos J followed her earlier ruling: she found that the Auditor-General had failed to discharge the onus he bore to establish that his non-compliance with Hartzenberg J’s orders was not wilful and mala fide. In Deyzel v Deyzel,28 however, van Rooyen AJ declined to follow De Vos J, preferring the Eastern Cape approach.

  5. The full court of the Eastern Cape has subsequently upheld and elaborated on the reasoning on Uncedo and Victoria Park. In Burchell v Burchell,29 Froneman J (Sandi and Dambuza JJ concurring) held that ‘civil contempt’ remains a criminal offence under the Constitution, and that a respondent in such proceedings is inevitably an ‘accused person’ under s 35 of the Bill of Rights. Froneman J pointed out that committal for contempt of court orders raises no conflict with freedom of speech30 or other fundamental rights, but that, on the contrary, compliance with court orders is of fundamental concern to a society that bases itself on the rule of law. The full court thus held that while the applicant has to prove the elements of civil contempt beyond reasonable doubt, the application procedure is constitutionally competent to accommodate the altered onus. The full court also found that since there is a purely civil aspect to the proceedings, a court may issue a declarator that a respondent is in contempt of court, established only on balance of probabilities, together with associated civil relief (such as not suspending the order pending appeal, and barring the contemnor from access to civil courts until the contempt is purged).


Constitutional characterisation of contempt of court

  1. The proper conclusion as to what onus is applicable in contempt proceedings cannot be deduced as a matter of simple typology from the fact that a public prosecution is competent. Beyers affirmed only that civil contempt has not divested itself of a criminal dimension: it did not hold that that its civil character had been erased (for the procedure is ‘tweeslagtig’, and not criminal only). This underlies the finding in Burchell that civil mechanisms designed to induce compliance, short of committal to prison, are competent even when proved only on balance of probabilities.

  2. But this appreciation unavoidably raises the question why a lesser onus should not also be appropriate in at least some committal proceedings, as CCII urged us to find. For though civil contempt applications generally encompass prayers for relief aimed at both punishment and enforcement – the relief sought and obtained in the present case seemingly an instance – an applicant may disavow a punitive purpose and claim committal solely to secure compliance. In such cases, counsel for CCII contended, only the civil aspect of the process is engaged, with the result that imposing a criminal standard of proof is not only inappropriate, but unfair to those entitled to enforce compliance.

  3. This would be correct if one were to deduce the standard of proof simply from the nature of the particular proceeding. But the question requires a broader approach. Looming over the debate about the typology of contempt committal is the more important question of constitutional characterisation, which the Eastern Cape decisions address: does the fact that imprisonment may be sought in committal proceedings purely for enforcement so affect the nature of the means employed that a lesser standard of proof can be justified? Differently put, do constitutional values permit a person to be put in prison to enforce compliance with a civil order when the requisites are established only preponderantly, and not conclusively? In my view they do not, and the Eastern Cape decisions that the criminal standard of proof applies whenever committal to prison for contempt is sought are correct.

  4. There are two principal reasons for this conclusion. The first is liberty: it is basic to our Constitution that a person should not be deprived of liberty, albeit only to constrain compliance with a court order, if reasonable doubt exists about the essentials. The second reason is coherence: it is practically difficult, and may be impossible, to disentangle the reasons why orders for committal for contempt are sought and why they are granted: in the end, whatever the applicant’s motive, the court commits a contempt respondent to jail for rule of law reasons; and this high public purpose should be pursued only in the absence of reasonable doubt.


First consideration: liberty, guilt and incarceration

  1. A long series of Constitutional Court (CC) decisions has established that it is generally impermissible to find an accused guilty of a criminal offence in the absence of conclusive proof of its essential elements. These decisions provide one of the leitmotifs of our democratic jurisprudence, and have led to the invalidation of a number of ‘reverse onus’ provisions, which placed on an accused the legal burden of disproving an essential element of the offence.31 The CC has held however that it is permissible in certain circumstances for an accused to bear the lesser evidential burden of having to advance evidence that raises a reasonable doubt about an element of a crime – absent which the offence is established beyond reasonable doubt.32

  2. The decisions deal with statutory presumptions and reverse onuses.33 But they undoubtedly entail that where the state prosecutes an alleged contemnor at common law for non-compliance with a civil order, the requisite elements must be established beyond reasonable doubt. In such a prosecution the contemnor is plainly an ‘accused person’ in terms of s 35(3) of the Bill of Rights, and enjoys the inter-related rights that s 35(3)(h) confers: to be presumed innocent, to remain silent in the face of the charges and not to testify during the proceedings. By developing the common law in conformity with the Constitution, the reverse onus the accused bore in prosecutions such as Beyers must now be reduced to an evidential burden (as Mbenenge AJ rightly envisaged in the second Uncedo decision). Once the prosecution has established (i) the existence of the order, (ii) its service on the accused, and (iii) non-compliance, if the accused fails to furnish evidence raising a reasonable doubt whether non-compliance was wilful and mala fide, the offence will be established beyond reasonable doubt: the accused is entitled to remain silent, but does not exercise the choice without consequence.34

  3. It should be noted that developing the common law thus does not require the prosecution to lead evidence as to the accused’s state of mind or motive: once the three requisites mentioned have been proved, in the absence of evidence raising a reasonable doubt as to whether the accused acted wilfully and mala fide, all the requisites of the offence will have been established. What is changed is that the accused no longer bears a legal burden to disprove wilfulness and mala fides on balance of probabilities, but to avoid conviction need only lead evidence that establishes a reasonable doubt.

  4. There can be no reason why these protections should not apply also where a civil applicant seeks an alleged contemnor’s committal to prison as punishment for non-compliance. This is not because the respondent in such an application must inevitably be regarded as an ‘accused person’ for the purposes of s 35 of the Bill of Rights. On the contrary, with respect to the careful reasoning in the Eastern Cape decisions, it does not seem correct to me to insist that such a respondent falls or fits within s 35. Section 12 of the Bill of Rights grants those who are not accused of any offence the right to freedom and security of the person, which includes the right not only ‘not to be detained without trial’,35 but ‘not to be deprived of freedom arbitrarily or without just cause’. 36 This provision affords both substantive and procedural protection,37 and an application for committal for contempt must avoid infringing it.

  5. And in interpreting the ambit of the right’s procedural aspect, it seems to me entirely appropriate to regard the position of a respondent in punitive committal proceedings as closely analogous to that of an accused person; and therefore, in determining whether the relief can be granted without violating s 12, to afford the respondent such substantially similar protections as are appropriate to motion proceedings. For these reasons, the criminal standard of proof is appropriate also here.

  6. I follow this path because the civil process for a contempt committal is an oddity that is distinctive in its combination of civil and criminal elements, and it seems undesirable to strait-jacket it into the protections expressly designed for a criminal accused under s 35.38 Certainly, not all of the rights under that provision will be appropriate to or could easily be grafted onto the hybrid process. For similar reasons, the High Court of Australia has observed, in the context of the English-derived process for contempt, that ‘to say that [civilly-initiated] proceedings for contempt are essentially criminal in nature is not to equate them with the trial of a criminal charge’.39

  7. It would certainly be odd to regard the applicant in such proceedings as a prosecutor, not only because of absence of office, but because of the presence of manifest personal interest. During argument there was debate about whether a civil court’s finding of contempt, with concomitant imposition of punishment, would count as a ‘previous conviction’ for purposes of s 271 of the Criminal Procedure Act 51 of 1977 (which permits the prosecution in a criminal trial to prove previous convictions in aggravation of sentence); or whether, if it were, it would feature in the South African Police Services’ SAP 69 register of previous convictions, and what mechanisms might be necessary to ensure that it was so recorded. Neither counsel ventured firm submissions, and the debate was inconclusive.

  8. And indeed, these questions are not before us now, and it is not necessary to decide them: I make only the point that they may be better answered not through a ‘rights-by-category’ analysis, in which the protections afforded depend on whether the respondent is an ‘accused person’ under s 35; but by considering the rights in that provision a ‘relevant background source’ that furnishes values instructive in interpreting the full range of constitutional protections to which the alleged ‘civil contemnor’ sought to be imprisoned as a punishment for disobeying a court order is entitled.40 Certainly, the requirement that proof should be conclusive, and not merely preponderant, seems to me to be among them.41

  9. Since the applicant in punitive committal proceedings must prove contempt beyond reasonable doubt, why should a lesser standard be warranted when committal is sought for coercion alone? In my view, there can be no reason. Pickering J pointed out in Uncedo42 that the application of two different standards of proof, depending on whether the initiator chooses to lay a criminal charge, or proceed civilly, is unwarrantable, because it introduces ‘a certain degree of arbitrariness’. This applies the more if the standard of proof were to depend on the objective with which the initiator proceeds, and would run counter to this court’s analysis in Beyers, which pointed to the ineluctably criminal dimension of the remedy granted even in proceedings aimed at coercion.43

  10. While the applicant may disavow punishment as a motive (a matter to which I return), the means the court is asked to employ remain the same: the public sanction of imprisonment for disobedience of a court order. The invocation of that sanction in my view requires conclusive proof. No less than punitive committal, purely coercive committal uses imprisonment, or its threat; and whenever loss of liberty for disobedience of an order of court is threatened it seems to me necessary and proper that the infraction should be proved conclusively.

  11. Counsel for CCII invoked cases where the Constitutional Court has upheld statutory provisions providing for imprisonment as a process in aid to procuring testimony by a recalcitrant witness in non-criminal proceedings,44 but as Froneman J pointed out in Burchell (para 12), these decisions do not support the contention that committal to prison for civil contempt for coercive reasons should be permitted on less stringent grounds than for the criminal offence. This case squarely raises the question of what standard of proof is constitutionally appropriate in determining whether coercive committal is justified. That question did not arise in the CC cases, which were concerned with other aspects of the procedural and substantive justification of committal. The CC therefore did not consider or decide the question of proof.45

  12. And as O’Regan J pointed out in De Lange v Smuts, the power to imprison for coercive and non-punitive purposes is ‘an extraordinary one’:

The power to order summary imprisonment of a person in order to coerce that person to comply with a legal obligation is far-reaching. There can be no doubt that indefinite detention for coercive purposes may involve a significant inroad upon personal liberty. Clearly it will constitute a breach of s 12 of the Constitution unless both the coercive purposes are valid and the procedures followed are fair. In this case there seems no doubt that the purpose is a legitimate one. It also seems necessary and proper, however, for the exercise of the power to be accompanied by a high standard of procedural fairness.’46

  1. Though O’Regan J dissented from the conclusion of the majority that the power of committal could be constitutionally exercised where a magistrate presided in an insolvency inquiry, there is nothing in the judgments of the other members of the court to suggest that anything less than a ‘high standard of procedural fairness’ is essential in cases of coercive committal: on the contrary, I read the judgments as endorsing the principle. That includes the degree of proof. In my view, ‘high’ procedural fairness requires proof beyond reasonable doubt in regard to wilfulness and mala fides even when coercion and not punishment is the object.


Second consideration: no ‘purely coercive’ contempt committal – the public vindication of judicial authority is always involved

  1. The preceding discussion has assumed that it is possible to disentangle an applicant’s reasons for seeking a respondent’s committal in contempt proceedings, and thus that it is possible for a civil contempt application or order to aim purely at the private object of compliance, drained of punitive dimension. This is not so. The High Court of Australia has expressed ‘considerable difficulty’ with the notion that, in some cases, the purpose or object of the proceedings is punitive and, in others, remedial or coercive. In any event, the court observed, the purpose of the proceedings is not the same as the purpose or object of the individual bringing them. And:

‘… [P]roceedings for breach of an order or undertaking have the effect of vindicating judicial authority as well as a remedial or coercive effect. Indeed, if the person in breach refuses to remedy the position, as is not unknown, their only effect will be the vindication of judicial authority.’

The court went on to hold that ‘purpose or object cannot readily be disentangled from effect’ and that it therefore had to be acknowledged that punitive and remedial objects are ‘inextricably intermixed’.47

  1. In Videotron Ltée v Industries Microlec Produits Électroniques Inc,48 the majority of the Supreme Court of Canada, despite the strong civil setting provided by the Quebec Code of Civil Procedure, similarly held that the penalty for contempt of court, even when used to enforce a purely private order, inevitably involves an element of public law, ‘because respect for the role and authority of the courts, one of the foundations of the rule of law, is always at issue’. And in England it has long been accepted that the applicant in contempt proceedings must establish the requisites beyond reasonable doubt.49

  2. In the United States, the constitutional setting differs markedly from that in other comparable jurisdictions, including ours, since the Due Process clause applies only when the proceedings are properly classified as criminal. The federal courts therefore acquire jurisdiction over state proceedings under that clause only when the proceedings can be so classified, and a bifurcated classification of contempt proceedings for the purposes of applying the federal guarantees of the United States constitution is therefore unavoidable. The approach of the US Supreme Court to the classification of contempt must be seen against this background. It means that typology is determinative of constitutional protection under the clause in question, whereas under the South African Constitution it is not.

  3. And the relevant decisions in the United States, though performing an obligatory bifurcated classification, express some measure of discomfort with it. Thus, the US Supreme Court has observed that it would be ‘misguided’ to attempt to classify relief by reference to the supposed purpose of the laws, since ‘In contempt cases, both civil and criminal relief have aspects that can be seen as either remedial or punitive or both: when a court imposes fines and punishments on a contemnor, it is not only vindicating its legal authority to enter the initial court order, but it also is seeking to give effect to the law’s purpose of modifying the contemnor’s behavior to conform to the terms required in the order’.50 The court has also referred to the distinction between civil and contempt fines as ‘somewhat elusive’.51

  4. Given our very different constitutional setting, the approach of the English, Australian and Canadian courts seems convincing to me. As they have found, there is no true dichotomy between proceedings in the public interest and proceedings in the interest of the individual, because even where the individual acts merely to secure compliance, the proceedings have an inevitable public dimension – to vindicate judicial authority. Kirk-Cohen J put it thus on behalf of the full court, ‘Contempt of court is not an issue inter partes; it is an issue between the court and the party who has not complied with a mandatory order of court’.52 Elaborating this, Plasket J pointed out in the Victoria Park Ratepayers case that contempt of court has obvious implications for the effectiveness and legitimacy of the legal system and the legal arm of government: there is thus a public interest element in every contempt committal.53 He went on to explain that when viewed in the constitutional context –

it is clear that contempt of court is not merely a mechanism for the enforcement of court orders. The jurisdiction of the superior courts to commit recalcitrant litigants for contempt of court when they fail or refuse to obey court orders has at its heart the very effectiveness and legitimacy of the judicial system. … That, in turn, means that the court called upon to commit such a litigant for his or her contempt is not only dealing with the individual interest of the frustrated successful litigant but also, as importantly, acting as guardian of the public interest’.

  1. These expositions seem to me compelling. A court in considering committal for contempt can never disavow the public dimension of its order. This means that the use of committals for contempt cannot be sundered according to whether they are punitive or coercive. In each, objective (enforcement) and means (imprisonment) are identical. And the standard of proof must likewise be identical.

  2. This approach conforms with the true nature of this form of the crime of contempt of court. As pointed out earlier (para 10), this does not consist in mere disobedience to a court order, but in the contumacious disrespect for judicial authority that is so manifested. It also conforms with the analysis in Beyers (para 11 above), where this court held that even though enforcement is the primary purpose of committal, it is nevertheless not imposed merely because the obligation has not been observed, ‘but on the basis of the criminal contempt of court that is associated with it’. The punitive and public dimensions are therefore inextricable: and coherence requires that the criminal standard of proof should apply in all applications for contempt committal.

  3. Finally, as pointed out earlier (para 23), this development of the common law not require the applicant to lead evidence as to the respondent’s state of mind or motive: once the applicant proves the three requisites (order, service and non-compliance), unless the respondent provides evidence raising a reasonable doubt as to whether non-compliance was wilful and mala fide, the requisites of contempt will have been established. The sole change is that the respondent no longer bears a legal burden to disprove wilfulness and mala fides on balance of probabilities, but need only lead evidence that establishes a reasonable doubt. It follows, in my view, that Froneman J was correct in observing in Burchell (para 24) that in most cases the change in the incidence and nature of the onus will not make cases of this kind any more difficult for the applicant to prove. In those cases where it will make a difference, it seems to me right that the alleged contemnor should have to raise only a reasonable doubt.

  4. To sum up:

  1. The civil contempt procedure is a valuable and important mechanism for securing compliance with court orders, and survives constitutional scrutiny in the form of a motion court application adapted to constitutional requirements.

  2. The respondent in such proceedings is not an ‘accused person’, but is entitled to analogous protections as are appropriate to motion proceedings.

  3. In particular, the applicant must prove the requisites of contempt (the order; service or notice; non-compliance; and wilfulness and mala fides) beyond reasonable doubt.

  4. But once the applicant has proved the order, service or notice, and non-compliance, the respondent bears an evidential burden in relation to wilfulness and mala fides: should the respondent fail to advance evidence that establishes a reasonable doubt as to whether non-compliance was wilful and mala fide, contempt will have been established beyond reasonable doubt.

  5. A declarator and other appropriate remedies remain available to a civil applicant on proof on a balance of probabilities.


Application to facts: did CCII show beyond reasonable doubt that the Auditor-General’s non-compliance was wilful and mala fide?

  1. The question therefore is whether CCII proved beyond reasonable doubt that the Auditor-General’s failure to comply timeously with Hartzenberg J’s order was wilful and mala fide. As explained earlier (para 4), the default fell into two categories – that regarding the second part of the order, which related to what the parties referred to as the ‘audit files’; and that relating to the first part, namely the draft reports. Regarding both delays, the Auditor-General committed himself to an extensive answering affidavit in which he volunteered an explanation. At the outset of his argument Mr Marcus conceded that the Auditor-General’s undemurring acquiescence in this procedure precluded a challenge, in the present case, to its constitutionality. But the discussion above has necessarily traversed the propriety of such proceedings, whose use (subject to case-by-case clarification of the respondent’s constitutional protections) has been found to pass constitutional muster.

  2. As mentioned earlier (para 4), after the proceedings were instituted, and before De Vos J heard the matter and gave judgment on 14 October 2004 (apparently on the same day), there was compliance in relation to the four categories of documents mentioned in part two of Hartzenberg J’s order (audit files; contracts; minutes; and working papers). After De Vos J granted the Auditor-General leave to appeal in November 2004, the Auditor-General supplied CCII with all the draft reports, so complying with part one.

  3. CCII’s complaint is thus two-fold: (i) that the Auditor-General had not complied with the bulk of the order when the present proceedings were instituted; (ii) that he supplied the draft reports only after De Vos J delivered judgment in October 2004. Issue (i) relates principally to the costs in the court below; but issue (ii) – on which the great bulk of the argument before us focused – goes to the heart of De Vos J’s finding of contempt and the suspended penalty she imposed for it.

  4. The 40 days Hartzenberg J granted for compliance with his order expired on 12 February 2003; but before this the Auditor-General applied for leave to appeal. This had the effect of suspending the order. The application was set down for hearing on 13 March 2003. But the Auditor-General withdrew it after discussion between the parties’ legal representatives seemed to result in a common understanding of the meaning and effect of the order.

  5. But this was not so. On 15 May 2003 (approximately forty days after the application for leave to appeal was withdrawn), the Auditor-General’s attorneys made a first delivery to CCII’s attorneys. It was a bundle of documents consisting of some 751 pages, together with a schedule listing the documents he refused to deliver. As regards draft versions of the joint investigating team’s report, he provided two documents: a draft report of the Public Protector consisting of 78 pages, and chapter 12 of a draft report prepared by the Auditor-General, consisting of 21 pages.

  6. CCII’s attorneys protested in a letter of 27 May 2003 that para 1.1 of Hartzenberg J’s order contained ‘no limitation’. The Auditor-General was obliged to provide ‘all and complete draft versions of the investigation report’: ‘By this is meant provision of copies of all drafts of the report of the three investigating agencies as well as those of the joint report.’ The Auditor-General’s attorneys in reply insisted that ‘Your clients’ entitlement in terms of the order of court was for documents relating to the reduced record. This relates to the draft report as well.’ The letter continued: ‘Our client is in the process of examining whether or not there are further documents which it may be obliged to disclose and will revert in due course.’

  7. On 12 June 2003, CCII launched the present proceedings. Apart from costs CCII sought (a) a declaration that the Auditor-General had failed to comply fully with the order of Hartzenberg J; (b) a declaration that he was in contempt of the order; (c) a direction that he comply fully with the order within two weeks; and (d) as a sanction for the contempt, the imposition of one month’s imprisonment, suspended on condition of timeous compliance with (c). Only thereafter was there compliance in regard to the audit reports.

  8. The Auditor-General’s answering affidavits – lodged on 31 July 2003, while the draft reports were still outstanding –

  1. asserted that it had always been his intention to comply with the order, and affirmed that he was committed to compliance, since ‘particularly given the constitutional obligations of the institution of Auditor-General it would be remiss in the extreme not to comply with a court order’; and

  2. stated that he was still ‘endeavouring to comply fully with the order’, was doing ‘everything possible to compile the documentation’ ordered to be disclosed, and that it was ‘plain’ that he had endeavoured, to the best of his ability, to comply with the order.

  1. With regard to the ‘audit files’, he set out the complexity and scope of the task required, as well as the administrative arrangements he undertook to ensure compliance. He emphasised that at some stage 12 persons, including himself, worked on the matter simultaneously: since ‘a proper and credible process had to be followed in considering these documents, any further resources assigned to the task would not have speeded up the process’:

Further, I wish to emphasise that no matter how vast the teams may have been, at some point each and every document had to be considered by me personally, so that I could satisfy myself that I was indeed complying fully with the Order.’

Regarding the first part of Hartzenberg J’s order, he stated that ‘My obligation relates only to the reduced record,’ and asserted that ‘all draft reports relating to the reduced record have been released’ to CCII.

  1. Mr Rogers urged us to endorse De Vos J’s finding that the Auditor-General was in clear contempt of Hartzenberg J’s ruling and that, given the clarity of the terms of that order, the only appropriate inference was that he had acted wilfully and mala fide.

  2. I am unable to accept this submission, or to agree with De Vos J’s approach to the evidence. Central here is that not only the Auditor-General committed himself to motion proceedings: CCII did too. Large in that choice loomed the fact that the parties were in dispute about the reasons and justifications for the admitted failure to comply timeously with Hartzenberg J’s order. The Auditor-General asserted that his default was unintentional: in the case of the audit files, it was because of the administrative burden of compliance; and in the case of the draft reports it was because of the interpretation he accorded the order. CCII asserted that there was no justification for the default and that the inference to be drawn from the Auditor-General’s own account was that his non-compliance was wilful and mala fide.

  3. How was this factual dispute to be resolved? CCII did not ask for the matter to be referred to oral evidence, or for the Auditor-General to be cross-examined. Had that happened, and had the Auditor-General given viva voce evidence – or declined to do so – the disputed facts would have been determined in effect by a live contest resulting in a trial of the issues. Instead, CCII chose to argue, on the opposing affidavits, that the requisites for contempt of court had been fulfilled.

  4. That conflicting affidavits are not a suitable means for determining disputes of fact has been doctrine in this court for more than 80 years.54 Yet motion proceedings are quicker and cheaper than trial proceedings, and in the interests of justice courts have been at pains not to permit unvirtuous respondents to shelter behind patently implausible affidavit versions or bald denials. More than sixty years ago, this court determined that a judge should not allow a respondent to raise ‘fictitious’ disputes of fact to delay the hearing of the matter or to deny the applicant its order.55 There had to be ‘a bona fide dispute of fact on a material matter’.56 This means that an uncreditworthy denial, or a palpably implausible version, can be rejected out of hand, without recourse to oral evidence. In Plascon-Evans Paints Ltd v Van Riebeeck Paints (Pty) Ltd,57 this court extended the ambit of uncreditworthy denials. They now encompassed not merely those that fail to raise a real, genuine or bona fide dispute of fact, but also allegations or denials that are so far-fetched or clearly untenable that the Court is justified in rejecting them merely on the papers.

  5. Practice in this regard has become considerably more robust, and rightly so. If it were otherwise, most of the busy motion courts in the country might cease functioning. But the limits remain, and however robust a court may be inclined to be, a respondent’s version can be rejected in motion proceedings only if it is ‘fictitious’ or so far-fetched and clearly untenable that it can confidently be said, on the papers alone, that it is demonstrably and clearly unworthy of credence.

  6. Can the Auditor-General’s version be rejected on the affidavits as ‘fictitious’, or as demonstrably uncreditworthy? In my view, clearly not. Regarding his collation of the audit files, which CCII received in adequate form only after the current proceedings were initiated, CCII understandably complains that on the Auditor-General’s own version too little resources were devoted to the task, and too late. But it is clear that he did not sit idly by. He assigned staff to the task and engaged himself in it as set out in the extract quoted above. He gives details of these efforts. And throughout, he asserts the good faith of his efforts to comply with the order.

  7. Mr Rogers for CCII subjected the Auditor-General’s account to searching criticism and on the affidavits alone there certainly appear to be gaps and insufficiencies in the account tendered. Despite this, I do not think that his assertions can be rejected as fictitious or as so implausible as to warrant dismissal without recourse to oral evidence.

  8. The draft reports stand on a different footing. The order Hartzenberg J granted is unambiguous. It requires the Auditor-General, without qualification, to hand over ‘all draft versions’ of the joint investigating team’s report. In his answering affidavit, the Auditor-General asserts only that this refers to the ‘reduced record’. His stance requires some background. In the proceedings before Hartzenberg J – the record of which the parties agreed during argument should be placed before us; rightly so, given its relevance – the Auditor-General’s answering affidavit made much of the bulk of the material that he and his staff would be required to peruse if the application for access were granted. In reply, Dr Richard Young, the managing director of CCII, explained that he was ‘not aware’ when launching the application of the bulk of the record. He therefore emphasised that CCII’s interest lay solely ‘in that portion of the record relating to its complaints’. The respondents knew, he asserted, that CCII’s complaints ‘relate exclusively’ to the acquisition of the corvettes and its deselection as a supplier (together with related issues). He continued:

Although [CCII] believes that the [Auditor-General] has overstated the magnitude of the burden of complying with [CCII’s] request, [CCII] is willing for purposes of the present proceedings to confine its request to that portion of the [Auditor-General’s] record which relates to the matters specified … above. I shall refer to this portion of the record as “the Reduced Record”.’

The deponent then challenged the Auditor-General to –

supplement his answering affidavit by indicating whether he is willing to give [CCII] access to the Reduced Record and, if not, to justify such refusal’.

  1. It is against the background of this fact – that CCII sought access only to ‘the Reduced Record’ – that Hartzenberg J issued his order. And the Auditor-General’s correspondence consistently claimed thereafter that the first part of the order, like the second, referred only to ‘the reduced record’. In his answering affidavit in the contempt proceedings, he also claimed that the draft reports to be furnished related solely to ‘the reduced record’, and asserted that he had ‘released all draft versions of the report submitted to Parliament in my possession to which [CCII] is entitled in terms of my understanding of the court order’.

  2. This was wrong. There is no ambiguity in Hartzenberg J’s order. But is it possible to find on the affidavits alone, as CCII urged, that the Auditor-General’s stance was wilful and mala fide? I do not think so. Telling in this regard is CCII’s own correspondence. Twice it stumbles, not meaningfully, but tellingly, over what ‘the reduced record’ constitutes and whether it encompasses the draft reports. Thus, its attorneys’ letter of 23 May 2003 clearly (and correctly) asserts that ‘there is no limitation in Paragraph 1 of the Order’. But the letter states earlier that CCII’s counsel informed the Auditor-General’s counsel ‘that they, our client and the writer all interpreted the Order as referring to the so-called reduced record’. Notable here is that ‘the Order’ is referred to without differentiating its parts or their application in relation to ‘the Reduced Record’ – which chimed with the Auditor-General’s stance regarding the draft reports. Similarly, in a letter of 27 May 2003, CCII’s attorneys again allude to the reduced record, stating that ‘so as not to allow any possibility for ambiguity to confuse your client about the documents which are due to our client under the Order, this includes all documents relating to the corvette component of the Strategic Defence Packages’ – which again accords with the Auditor-General’s claim that only those draft reports bearing on the reduced record had to be released.

  3. These statements must of course be read in the light of CCII’s sustained insistence that the first part of the order encompassed all draft reports, not only those in the ‘reduced record’; but the ambiguity of expression is not only unmistakable, but significant, for it runs counter to CCII’s submission that there was no rational basis or explanation for the Auditor-General’s understanding of the order. In my view, the Auditor-General’s claim that he so understood the order – though clearly wrong – is not entirely incapable of comprehension.58 Mr Rogers for CCII emphasised that the Auditor-General had not claimed to rely on legal advice in taking his stand on the meaning of the order. That is true; but his stance is nevertheless not capable of being rejected on the papers as ‘fictitious’ or palpably uncreditworthy, without his being afforded an oral hearing.

  4. In the light of the proper approach to deciding factual disputes in motion proceedings, I should add that on the particular form of process the parties committed themselves to in this case I do not think that it would make any difference had the onus been only proof on balance of probabilities. The accepted approach requires that, subject to ‘robust’ elimination of denials and ‘fictitious’ disputes, the court must decide the matter on the facts stated by the respondent, together with those the applicant avers and the respondent does not deny. On that approach, since the Auditor-General’s version cannot legitimately be ‘robusted’ away, his factual assertions, including those regarding his state of mind, must be accepted as established. The proven facts thus establish more than just a reasonable doubt, but a factual picture that entails acceptance of the Auditor-General’s version; though that is incidental to the form of the proceedings before us.

  5. To summarise: On the accepted test for fact-finding in motion proceedings, it is impossible to reject the Auditor-General’s version as ‘fictitious’ or as clearly uncreditworthy. There is a real possibility that if a court heard oral evidence on the factual disputes between the parties, it might accept the Auditor-General’s version, or at least find that there was reasonable doubt as to whether the delay in complying with the orders of Hartzenberg J was wilful and mala fide. CCII therefore failed to prove that the default was wilful and mala fide.

  6. The finding of contempt and with it the penalty cannot stand. That is not however an end of the matter. The first part of the order De Vos J granted was a declaration that the Auditor-General ‘has failed to comply fully with this court’s order of 25 November 2002’. The third part was a directive to the Auditor-General ‘to comply fully with the said order within a period of four weeks from the date of this order’. Those portions of the order were plainly justified. So although CCII has failed in its quest for a full finding of contempt, with a concomitant penalty, it was entitled at least to the declarator and to the directory order.

  7. For this reason, the costs order in the court below should remain undisturbed. And even though the Auditor-General has had substantial success on appeal, the litigation and the central question of principle in this court had a novel constitutional character which in my view would make it unjust to burden CCII with the costs of the proceedings in this court. The parties should therefore bear their own costs of appeal.

  8. The appeal accordingly succeeds. The order of the court below is set aside to the extent that the finding of contempt and the associated penalty are set aside. The appellant is to pay the respondent’s costs in the court below. There is no order on the costs of the appeal.



E CAMERON

JUDGE OF APPEAL



CONCUR:

HOWIE P

CACHALIA AJA

HEHER JA:

[68] The facts which Cameron JA has so meticulously analysed satisfy me that the appellant must succeed, whether one applies the civil or criminal standard of proof to them. There is no doubt that the appellant was at all material times able to comply with the terms of the order. But, without testing under cross-examination, the materials were lacking for a rejection of the appellant’s (non-wilful and bona fide) state of mind as expressed in his affidavit and supported by the trend of the correspondence from his attorneys.

[69] Since Cameron JA has thought it necessary to undertake an analysis of the onus in civil contempt proceedings I deem it advisable to express my views on that subject lest silence (or equivocation) be taken for assent.

[70] I agree that s 35 of the constitution is not engaged by the substance of such proceedings. For the reasons which follow the influence of that section on the protection afforded by s 12(1)(a) and (b) must be very slight.

[71] I also agree that, since all applications for committal for civil contempt carry the threat of imprisonment, s 12 is immediately engaged and the respondent must be accorded the widest procedural fairness in the conduct of the proceedings which is consistent with the nature and purpose of the remedy. Nevertheless I differ strongly from Cameron JA that the necessary and proper standard of fairness demands (a) that the applicant prove his case beyond a reasonable doubt, and (b) that the existing common law onus which rests on a respondent to prove absence of wilfulness and absence of mala fides requires developing in such a manner as to burden the applicant with proof of wilfulness and mala fides.

[72] The critical point of departure between us seems to be Cameron JA’s acceptance of a material difficulty in separating coercive (or remedial) orders of imprisonment made in civil contempt proceedings from punitive orders. This supposed problem is one which recurs in judgments in many jurisdictions. In my view it is overstated. Its solution is cardinal to the proper categorization of civil contempt proceedings and, as I shall attempt to show, affords the opportunity to develop our common law in accordance with constitutional values.

[73] Upon proper analysis the distinction between coercive and punitive orders has something to do with the intent of an applicant or the court but much to do with the consequences of the order. It is the latter aspect to which any judicial officer who is required to consider whether an order of committal for contempt of court should be granted should pay careful attention.

[74] The following are, I would suggest, the identifying characteristics of a coercive order:

1. The sentence may be avoided by the respondent after its imposition by appropriate compliance with the terms of the original (breached) order ad factum praestandum together with any other terms of the committal order which call for compliance.59 Such avoidance may require purging a default, an apology or an undertaking to desist from future offensive conduct.60

2. Such an order is made for the benefit of the applicant in order to bring about compliance with the breached order previously made in his favour.

3. Such an order bears no relationship to the respondent’s degree of fault in breaching the original order or to the contumacy of the respondent thereafter or to the amount involved in the dispute between the parties.

4. Such an order is made primarily to ensure the effectiveness of the original order and only incidentally vindicates the authority of the court.

[75] By contrast a punitive order has the following distinguishing features:

1. The sentence may not be avoided by any action of the respondent after its imposition.

2. The sentence is related both to the seriousness of the default and the contumacy of the respondent.

3. The order is influenced by the need to assert the authority and dignity of the court and as an example for others.

4. The applicant gains nothing from the carrying out of the sentence.

[76] The differences are marked and important. They emphasise that a coercive order of imprisonment is one to which a respondent willingly (if reluctantly) and defiantly submits in order to frustrate the rights of another party. If he is ‘deprived’ of his liberty it is because he has, with knowledge of the order and the consequences of disobedience, elected to flout the order. Such an attitude has nothing to do with an onus of proof: the respondent would or would not submit or comply irrespective of the onus. Nor can one properly describe as ‘punishment’ that confinement to which a defendant of his own choice submits to serve his own ends. So understood, the circumstances of a coercive detention (and the procedure which is fair and appropriate to its imposition) stand at a vast remove from the case of enforced deprivation of liberty against which s 12 is primarily concerned to guard.61

[77] Of course there is a public dimension to both categories of order. But its emphasis is not the same, as I have pointed out. In any event, the public interest in having court orders which do not contain empty promises is a strong factor in favour of retaining the lighter onus.

[78] I consequently do not accept that a party in civil proceedings who exposes himself to the deprivation of freedom which flows from civil contempt and a consequent coercive order against himself deserves or needs an extension or adaptation of the common law to satisfy the imperatives of s 12(1). In the circumstances the existing procedures are entirely consonant with the constitutional values which underpin s 12.

[79] In reaching this conclusion I am very conscious of the strong body of judicial opinion which has voiced a conclusion contrary in its tenor to my own. I venture to suggest that there is, generally, an absence of consideration of the aspects which I have set out shortly above. An exception is Hicks v Feiock [1988] USSC 72; 485 US 624 (1988) in which the Supreme Court, despite differences in the result, was unanimous in finding that civil contempt proceedings are primarily coercive in nature and require proof on a balance of probabilities. I disagree with my colleague that differences in the constitutional backgrounds of the United States and South Africa have any significant bearing on the plain logic which it espouses and which is, in my view, of equal relevance to the procedural safeguards against abuse of the liberty provision in s 12(1) and to remedies for the breach of civil orders as our legal system knows them.62

[80] I would go further. The extension of the criminal standard of proof to civil contempt would have harmful consequences. In my experience the ordinary litigant (often an indigent woman63) finds it difficult enough under present procedures to pin down a party who is determined to avoid the consequences of a judgment. Absence of wilfulness and mala fides are frequently highly subjective and the respondent’s protestations often serve to carry the day, particularly as these are matters within his own ken and the applicant seldom has the means to pursue the enquiry with the necessary vigour. If the onus were to be increased to one beyond reasonable doubt the efficacy of the remedy (and with it the worth of a civil judgment) would be reduced, to the detriment of justice.64

[81] I should also add that, in principle, it would be wrong and unfair to align the frailty of the subject with the power of the State by requiring the former to discharge the criminal onus without comparable means to do so, unless such a conclusion cannot be avoided.

[82] That is indeed the only conclusion in punitive proceedings for contempt. For that reason the law does require development: a judicial officer who has found a litigant in civil proceedings to be in contempt and who forms the opinion that a punitive sentence may be warranted, should (whether or not he imposes a coercive sentence) refer the matter to the Director of Public Prosecutions with a view to prosecution in a criminal court. This would in my view be a desirable and justified development of the common law to ensure that those forms of the remedy of contempt of court (and the concomitant procedures) which are criminal in substance are tried in accordance with criminal standards, while leaving those that are truly civil in history, objectives and effects to be treated, as they always have been, according to civil standards.

[83] The existing reverse onus of proof renders the prospect of a finding against a respondent in contempt applications more likely than the application of a heavier onus would. Since such an onus has a tendency towards deprivation of the freedom of the subject it must be able to withstand constitutional scrutiny at the previously mentioned level of procedural fairness.

[84] Applying the test of examination of the nature of the deprivation and its purpose (De Lange v Smuts NO at para 143), I am satisfied that the existing procedure survives scrutiny. The following considerations militate against a development of the common law as Cameron JA proposes.

1. The existing reverse onus is a rational response to a proved breach of a court order which confers a right of enforcement on a party. It is proper and satisfies one’s sense of justice that the breaching party should be required to justify non-compliance.

2. As I have pointed out earlier, the subject-matter of the reverse onus often lies peculiarly within the knowledge of the respondent. In such circumstances equity favours the holder of the order and demands that the already defaulting respondent not only appear to play open cards but that he satisfy the court that he is now indeed doing so.

3. The proceedings are civil in substance and the coercive imprisonment which may follow is a civil remedy. As I have attempted to show above the deprivation of freedom which is threatened is of a singular nature not warranting special safeguards.

4. There is no evidence in the long history of the remedy that injustice has ever flowed from the application of the reverse onus in its present form.

5. Abolishing the existing reverse onus will simply render the remedy less effective without tending towards a corresponding advantage in the administration of justice.




J A HEHER

JUDGE OF APPEAL



CONCUR:
FARLAM JA

1 Constitution Chapter 9, s 181(1) – ‘The following institutions strengthen constitutional democracy in the Republic: … (e) The Auditor-General’. Section 181(2): ‘These institutions are independent, and subject only to the Constitution and the law, and they must be impartial and must exercise their powers and perform their functions without fear, favour or prejudice.’

2 CCII Systems (Pty) Ltd v Fakie NO 2003 (2) SA 325 (T).

3 In full (2003 (2) SA 325 (T) at 335-336):

1 The first respondent [the Auditor-General] is ordered to provide the applicant by no later than 40 Court days from the date of this order with the following records:

1.1 all draft versions of the report submitted to Parliament by the joint investigating team regarding the so-called Strategic Defence Packages for the procurement of armaments for the South African National Defence Force;

1.2 in respect of all audit files concerning the Strategic Defence Packages for the procurement of armaments for the SA National Defence Force from 1 January 1998 to 20 November 2001 dealing with:

1.2.1 the de-selection of the applicant as a supplier of the combat suite's information management system and the selection instead of the detexis diacerto combat suite databus;

1.2.2 the selection of the supplier of the systems management system, navigation distribution system and the integrated platform management system simulator;

1.2.3 the role of African Defence Systems (Pty) Ltd, a company controlled by Thomson-CSF of France (which later changed its name to Thales International), in the supply of the combat suite for the Corvettes and its conflict of interest by virtue of its involvement in the supply of the Corvettes at various different levels, namely as:

1.2.3.1 a member of the consortium constituting the prime contractor for the supply of Corvettes;

1.2.3.2 the supplier of the combat suite and at the same time being the combat suite integrator;

1.2.3.3 the supplier of various systems and subsystems for the combat suite, including the SMS and the combat management system; and

1.2.3.4 an associate company (ie a company in the Thomson-CSF group) of the supplier of the Detexis system;

1.2.4 the conflict of interest of Shamin Shaikh as:

1.2.4.1 the Department of Defence's Chief of Acquisitions and chairperson or member of various committees and boards involved in the assessment of the SDP; and

1.2.4.2 brother of Schabir Shaikh, who at all material times had an indirect interest in ADS;

1.3 all the documents and records in respect of which [he] has no objection in terms of chap 4 or s 12 of Act 2 of 2000; and

1.4 a list of all the documents and records in respect of which [he] objects in terms of the provisions of the aforesaid Act 2 of 2000, setting out clearly and concisely (a) a description of the document or record, (b) the basis for the objection, (c) an indication if the objection relates to the whole document or only to portions thereof and if so, (d) to which portions.’

2. The respondents [the Auditor-General, the Public Protector, the National Director of Public Prosecutions and the Minister of Defence] are ordered jointly and severally to pay the applicants’ costs of the application inclusive of the costs of two counsel.’

4 S v Beyers 1968 (3) SA 70 (A).

5 Melius de Villiers The Roman and Roman-Dutch Law of Injuries (1899) page 166: ‘Contempt of court … may be adequately defined as an injury committed against a person or body occupying a public judicial office, by which injury the dignity and respect which is due to such office or its authority in the administration of justice is intentionally violated.’ Cf Attorney-General v Crockett 1911 TPD 893 925-6 per Bristowe J: ‘Probably in the last resort all cases of contempt, whether consisting of disobedience to a decree of the Court or of the publication of matter tending to prejudice the hearing of a pending suit or of disrespectful conduct or insulting attacks, are to be referred to the necessity for protecting the fount of justice in maintaining the efficiency of the courts and enforcing the supremacy of the law.’

6 S v Mamabolo [2001] ZACC 17; 2001 (3) SA 409 (CC) para 14, per Kriegler J, on behalf of the court (where contempt of court in the form of scandalising the court was in issue).

7 Per Sachs J in Coetzee v Government of the Republic of South Africa [1995] ZACC 7; 1995 (4) SA 631 (CC) para 61, quoted and endorsed by the court in Mamabolo (above). In Coetzee, statutory procedures for committal of non-paying judgment debtors to prison for up to 90 days – which the statute classified as contempt of court – were held unconstitutional.

8 The Roman-Dutch law seems to offer no trace of private enforcement of criminal contempt remedies for disobedience of a civil order: see Melius de Villiers, note 5 above, pages 166-173. Cf Attorney-General v Crockett 1911 TPD 893 917 922 where it was held that, as regards criminal practice in matters of contempt, English procedure should be followed.

9 Although money judgments cannot ordinarily be enforced by contempt proceedings, ‘it is well established that maintenance orders are in a special category in which such relief is competent’: Bannatyne v Bannatyne [2002] ZACC 31; 2003 (2) SA 363 (CC) para 18.

10 Cape Times Ltd v Union Trades Directories (Pty) Ltd 1956 (1) SA 105 (N) 120D-E: ‘Generally speaking, punishment by way of fine or imprisonment for the civil contempt of an order made in civil proceedings is only imposed where it is inherent in the order made that compliance with it can be enforced only by means of such punishment.’

11 JRL Milton ‘Defining Contempt of Court’ (1968) 85 SALJ 387: ‘The concept of contempt of court is one which bristles with curiosities and anomalies. Of the various examples which may be chosen to illustrate this point perhaps the most striking is that of the classification of contempts of court into civil contempt (or contempt in procedure) and criminal contempt.’

12 Frankel Max Pollak Vinderine Inc v Menell Jack Hyman Rosenberg & Co Inc [1996] ZASCA 21; 1996 (3) SA 355 (A) 367H-I; Jayiya v Member of the Executive Council for Welfare, Eastern Cape 2004 (2) SA 602 (SCA) paras 18 and 19.

13 Consolidated Fish (Pty) Ltd v Zive 1968 (2) SA 517 (C) 524D, applied in Noel Lancaster Sands (Edms) Bpk v Theron 1974 (3) SA 688 (T) 691C.

14 Noel Lancaster Sands (Edms) Bpk v Theron 1974 (3) SA 688 (T) 692E-G per Botha J, rejecting the contrary view on this point expressed Consolidated Fish v Zive (above). This court referred to Botha J’s approach with seeming approval in Frankel Max Pollak Vinderine Inc v Menell Jack Hyman Rosenberg & Co Inc [1996] ZASCA 21; 1996 (3) SA 355 (A) 368C-D.

15 See the formulation in S v Beyers 1968 (3) SA 70 (A) at 76E and 76F-G and the definitions in Jonathan Burchell Principles of Criminal Law (3ed, 2005) page 945 (‘Contempt of court consists in unlawfully and intentionally violating the dignity, repute or authority of a judicial body, or interfering in the administration of justice in a matter pending before it’) and CR Snyman Strafreg (4ed, 1999) page 329 (‘Minagting van die hof is die wederregtelike en opsetlike (a) aantasting van die waardigheid, aansien of gesag van ‘n regterlike amptenaar in sy regterlike hoedanigheid, of van ‘n regsprekende liggaam, of (b) publikasie van inligting of kommentaar aangaande ‘n aanhangige regsgeding wat die strekking het om die uitstlag van die regsgeding te beïnvloed of om in te meng met die regsadministrasie in daardie regsgeding’).

17 See the remarks of Steyn CJ at 81A-B.

18 1968 (3) SA at 76E-G.

19 Drawing on the preceding analysis in Afrikaanse Pers-Publikasies (Edms) Bpk v Mbeki 1964 (4) SA 618 (A) 626.

20 1968 (3) SA at 80C-H (my translation throughout).

21 Noel Lancaster Sands (Edms) Bpk v Theron 1974 (3) SA 688 (T) 691A-D; Putco Ltd v TV & Radio Guarantee Co (Pty) Ltd 1985 (4) SA 809 (A) 836D-E; Frankel Max Pollak Vinderine Inc v Menell Jack Hyman Rosenberg & Co Inc 1996 (3) SA 355 (A) 367-8.

22 See the exposition by Pickering J in Uncedo Taxi Service Association v Maninjwa 1998 (3) SA 417 (E) 425G-426C and Frankel Max Pollak Vinderine Inc v Menell Jack Hyman Rosenberg & Co Inc [1996] ZASCA 21; 1996 (3) SA 355 (A) 367J.

23 1998 (3) SA 417 (E). The Zimbabwe Supreme Court followed Uncedo in a related setting in In re Chinamasa 2001 (2) SA 902 (ZSC) 922E-F and 924-5.

24 Constitution s 35(1): ‘Everyone who is arrested for allegedly committing an offence has the right – (a) to remain silent …’ Section 35(3): ‘Every accused person has a right to a fair trial, which includes the right – (a) to be informed of the charge with sufficient detail to answer it; … (f) to choose, and to be represented by, a legal practitioner, and to be informed of this right promptly; (g) to have a legal practitioner assigned to the accused person by the state and at state expense, if substantial injustice would otherwise result, and to be informed of this right promptly; (h) to be presumed innocent, to remain silent, and not to testify during the proceedings; …’.

27 2004 (4) SA 350 (T).

28 Case 19869/05 (T), judgment of 21 December 2005.

29 Judgment dated 3 November 2005, [2006] JOL 16722 (E).

30 As exemplified in S v Mamabolo (above).

31 See most recently S v Singo [2002] ZACC 10; 2002 (4) SA 858, 2002 (2) SACR 160 (CC), where Ngcobo J collates and analyses much of the preceding jurisprudence.

32 See for instance Osman v Attorney-General, Transvaal [1998] ZACC 14; 1998 (4) SA 1224, 1998 (2) SACR 493 (CC) paras 22-23 and S v Manamela [2000] ZACC 5; 2000 (3) SA 1 (CC), 2000 (1) SACR 414 (CC) paras 52-59.

33 As Froneman J pointed out in Burchell (para 15), the CC, though acknowledging that the right to individual freedom and security is not absolute, has only once sanctioned a legislative provision that places a legal onus on an individual deprived of freedom, namely in bail applications. There, an important consideration was the wording of the constitutional provision permitting deprivation of liberty on arrest (everyone arrested for an offence has the right ‘to be released from detention if the interests of justice permit, subject to reasonable conditions’: Bill of Rights s 35(1)(f); S v Dlamini [1999] ZACC 8; 1999 (4) SA 623 (CC) paras 6 and 38).

34 Osman v A-G Transvaal [1998] ZACC 14; 1998 (4) SA 1224 (CC) para 22.

35 Bill of Rights s 12(1)(b).

36 Bill of Rights s 12(1)(a).

37 See Bernstein v Bester NO [1996] ZACC 2; 1996 (2) SA 751 (CC) paras 145-146 (O’Regan J) and De Lange v Smuts NO [1998] ZACC 6; 1998 (3) SA 785 (CC) paras 22-25 (Ackermann J) (both dealing with the comparable provision under the interim Constitution).

38 In re Dormer (1891) 4 SAR 64 at 85 per Kotzé CJ (‘Contempts of court are certainly in some respects analogous to criminal offences, but they are a distinct species of offence, to which a special mode of summary procedure is applicable, and do not admit of the ordinary and usual forms and modes of criminal procedure’), applied in Afrikaanse Pers-Publikasies (Edms) Bpk v Mbeki 1964 (4) SA 618 (A) 626.

39 Witham v Holloway (1995) 131 ALR 401 (HC of A) 408, per Brennan, Deane, Toohey and Gaudron JJ.

40 Cf the approach of Sachs J in Coetzee v Government of the Republic of South Africa [1995] ZACC 7; 1995 (4) SA 631 (CC) para 43.

41 Compare Nel v le Roux [1996] ZACC 6; 1996 (3) SA 562 (CC) para 11, where it was held that a recalcitrant witness who is examined under s 205 of the Criminal Procedure Act 51 of 1977 under the procedure of s 189 is not an ‘accused person’ and therefore not entitled ‘directly’ to fair trial rights, but that such an examinee is ‘unquestionably entitled to procedural fairness’.

42 Uncedo 1998 (3) SA 417 (E) 427I-J.

43 In Hicks v Feiock [1988] USSC 72; 485 US 624 (1988), the question was the classification of relief imposed in a state court contempt proceeding as civil or criminal in nature, for purposes of applying the Due Process clause and other provisions of the United States constitution, since the protections in question do not apply when the relief is civil in nature. That constitutional setting differs considerably from ours, which in my view offers more varied possibilities in characterising the proceeding in question, and in determining the appropriateness of the applicable protections. Despite the apparently greater rigidity of the distinction in United States constitutional law, the majority of the court observed that in contempt cases, both civil and criminal relief ‘have aspects that can be seen as either remedial or punitive or both’ (485 US at 625), and that the ‘civil’ and ‘criminal’ labels of the law ‘have become increasingly blurred’ in state law codifications (485 US at 631). See too International Union, United Mineworkers of America v Bagwell [1994] USSC 43; 512 US 821 (1993) 826-830, dealing with ‘the somewhat elusive distinction between civil and criminal contempt fines’.

44 Nel v le Roux [1996] ZACC 6; 1996 (3) SA 562 (CC) (committal of recalcitrant witness under procedures in the Criminal Procedure Act 51 of 1977); De Lange v Smuts NO [1998] ZACC 6; 1998 (3) SA 785 (CC) (permitting committal of recalcitrant witnesses in sequestration proceedings by magistrate,).

45 The same applies to Bannatyne v Bannatyne [2002] ZACC 31; 2003 (2) SA 363 (CC), on which counsel for CCII relied: but there, in upholding the importance of contempt committal against maintenance defaulters as process in aid, the CC did not consider the standard of proof.

46 [1998] ZACC 6; 1998 (3) SA 785 (CC) para 147.

47 Witham v Holloway (1995) 131 ALR 401 (HC of A) 407-408.

48 (1992) 96 DLR (4th) 376 (SCC) 398. The vigorous dissent of l’Heureux-Dubé J turned on the special nature of injunctive relief in its specific setting in the Quebec Code of Civil Procedure: see pages 386f, 387g, 387-388 and especially 395g.

49 Re Bramblevale Ltd [1969] 3 All ER 1062 (CA) and the cases following it.

50 Hicks v Feiock [1988] USSC 72; 485 US 624 (1985) at 635.

51 International Union, United Mineworkers of America v Bagwell 512 US 821 (1993) 830.

52 Federation of Governing Bodies of South African Schools (Gauteng) v MEC for Education, Gauteng 2002 (1) SA 660 (T) 673D-E (Southwood & Basson JJ concurring).

53 Victoria Park Ratepayers (above) para 5.

54 Frank v Ohlsson’s Cape Breweries Ltd 1924 AD 289 at 294, per Innes CJ.

55 Peterson v Cuthbert & Co Ltd 1945 AD 420 at 428, per Watermeyer CJ.

56 Room Hire Co (Pty) Ltd v Jeppe Street Mansions (Pty) Ltd 1949 (3) SA 1155 (T) at 1162-1164, per Murray AJP.

57 [1984] ZASCA 51; 1984 (3) SA 623 (A) at 634-635, per Corbett JA.

58 There is long-standing authority that a misunderstanding as to the true meaning of an order negatives an inference that non-compliance is wilful: Botha v Dreyer (1880) 1 EDC 74.

59 The defendants in such proceedings carry ‘the keys of their prison in their own pockets’: Shillitani v United States [1966] USSC 110; 384 US 364 at 368 (1966).

60 In the words of Sachs J in Coetzee v Government of the Republic of South Africa; Matiso and others v Commanding Officer, Port Elizabeth Prison, and others [1995] ZACC 7; 1995 (4) SA 631 (CC) at para 43 fn 34, imprisonment of this kind is regarded in the jurisprudence of the United States as ‘a flexible remedial instrument for failure to fulfil an obligation’; cf Chinamora v Angwa Furnishers 1998 (2) SA 432 (ZSC) at 447F-G.

61 There is an intermediate kind of ‘contempt’ proceeding which does not fit readily into either category. Nel v Le Roux [1996] ZACC 6; 1996 (3) SA 562 (CC) and De Lange v Smuts NO [1998] ZACC 6; 1998 (3) SA 785 (CC) provide examples. Such cases do not involve civil contempt in the sense that we are concerned with. Committal is there the consequence of a refusal to comply with a statutory coercion imposed in the public interest. It is in this context that the dictum of O’Regan J which Cameron JA quotes at para 31 of his judgment has to be understood.

62 In Witham v Holloway 131 ALR 401 (HC of A) at 418; 183 CLR 525 at 547, McHugh J said,

. . . the chief reason for rejecting the United States approach of classifying proceedings for contempt according to their objective is that it leads to the practical problems to which I have referred.’

I have attempted to address the perceived practical objections in paras 69 to 71 above. The learned Judge seems to have encountered no obstacle in the constitutional peculiarities of that country. Nor indeed did the majority of the court express any such reservation in considering the America approach (at 406-7).

63 See Bannatyne v Bannatyne [2002] ZACC 31; 2003 (2) SA 363 (CC) para 27:

Systemic failures to enforce maintenance orders have a negative impact on the rule of law. The courts are there to ensure that the rights of all are protected. The Judiciary must endeavour to secure for vulnerable children and disempowered women their small but life-sustaining legal entitlements. If court orders are habitually evaded and defied with relative impunity, the justice system is discredited and the constitutional promise of human dignity and equality is seriously compromised for those most dependent on the law.’

64 In addition, as respondent’s counsel submitted, coercive execution by way of attachment and sale of property is not available to a civil litigant who has obtained an order ad factum praestandum. Contempt proceedings constitute the primary and, sometimes, the only method of enforcement of such orders. See also in this regard Witham v Holloway, fn 4 above, at 419 lines 10-22.