Ascendis Animal Health (Pty) Ltd v Merck Sharpe Dohme Corporation
Res judicata examined and Constitutional Court failed to achieve a majority decision which meant that the judgment of the high court was not overturned.
“Just or unjust? In my respectful view justice points to treating the Supreme Court of Appeal judgment refusing Ascendis’s revocation argument as conclusive of the patent’s validity. There can be no reason why res judicata does not apply, appropriately expanded. No doubt this prejudices Ascendis. Of greater consequence is the destabilisation of our patent litigation system, through creating a perilous and novel default. This would be to the prejudice of all. The courts in both the amendment proceedings and the interdict application were alive to this fact. The issue of validity between the parties is long-worn and settled. I would grant leave to appeal, but dismiss the application.” [para 136]
(CCT 212/18)  ZACC 41 (24 October 2019)
On appeal from the order of the High Court of South Africa, Gauteng Division, Pretoria:
1. Leave to appeal is granted.
2. The appeal is dismissed.
3. Each party is to pay its own costs.
Coram: Mogoeng CJ, Cameron J, Froneman J, Jafta J, Khampepe J, Ledwaba AJ, Madlanga J, Mhlantla J, Nicholls AJ and Theron J
Judgments: The Court:  to 
Khampepe J:  to 
Cameron J:  to 
Heard on: 9 May 2019
Decided on: 24 October 2019
“It is correct that, as a “non-examining country”, government does not examine a patent’s validity upon registration. The concerns this triggers suggest we should resist bifurcation, as developments in foreign jurisdictions warn. Imposing bifurcation, in the absence of any firm foundation, magnifies and does not diminish concern about the ease with which validity challenges can be initiated. Bifurcation unnecessarily points us toward an invidious choice: should we accept bifurcation, and with it allow successive validity challenges, or should we accept bifurcation with only single validity challenges, at the risk of detriment to the less-resourced? And if we wish to provide a salve for the exorbitant costs of patent litigation, then to allow multiple validity challenges is a remedy only the ultra-well-resourced could enjoy.
That need not be the way. An alternative lies in non-bifurcation, matched with equitable application of res judicata. As a result of over-zealous extension, barring a claim or defence may wreak “great hardship and even positive injustice to individuals”. Recently this Court overrode even res judicata proper, in exceptional circumstances, because justice demanded it. ” [paras 127 – 128]
‘Patents Act 57 of 1978 — section 61 — causes of action — res judicata — issue estoppel — bifurcated proceedings — amendment of pleas”
“The following explanatory note is provided to assist the media in reporting this case and is not binding on the Constitutional Court or any member of the Court.
On Thursday, 24 October 2019 at 10h00, the Constitutional Court handed down judgment in an application for leave to appeal against a judgment by the Commissioner of Patents which refused an application by Ascendis Animal Health (Pty) Limited to amend its plea and granted an application by Merck Sharpe Dohme Corporation, Merial Limited and Merial South Africa (Pty) Limited to amend their plea in patent infringement proceedings, regarding Merck and Merial’s patent for an anti-parasitic formulation.
In June 2011, Ascendis instituted patent revocation proceedings against Merck and Merial’s patent, claiming that the patent was invalid as it lacked novelty and inventiveness. In October 2011, while the revocation proceedings were ongoing, Merck and Merial instituted proceedings claiming damages against Ascendis, in which they claimed that Ascendis had infringed and was continuing to infringe Merck and Merial’s patent.
Merck and Merial stayed their infringement action pending the finalisation of Ascendis’s revocation proceedings. At the revocation hearing, Ascendis, without bringing a formal application, requested to proceed first with the lack of novelty claim and, only if this was unsuccessful, move on to argue the obviousness claim. Ascendis proceeded to attack the patent exclusively based on lack of novelty. The Commissioner found in favour of Ascendis and revoked the patent for its lack of novelty. The Supreme Court of Appeal overturned this finding on appeal and found in favour of Merck and Merial.
After the ruling by the Supreme Court of Appeal, Ascendis sought to amend its plea in the infringement action, to remove the defence of lack of novelty and to add an additional defence. This was that the patent was invalid because of inutility. Merck and Merial sought to amend their pleadings in the same proceedings, to submit that all challenges to their patent were res judicata (already decided by a court) since the validity of the patent had already been finally upheld by the Supreme Court of Appeal.
The Commissioner, in hearing the infringement action, held that in a revocation application, the grounds on which an applicant may rely to attack the validity of a patent are not separate causes of action. The Commissioner held that Ascendis had abandoned obviousness as a ground of invalidity and that it had not reserved the right to rely upon a separate ground of alleged invalidity. The validity of the patent was res judicata between the parties. Ascendis sought to appeal against this decision to the Supreme Court of Appeal, but leave to appeal was refused.
Before the Constitutional Court, Ascendis contended that the Commissioner extended the rule of res judicata beyond the boundaries within which it previously has been applied. It contended that the different subsections dealing with invalidity in section 61 of the Patents Act 57 of 1978 each require different facts to be proved. They therefore constitute separate, individual causes of action. Finally, Ascendis contended that the Commissioner erred in holding that the alleged procedural error by its counsel could preclude Ascendis from raising a different basis for invalidity than what had already been ruled upon, as a defence in separate proceedings.
Merck and Merial contended that the Commissioner correctly applied the res judicata principle and “the once and for all” rule. The defence Ascendis sought to rely on constitutes the same cause of action, namely that the patent was invalid, which the Supreme Court of Appeal had already conclusively dismissed. Further, even if the matter is not res judicata, litigating the validity of a patent in the way Ascendis seeks to do would constitute an abuse of process.
In an evenly split decision by the Constitutional Court, the first judgment penned by Khampepe J (with Froneman J, Ledwaba AJ, Nicholls AJ and Theron J concurring) would have granted leave to appeal and upheld the appeal. The first judgment held that the Patents Act endorses two separate proceedings with their own rules and remedies. It would be improper for the Court to conclude that the findings in the revocation proceedings have a final effect on the infringement action in regard to causes of action which have not been decided in revocation proceedings. Further, the first judgment held that each of the grounds of revocation found in section 61 of the Patents Act require different material facts to be proved. Accordingly, they constitute separate, distinct and independent causes of action that can be raised even after one has been dismissed.
Hence, according to the first judgment, absent any final and definitive decision made by a competent court, Ascendis’s claim of inutility was not a cause of action that could be said to be res judicata.
Furthermore, according to the first judgment, Ascendis’s claim of obviousness could also not be said to have been decided on its merits in the revocation proceedings. However, Ascendis’s seeming procedural error in abandoning this claim should not exclude it from the parallel proceedings. In addition, the first judgment found that in the revocation proceedings before the Commissioner, the parties pleaded the obviousness claim which placed the Commissioner under a duty to pronounce on the status of that claim.
In respect of Merck and Merial’s contention that Ascendis’s conduct in the litigation proceedings constituted abuse of process, the first judgment could not make a conclusive finding of abuse of process by Ascendis, as Merck and Merial did not fully plead and argue abuse of process on the facts.
The second judgment penned by Cameron J (with Mogoeng CJ, Jafta J, Madlanga J and Mhlantla J concurring) would have granted leave to appeal but dismissed the appeal. The second judgment found that the Commissioner correctly discerned the “real issue” between the parties. This was whether the judgment of the Supreme Court of Appeal, which rebuffed Ascendis’s attempt to revoke Merck’s patent, had conclusively determined the validity of the patent. This was the issue before the Constitutional Court too. That issue had already been conclusively determined between the parties when the Supreme Court of Appeal ruled that the patent was valid. It was accordingly res judicata. The second judgment further held that the default position postulated by the first judgment would result in incoherence and affront the long-held judicial caution against piecemeal litigation. The default position should rather be that a previously unsuccessful revocation applicant is generally, though not invariably precluded from raising the validity of the patent as a defence to a subsequent damages claim. If this is unfair, then a court should allow the latter defence to be raised, but only in exceptional circumstances.
In the result, five Judges of the Constitutional Court held that the appeal should be upheld as set out in the judgment of Khampepe J and five Judges held that the appeal should be dismissed as set out in the judgment of Cameron J. There is thus no majority decision of the Constitutional Court. Accordingly, the judgment and order of the Commissioner stands.”
Quotations from judgment
Note: Footnotes omitted and emphasis added
 The first judgment in this matter was written by Khampepe J and was concurred in by Froneman J, Ledwaba AJ, Nicholls AJ and Theron J. The effect of the first judgment is that five members of the Court held that the application for leave to appeal should be granted and that the appeal should be upheld.
 The second judgment in this matter was written by Cameron J and was concurred in by Mogoeng CJ, Jafta J, Madlanga J and Mhlantla J. The effect of the second judgment is that five members of the Court held that leave to appeal should be granted and the appeal should be dismissed.
 There is thus no majority decision of this Court. The result is that the judgment and order of the High Court of South Africa, Gauteng Division, Pretoria stands.
 Considerations of fairness dictate that each party is ordered to pay its own costs in light of there not being a majority decision in this matter.
KHAMPEPE J (Froneman J, Ledwaba AJ, Nicholls AJ and Theron J concurring):
Nature of proceedings
 This is an application for leave to appeal against the judgment and order of van der Westhuizen J, sitting as the Commissioner of Patents in the High Court of South Africa, Gauteng Division, Pretoria (the High Court). The judgment of the High Court concerns two amendment applications made in patent infringement proceedings instituted by the respondents, as plaintiffs, against the applicant, as defendant.
 The High Court refused to grant leave to appeal against the dismissal of the applicant’s application to amend its pleadings before the High Court. The Supreme Court of Appeal dismissed the application for leave to appeal. The applicant now knocks on the doors of this Court seeking to overturn the judgment and order of the High Court.
Background facts and litigation history
 In June 2011, the applicant, Ascendis Animal Health (Pty) Limited, filed an application for the revocation of the South African Patent 1998/10975 (the 1998 patent) in terms of section 61(1) of the Patents Act (the Act) on the basis that the invention was not patentable in terms of section 25 of the Act (revocation proceedings). The applicant pleaded that
- the 1998 patent was not new in terms of the definition in section 25(5) and (6) (novelty); and
- that it lacked an inventive step in terms of the definition of section 25(10) (obviousness).
 During the exchange of pleadings and relevant documents between the parties in the revocation proceedings, the first and second respondents, Merck Sharp Dohme Corporation and Merial Limited, instituted proceedings against the applicant for infringing the 1998 patent (the infringement action). The respondents contended that since August 2011 the applicant had made, sold, used, offered for sale and / or imported an anti-parasitic formulation under the trademark entitled Ivermax LA Platinum and, due to these actions, had infringed and was continuing to infringe claims 1 to 7, 18 to 23 and 29 of the 1998 patent.
 The revocation proceedings turned on the validity of the 1998 patent in light of the disclosure in a separate patent registered in 1992; whereas the infringement action turned on the unauthorised use, and transgression, of the 1998 patent. However, underpinning both proceedings, albeit in different contexts and with different forms of relief sought, was the question of the validity of the 1998 patent. As evidence had already been led (by way of affidavits) in the revocation proceedings, the parties agreed to stay the infringement proceedings and proceed with the revocation proceedings to finality.
 In August 2012, the applicant sent a letter to the respondents advising them that it intended to first argue the novelty claim separately from the obviousness claim and, should this argument fail, the applicant would then apply to refer the obviousness claim for oral hearing where the applicant would lead evidence on this question. The logic behind this approach was that should the Commissioner of Patents find that the patent lacked novelty (on paper), then it would not need to hear oral evidence as the patent would have already been found to be invalid and the obviousness claim would not take the matter any further. The respondents opposed this and alleged that a referral of this nature was not in terms of a legitimate procedure or even appropriate. Subsequently, there was no agreement between the parties on the purported separation of the issues.
 Despite the objection raised by the respondents, the applicant persisted and asserted in its heads of argument that should the Commissioner of Patents find that the invention is not novel, then an application would be moved for the matter to be referred for the hearing of oral evidence on the second challenge to the patent – the ground of obviousness.
 The Commissioner of Patents, Teffo J, revoked the 1998 patent for lack of novelty, in light of the disclosure of the 1992 patent. She held that claims 1 to 29 of the 1998 patent failed to satisfy the requirements of novelty as set out in section 25(5) and (6) of the Act and therefore fell to be revoked. Importantly, her finding was limited to novelty and did not address the obviousness claim. No reasons were advanced explicating the status of the obviousness claim or why it was not addressed.
 Aggrieved by the decision, the respondents approached the Supreme Court of Appeal on the novelty claim. The applicant made no attempt to appeal the failure to pronounce on the obviousness claim. It also did not advance any argument that the question of obviousness had been held over (by agreement or by order of court) for determination at a later date, pending the outcome of the novelty claim.
Supreme Court of Appeal
 The Supreme Court of Appeal found, without mentioning the obviousness claim, that there was no merit in the attack on the patent’s validity under the novelty claim and that the revocation application should have been dismissed with costs. The Supreme Court of Appeal ordered that each of the claims of the 1998 patent be certified as being valid in terms of section 74 of the Act.
 Subsequently, in light of the Supreme Court of Appeal’s decision, the applicant filed a notice of intention to amend its plea in the infringement action. The applicant, through this amendment, sought to remove the novelty defence, keep the obviousness defence and introduce a new defence, inutility, in terms of section 61(1)(d) of the Act.
 The respondents opposed the applicant’s amendment application and filed an application to amend their replication to the plea so as to plead res judicata in light of the Supreme Court of Appeal’s judgment. Furthermore, the respondents, believing that the applicant was frustrating the proceedings, as the 1998 patent was set to expire in 2018, applied for an interim interdict prohibiting the applicant from further infringing the 1998 patent pending the outcome of the infringement action or until the expiry of this patent.
Interim interdict proceedings
 Louw J held that it is trite that generally, piecemeal litigation is to be avoided. He found that the applicant was attempting to re-try the matter by amending its pleadings in the infringement proceedings and this amounted to piecemeal litigation. Louw J held that the applicant was attempting to do what the court in Chiron Corporation warned against, which is allowing “a party with more than one cause of action or more than one ground of defence to advance one of them and then keep the others in reserve for a rainy day.”
 As obiter (in passing), Louw J further remarked that if the applicant had in its heads of argument informed the Supreme Court of Appeal that it still intended to proceed with the obviousness claim, the Supreme Court of Appeal would have most likely refused to entertain the appeal as it is strongly opposed to hearing piecemeal appeals. Unfortunately, the applicant did not inform or raise this with the Supreme Court of Appeal.
 Accordingly, he held that the applicant’s attempt to rely on invalidity in the infringement action based on obviousness amounted to piecemeal litigation as the Supreme Court of Appeal had finally decided the issue of the 1998 patent’s validity. He further found that the respondents satisfied the requirements of an interim interdict, and therefore issued an interim order interdicting the applicant from infringing the patent, pending the final determination of the main application, and reserving the question of costs for determination in the action.
Amendment applications in the High Court
 Before van der Westhuizen J, the parties sought to amend their pleadings in the infringement action respectively as described above.
 The applicant’s counsel contended that each ground for revocation of a patent, as provided in the Act, is a separate and distinct cause of action, and therefore by advancing a new and further ground of invalidity, the applicant was advancing a new cause of action that may be introduced by way of an amendment.
 Van der Westhuizen J found no merit in this submission. He found that the cause of action in an application for revocation of a patent is that of the invalidity of the patent. The validity of the patent may be challenged and attacked on various grounds listed in the Act. A party may therefore elect to rely on one or more of these listed grounds, but this does not mean that they can rely on them in a piecemeal fashion.
 To reinforce his conclusion, van der Westhuizen J highlighted that the Supreme Court of Appeal, per Ponnan JA, certified that all the claims of the 1998 patent are valid in accordance with section 74 of the Act – an indication that the issue of the validity of the 1998 patent is res judicata between the parties.
 In finding that the matter is res judicata, the Commissioner discussed the requirements for a successful reliance on the doctrine of res judicata, namely: (i) same parties (the parties are the same); (ii) the same cause of action (the invalidity of the 1998 patent); (iii) the same relief (the revocation of the 1998 patent); and (iv) a final judgment (the Supreme Court of Appeal judgment). The Commissioner held that the requirements were, indeed, fulfilled.
 It was held that since the cause of action is the invalidity of the patent, the matter was thus res judicata. Accordingly, the High Court dismissed the application for the amendment of the applicant’s plea and upheld the application for the amendment of the respondent’s plea.
 The Commissioner refused leave to appeal. An application to the Supreme Court of Appeal for leave to appeal was also refused. The application before us is for leave to appeal against the judgment refusing the applicant’s application to amend its pleadings in the infringement action.
 The issues for determination as I see them are:
- 1. Whether this matter engages this Court’s jurisdiction.
- 2. Whether all of the subsections in section 61 of the Act constitute a single cause of action or whether each subsection constitutes a separate and independent cause of action.
- 3. Whether the defence raised in the infringement action is res judicata.
Jurisdiction and leave to appeal
 Section 167(3)(b) of the Constitution sets out the grounds on which this Court’s jurisdiction may be engaged. This section confers jurisdiction on this Court to hear matters that raise constitutional issues or that raise an arguable point of law of general public importance which ought to be considered by this Court. In order for this Court to adjudicate and pronounce on a matter, a litigant has the onus of proving that the matter engages either of the two grounds or both.
Access to courts
 The first contention that the applicant advances is that this matter implicates the constitutional right of access to courts in terms of section 34 of the Constitution as it concerns the alleged unconstitutional deprivation of the applicant’s right to defend itself in a patent infringement action. Specifically, the applicant contends that its right to a fair hearing, including a hearing on the merits of the dispute, is being unjustifiably limited.
 The right of access to courts is a crucial provision in ensuring a constitutional democracy, and is an indispensable constituent of the rule of law, a founding value of the Constitution. It can thus not be gainsaid that an integral principle of the rule of law is the power given to people to challenge and demur against the legality of any conduct or law. It is thus evident that this right is instrumental to the achievement and enjoyment of the other rights and protections provided in the Constitution.
 Of importance to this case is the component of the provision that confers a right to the parties to have their disputes decided before a court. This element of the right directly grants the right to access to courts, and guarantees that litigants may bring their case before a court. Not only does this provision guarantee a litigant access to courts, it also confers on a litigant the right to a fair hearing.
 It is well-established that res judicata implicates the rights contained in section 34. However, the High Court, as will become evident later, extended the application of res judicata and as a result, adversely affected the right by denying the applicant an opportunity to raise a defence, which potentially taints the fairness element of the hearing. This prima facie extension of res judicata interferes with the applicant’s constitutional right to have the merits of the separate, undecided causes of action heard in court and thus gives this Court jurisdiction to decide the matter.
Arguable point of law
 There are three primary arguments that the applicant belabours in this regard. First, the applicant contends that the High Court extended the principle of res judicata drastically from the common law position. Second, the applicant submits that this matter raises a question of whether it is permissible that an alleged procedural misstep, which leads to a failure to formally separate the various grounds of invalidity in one matter, should deprive the applicant from relying on the undecided ground of invalidity as a defence in separate, but concurrent proceedings as a result of the application of res judicata. Third, the applicant submits that this matter involves the interpretation of section 61 of the Act and a determination of whether the grounds housed under this section constitute a single cause of action or multiple, distinct causes of action.
 The High Court found that the case was res judicata and that the grounds listed in section 61 are a single cause of action. These findings may have significant, adverse ramifications for future potential litigants who might be interested in bringing a revocation case. As will be seen later, there are also reasonable prospects of success on the question of whether the High Court extended and misapplied the principle of res judicata. Therefore, this matter begets an arguable point of law.
 The respondents submit that the infringement action only concerns a claim for damages, and that the applicant is attempting to avoid financial liability by not paying the damages. In essence, the pith of the respondents’ submissions is that it is not in the interests of justice for this Court to adjudicate this matter as the matter does not transcend the narrow interests of the parties, especially that of the applicant.
 While there is merit to this argument, it unduly simplifies the issue before this Court. This Court is not tasked with determining the outcome of the infringement action or the damages thereof. The applicant knocks on our door, asking that we answer the following question: in infringement proceedings, is the defence of invalidity of a patent (on the grounds of obviousness and inutility) res judicata, where that same defendant lost revocation proceedings based on the ground of novelty? To answer this question, this Court has to ascertain whether the different grounds housed under section 61 of the Act constitute a single cause of action or not. This is an interpretational exercise of provisions contained in legislation, which surely triggers our jurisdiction on its own.
 These questions undoubtedly go beyond the narrow interests of the parties. It is not disputed that the applicant approached this Court with these questions to safeguard its own interests. However, that does not denude the legal questions of their objective public importance. Any order given by this Court will have a wider impact in the arena of patent litigation. Accordingly, the infringement action and the subsequent order arising from it may bind only the parties; however, there is a general public interest in this decision that goes beyond the narrow interests of the parties because it requires the interpretation of section 61 of the Act.
 In light of the above, this matter raises a constitutional issue in relation to access to courts and various arguable points of law of general public importance. The arguments raised by the applicant have strong prospects of success. Leave to appeal is accordingly granted.
 The issue before the Court is whether the High Court’s finding that the validity of the 1998 patent was res judicata prevented the applicant’s amendment of its plea in the infringement proceedings. In oral argument before this Court, the respondents’ counsel contended that, even if the High Court erred in concluding that the validity of the 1998 patent was res judicata, the applicant’s attempts to raise the obviousness and inutility defences during the infringement proceedings amounted to an abuse of process.
 Although there may be compelling arguments in this case for the application of the doctrine of abuse of process, I am of the view that it is not properly before this Court. In that regard, I note that the issue of abuse of process was not raised in the High Court and is raised only, for the first time, in oral proceedings before this Court. Although some of the policy considerations implicit in the principle of res judicata intersect with those that motivate abuse of process concerns, the two concepts remain analytically distinct. In this regard, the primary concern, in my view, of res judicata is to ensure judicial certainty, and prevent contradictory conclusions on the same causes of action. Abuse of process concerns are motivated by the need to protect the “the integrity of the adjudicative functions of courts”, doing so ensures that procedures permitted by the rules of the Court are not used for a purpose extraneous to the truth-seeking objective inherent to the judicial process.
 In Janse van Rensburg, Heher JA aptly stated that a party who relies on abuse of process must pertinently plead the abuse, substantiated by facts and lay a foundation in fact, which could enable the opposing parties to deal with that reliance. Crucial here is the rationale that the doctrine of abuse of process must only be applied and imposed when the facts of the matter support the conclusion that the impugned conduct amounts to an abuse. Should this doctrine be applied where the facts do not point to an abuse of process, a party may be barred from advancing a bona fide challenge or defence in litigation. Was abuse of process pertinently pleaded in this matter? Simply not. The respondents did not fully plead and argue that the conduct of the applicant amounts to an abuse of process on the facts. For this reason, I am not convinced that this Court should make a conclusive finding, absent the supporting facts, that the applicant abused the process.
 As has been reiterated on numerous occasions, this Court’s reluctance to determine matters as a court of first instance is borne out of the practical reality that “this Court functions better when it is assisted by a well-reasoned judgment (or judgments) on the point in issue”. In this case, the question before us is whether the application of res judicata prevents the applicants from amending their plea to include certain defences in the infringement proceedings. This question has been fully ventilated in the High Court and in both the written and oral submissions before this Court. Accordingly, that will be the focus of my judgment.
Scheme of the Act
 In determining the res judicata point, it will be instructive to look at the scheme of the Act.
 According to the Act, any person may apply for the revocation of a patent, in the prescribed manner, at any time. All proceedings dealing with the revocation of a patent are dealt with comprehensively in chapter X of the Act. The chapter, through section 61(1), provides an exhaustive list of the grounds for the revocation of a patent. The Regulations to the Act then set out the applicable procedure to be followed when an application for revocation is launched. The Regulations make it clear that, unless directed by the Commissioner, all evidence relevant to the proceedings must be reduced to an affidavit. The proceedings therefore take on the form of a “full-dress debate” where the onus to prove the grounds for the revocation of the patent rests with the applicant.
 The Act, through chapter XI, thereafter guides the hand of patentees, this time, in the prosecution of infringement proceedings. The process followed in infringement proceedings takes the form of an action and, unlike revocation proceedings, is destined for oral evidence from its inception. As the patentee is the only one entitled to institute an action for the infringement of his or her patent, the burden of proof rests upon him or her. Furthermore, it is commonly understood that counterclaims for infringement or revocation are not mandatory counterclaims in either type of patent litigation.
 It is evident that the Act envisaged these two different proceedings to ultimately have two different consequences, with different forms of relief. It would be remiss if I did not illustrate the different effects of the two proceedings, which further support the argument that the Act envisages the two proceedings as separate. Revocation removes the patent from the Patents Register, a remedy in rem (against a thing). Invalidity defences in the infringement action, on the other hand do not remove the patent from the Patents Registrar, but render a patent unenforceable against the party who successfully challenges its validity but may still be enforced against third parties. Revocation is governed by section 61 and infringement by section 65 of the Act. The co-existence of dual proceedings in this context is not unusual – it is common cause between the parties that it is the norm in patent litigation to have parallel proceedings.
 When the Legislature has endorsed two separate proceedings with their own rules and remedies then it would be wrong for this Court to conclude that the findings in the revocation proceedings have a final effect on the infringement action for causes of action which have not actually been adjudicated upon in revocation proceedings. It might lead to the dual proceedings being nugatory and of no effect because a finding on the revocation proceedings could be carried over to the infringement action.
 I interpose to make a few salient points.
- First, the remedies sought in the revocation proceedings are distinct and materially different from the infringement proceedings. More specifically, the remedy sought in revocation proceedings is in rem whereas the remedy in infringement proceedings is in personam (against that person).
- Second, a party may institute revocation proceedings any time; whereas, no proceedings for infringement of a patent may be instituted within a period of nine months from the date of its sealing unless, on good cause shown, the Commissioner of Patents may direct otherwise.
- Third, infringement proceedings are inherently action proceedings; whereas revocation proceedings are predominantly application proceedings.
It is perspicuous from the reading of the Act and the language used, that the Legislature intended for there to be two separate proceedings. We should be slow to undermine the objective of the Legislature unless it leads to a constitutionally untenable outcome or an intolerable absurdity.
Do the grounds under section 61 constitute a single cause of action?
 The High Court answered this question in the affirmative when it held that section 61 constitutes a single cause of action which it found was the invalidity of the patent. For the following reasons I find that this is not correct.
 The leading case on the definition of “cause of action” is McKenzie. In McKenzie, Maasdorp JA approved the definition set out in the English case of Cook v Gill and defined “cause of action” as:
“[E]very fact which it would be necessary for the plaintiff to prove, if traversed, in order to support his right to judgment of the Court. It does not comprise every piece of evidence which is necessary to prove each fact, but every fact which is necessary to be proved.”
 Over a decade after McKenzie, the court in Abrahamse & Sons explicated this phrase as follows:
“The proper legal meaning of the expression ‘cause of action’ is the entire set of facts which give rise to an enforceable claim and includes every fact which is material to be proved to entitle a plaintiff to succeed in his claim. It includes all that a plaintiff must set out in his declaration in order to disclose a cause of action.”
 Of significance is the fundamental distinction that the court in McKenzie drew between the material facts which the applicant is required to prove in order to establish his or her case (facta probanda), and the evidence which the plaintiff must advance in order to establish those material facts (facta probantia). What this amounts to is that the “cause of action” in a particular case consists of the facta probanda as opposed to the facta probantia. In simple terms, the Court in McKenzie endorses the view that the central basic facts of the case are not to be confused with the various items of evidence required to prove those facts.
 More recently, Corbett JA cited the above cases with approval. To this end, cause of action means every fact that needs to be proved in order to support a litigant’s right to a judgment. It does not comprise every piece of evidence which is necessary to prove each fact, but every fact which is necessary to be proved.
 In light of the definition of cause of action and the distinction between facta probanda and facta probantia, I do not think that the grounds of revocation constitute a single cause of action. The opposite is true. Each of the grounds of revocation as set out in section 61 of the Act constitute separate, distinct and independent causes of action because the facta probanda that need to be proven for each ground are different.
Although the legal conclusion that results from claims of either novelty, obviousness or inutility may be the same (in other words, the finding of a patent’s invalidity) it does not follow that they all represent a single cause of action. The facts required to prove a claim of novelty, inutility and obviousness are markedly different as the elements constituting each ground are different.
 On a simple level, this is obviously so. In layman’s terms, novelty requires that something is new or different than what existed before (the prior art). Obviousness accepts that it is new, but it is an obvious step that anyone who is versed in the prior art could make because while it is a step, it is not an inventive step. Inutility is the contention that the blueprint of the patent itself does not lead to the claimed usefulness or result.
 In relation to the ground of novelty, the case of Merck Sharp Dohme Group is instructive. For a court to find that a patent lacks novelty, there are two definite inquiries that are imperative for a court to undertake, namely:
- a) Does a given occurrence or disclosure or the like (an alleged anticipation) in fact qualify as an effective anticipation?
- b) If it does, is the alleged anticipation an enabling disclosure of such a measure as to destroy the novelty of the alleged invention claimed in a given claim?
 These inquiries have been applied and cemented in case law. For instance, in Synthon BV, Lord Hoffmann held that when deciding on the ground of novelty:
“It is very important to keep in mind that disclosure and enablement are distinct concepts, each of which has to be satisfied and each of which has its own rules . . . in deciding whether there has been anticipation, there is a serious risk of confusion if the two requirements are not kept distinct.”
 Although obviousness is the most commonly raised ground when a party wishes to question the validity of a patent, it is difficult to construct and contrive an acceptable general principle setting out how a court reaches a decision on the lack of invention because all of this is fact-sensitive. Obviousness, unlike the other grounds, is not decided in terms of theoretical definitions and abstract legal principles proffered by courts – frequently, the facts and circumstances surrounding the invention are pertinent.
 In Roman Roller CC, the earliest case dealing with obviousness under the Act, Corbett CJ held that—
“As section 25(1) indicates, an invention is deemed to involve an inventive step if it is not obvious to a person skilled in the art, having regard to the state of the art at the relevant time. Conversely, if the invention is obvious to such a person, then the invention is deemed not to involve an inventive step and to be invalid on the ground of obviousness.
In order to apply these provisions to a particular case it is necessary first to determine:
(i) what the art is to which the invention relates,
(ii) what the state of this art was at the relevant time and
(iii) who is to be regarded as a ‘person skilled in the art’.”
 After determining the three factors laid out by Corbett CJ, the subsequent step is to question whether the patent involves an inventive step. What this requires is asking
“whether, in the light of the matter available to the public at the priority date, the step forward taken by the invention would have been obvious to the skilled addressee”.
 Of paramount importance is that for obviousness to arise as an issue, it means that the patent must have survived a novelty claim. In other words, only a patent that has been found to be novel can be attacked on the basis of obviousness.
 The ground of inutility provides that a patent may be revoked on the ground that the invention as illustrated or exemplified in the complete specification concerned cannot be performed or does not lead to the results and advantages set out in the complete specification. This boils down to whether the patent is useful as illustrated or exemplified. The term “useful” in patent law bears the “special meaning of effective to produce the result aimed at” or promised.
Murphy J in Merial found that—
“inutility is assessed with reference to the body of the specification, in particular as illustrated in drawings or as exemplified by way of example. Inutility is a question of fact in respect of which expert opinion evidence is admissible.”
 It is thus evident that novelty, obviousness and inutility require different facta probanda and therefore they are different causes of action. There is no merit to the respondents’ assertion, supported by the High Court, that the cause of action is the general invalidity of the patent. The different grounds may lead to a finding of invalidity but the mere fact that the result or legal conclusion is identical does not mean that those grounds are a single cause of action, grouped together under section 61. It is not a denouement supported by any law. The grounds I have discussed above illustrate that each ground requires a different set of facts to sustain the claim and relief sought.
 Janse van Rensburg buttresses this conclusion. The Supreme Court of Appeal held that related claims under different sections of the same Act, in that case the Insolvency Act, constitute separate causes of action, even where the consequences of applying the different sections would be the same. The Court also held that the sections showed no commonality of cause of action and were designed to provide remedies for different circumstances, notwithstanding the common outcome provided for in each.
 Analogously, the different sub-sections in the Act, each concerned with the invalidity of the patent and notwithstanding the common outcome, do not provide for a common cause of action. Furthermore, the different sections of the Act that govern the revocation proceedings and the infringement proceedings may focus on the invalidity of a patent (as a conclusion) but that does not elevate the invalidity of a patent into a cause of action.
 The Court in Bisonboard held that it is a well-established principle of our law that there is a distinction between causes of action on the one hand and legal proceedings on the other. The result of this distinction is that it is not the legal proceedings that will be terminated by res judicata, but the individual causes of action that have been decided. The High Court appears to have found that the proceedings were res judicata on the basis that the legal proceedings have a similar outcome. This is clearly wrong. The applicant relied on different causes of action and on the strength of that, the matter could not have been res judicata.
 Accordingly, the grounds housed under section 61 are separate and distinct causes of action. This leads to the question of whether the High Court was correct in finding that the matter was res judicata.
Is the defence in the infringement action res judicata?
 As a result of finding that the cause of action is the invalidity of the patent, the High Court concluded that the defence raised by the applicant in the infringement action was res judicata. Due to the incorrect findings in relation to the cause of action and the wrongful description of the relief sought as being the same in both proceedings, the subsequent finding by the High Court that the defence was res judicata is tainted and incorrect.
 Res judicata strictly means that a matter has already been decided by a competent court on the same cause of action and for the same relief between the same parties.
In Evins, Corbett JA stated that:
“Closely allied to the ‘once and for all’ rule is the principle of res judicata which establishes that, where a final judgment has been given in a matter by a competent court, then subsequent litigation between same parties, or their privies, in regard to the same subject-matter and based upon the same cause of action is not permissible and, if attempted by one of them, can be met by the exceptio rei judicatae vel litis finitae. The object of this principle is to prevent the repetition of lawsuits, the harassment of a defendant by a multiplicity of actions and the possibility of conflicting decisions.”
 In essence, the crux of res judicata is that where a cause of action has been litigated to finality between the same parties on a previous occasion, a subsequent attempt to litigate the same cause of action by one party against the other party should not be allowed. The underlying rationale for this principle is to ensure certainty on matters that have already been decided, promote finality and prevent the abuse of court processes.
 The requirements of res judicata, although trite, can be summed up as follows:
- (i) there must be a previous judgment by a competent court
- (ii) between the same parties
- (iii) based on the same cause of action, and
- (iv) with respect to the same subject-matter, or thing.
In a Lesotho case, Masara, the Court of Appeal stated that the defence of res judicata requires that a party must establish that the present case and the previous case are based on the same set of facts that have been finalised by a competent court or tribunal by the same parties on the merits of the same cause of action.
 In Molaudzi, this Court relied on the Canadian case of Amtim Capital Inc, which provided that the purpose of res judicata is to balance the public interest in the finality of litigation with the public interest of ensuring a just result on the merits. Therefore, this principle should not be used mechanically.
 I now turn to whether relying on the new grounds of invalidity, already held to be separate causes of action in the infringement proceeding, are in fact res judicata.
I will start with the inutility ground.
 The claim of inutility was neither pleaded nor argued in the Teffo J judgment and, therefore, no decision was made on the merits of this ground of invalidity. Absent a final and decisive decision made by a competent court or tribunal, this cause of action cannot be said to be res judicata. As I have stated, each of the three grounds in question, inutility, lack of novelty and obviousness, require different facts to be proven and thus constitute different causes of action.
Thus, while the litigation may be between the same parties, the remaining elements of res judicata namely:
- (i) the same cause of action and
- (ii) a final decision on the merits by a competent court have not been satisfied.
This ground is not res judicata and the High Court erred in finding that it is.
 It is clear that obviousness was also never decided on the merits in the Teffo J judgment. However, because it was pleaded by the applicant (and arguably subsequently abandoned) in those proceedings, it requires special attention here. While obviousness has not been heard on the merits because of the supposed procedural errors committed by the applicant, there may still be reason to exclude it from the parallel proceedings.
However, for the following reasons, I am not convinced that the obviousness claim should be excluded.
 As discussed above, it is important to separate the two proceedings that normally take place in patent litigations. The revocation proceedings and infringement action are two separate and independent proceedings. This is clear from the Act. The two proceedings may be linked and due to the nature of this type of litigation, most often are – but they remain fundamentally independent of each other.
There are two forms of relief sought in terms of both proceedings: if a party successfully challenges the validity of a patent under any one of the grounds, the patent is removed from the Patents Register; and if a party successfully defends itself in an infringement action by undermining the validity of the patent, the patent remains on the Patents Register but it will not be enforceable against the defendant in the action.
 The respondents strongly rely on the alleged procedural error committed by the applicant in failing to separate the different claims in the revocation proceedings. In my view, the procedural error has been overstated.
- First, the revocation proceedings were application proceedings and rule 33(4) applies to action proceedings, not application proceedings. An argument could have been advanced that a special case in terms of rule 33(1) would have been appropriate in these circumstances. However, no argument of this nature was advanced in the High Court or before this Court. Rule 6(5)(d)(iii) provides a respondent in application proceedings with the means of raising a point of law, but not an applicant.
 Despite this, Teffo J in the revocation proceedings had a duty to provide reasons why the case was decided in the way that it was. She had a responsibility to the parties and the public at large to furnish reasons for her decision and to pronounce on the status of the obviousness claim. This is especially so when the parties have pleaded this claim.
As stated in Stuttafords Stores—
“This Court has stated that furnishing reasons in a judgment— explains to the parties, and to the public at large which has an interest in courts being open and transparent, why a case is decided as it is. It is a discipline which curbs arbitrary judicial decisions.”
 In any event, I am not convinced by the respondents’ argument that an alleged procedural error that occurs in one of the proceedings (for example, the revocation proceedings) should automatically taint, as it were, the other independent proceedings. Especially in light of the fact that the evidentiary burden and relief sought are different in both proceedings. In the infringement action, where the applicant is defending a claim brought against it, I see no reason why the applicant cannot rely on a ground that has no previous finding on the merits. The applicant is now raising obviousness as a defence in infringement proceedings.
 In my view, the alleged procedural error is a red herring and only serves to cloud the real issue at hand. The real issue is whether the applicant can raise and rely on (as a defence in the infringement action) a different cause of action, which was not adjudicated upon in the revocation proceedings and regarding which there is no decision on the merits.
Surely, the answer must be yes. I am not convinced that there is any special reason to extend a ruling of res judicata to include the same legal conclusion. This holds even when an alleged procedural error by the applicant is the underlying reason why the cause of action was not heard on the merits or formally separated in the first proceedings.
 It must be unequivocally stated that the applicant is not asking this Court to overlook or even condone its alleged procedural error. Counsel for the applicant conceded during oral hearing that the applicant ought to forfeit its claim in the revocation proceedings and rightfully so.
 The High Court erroneously held that the certification of the claims in terms of section 74(1) of the Act by the Supreme Court of Appeal further evinced that the matter was res judicata in that the patent had been certified as being valid and validity could not be further challenged on any other ground. Could this be the significance of section 74(1) of the Act? Surely, the answer is no.
 Whenever the validity of a claim contained in a patent is unsuccessfully challenged in any proceeding, a Commissioner of Patents may issue a certificate confirming that the contested claim is valid. However, the certificate of validity is a misnomer in that the certificate does not purport to be a certificate that a claim of a patent is valid; instead the object and import of the certificate is to certify that the validity of the claim was challenged and that the claim was declared to be valid. As Burrell explains, it would be desirable to call it a certificate of contested validity. This would be in line with section 65 of the United Kingdom’s Patents Act, which contains an almost identical provision but refers to it as a “certificate of contested validity”.
 Therefore, the Supreme Court of Appeal merely certified that certain claims under novelty were at issue and were found to be valid. Thus, it certified the contested validity of the 1998 patent and did not purport to certify the patent itself as being valid as a whole.
 Section 74(2) is also pivotal in this case. Section 74(2) provides that where any party attempts unsuccessfully, to challenge a claim which has been certified in terms of section 74(1) in subsequent proceedings, that party must pay to the patent-holder party his full costs, charges and expenses as between patent agent or patent attorney or attorney and client insofar as that claim is concerned, unless a court orders otherwise. This is an important internal mechanism which has the effect of discouraging parties from repeatedly litigating certified claims in other proceedings as any party doing so will be mulcted in costs, on a seemingly punitive scale. In this case, if the applicant tried to re-litigate novelty relying on the certified claims, it would run the risk of being mulcted in costs.
 In my view, this provision suggests that the Act itself does not prohibit subsequent litigation on the validity of a patent. This is buttressed when one looks at the plain language of the section which provides that it is not limited to a third party but it is open to any party. Thus, in light of the specific provision in this specific Act, it suggests that a party can re-litigate a certified claim (with the threat of a cost order, of course). In effect, the import of this provision is that the Act was clearly designed to provide a remedy to circumvent and discourage, not prohibit, the re-litigation of claims that have been unsuccessfully contested and challenged.
Rule 28 of the Uniform Rules of Court
 In addition, to the reality that the amendment in the infringement proceedings was not res judicata, it should have been allowed under rule 28(1) of the Uniform Rules of Court in any event.
 Rule 28(1) provides:
“Any party desiring to amend any pleading or document other than a sworn statement, filed in connection with any proceedings, shall notify all the parties of his intention to amend, and shall furnish particulars of the amendment.”
 It is evident that this rule is an enabling rule and amendments should generally be allowed unless there is good cause for not allowing an amendment. This was enunciated in Moolman where the court held that:
“[T]he practical rule adopted seems to be that amendments will always be allowed unless the application to amend is mala fide or unless such amendment would cause an injustice to the other side which cannot be compensated by costs, or in other words, unless the parties cannot be put back for the purposes of justice in the same position as they were when the pleading which it is sought to amend was filed.”
 Therefore there appears to be no good reason why the High Court refused to allow the amendment of the applicant’s plea in light of this enabling rule and the principle laid out in Moolman. The respondents did not show that the amendments were requested mala fide and did not argue that there is prejudice that would be suffered by it that cannot be cured by an appropriate costs order.
 In light of the rule 28 finding above, one may logically question the need to make a pronouncement on whether section 61 of the Act constitutes a single cause of action and whether the matter is in fact res judicata. The answer lies in the actual finding of the High Court. The High Court went into an interpretative exercise in relation to section 61 and made a definitive finding in this regard. This finding, as illustrated above, was plainly wrong and it warrants the attention of this Court.
A refusal and reluctance by this Court to overturn the findings of the High Court could result in that court allowing the amendments to be effected but then deciding that the matter on the merits is res judicata, based on its existing interpretation of section 61. The question whether the amendments should be allowed is not the only issue with the High Court’s judgment. It is one of many questions, the most pertinent being the two issues identified by this Court. Furthermore, I discern no plausible reason as to why this Court should just decide this matter on rule 28 solely, in light of the clearly wrong findings of the High Court.
 As has been noted at the beginning of this judgment, the Court has not been called to pronounce on whether the applicant’s failure to separate other claims for the invalidity of the 1998 patent may still constitute an abuse of process. There may be compelling arguments to suggest that a litigant with access to the legal resources that the applicant has should not be allowed a second bite at the cherry.
However, the High Court’s decision to refuse the applicant’s plea amendment has prematurely stifled this line of enquiry. It is clear from the preceding discussion that the applicant should not have been refused this amendment to its plea. However, once the amendment has been accepted, the question remains open in possible trial proceedings that the applicant’s conduct still somewhat amounts to an abuse of court process.
As was noted in Janse van Rensburg,
“it is for the party who relies on the application of the rule pertinently to plead such reliance and lay a foundation in fact which would enable the opposing parties to deal with such reliance.”
Accordingly, this is a factual investigation which is better left in the discretion of the High Court.
The second judgment
 I have had the pleasure of reading the judgment penned by my brother Cameron J (second judgment). I understand the nub of the second judgment as purporting to relax, and perhaps even extend, the principle of res judicata in this case so that it applies even when the causes of action and the remedy are not the same – on the basis of justice, equity and fairness. For the reasons that follow, I cannot agree with this reasoning and outcome.
 The second judgment is clear that the principle of res judicata should apply when the same “issue” is at play. But, the second judgment never defines what an “issue” is, and never explains why we ought to reinvent res judicata by replacing “cause of action” with “issue”. Nor does the second judgment explain why its concerns cannot be addressed by the clear case law directed at abuse of process claims.
 This last problem is best illustrated in the second judgment:
“If an alleged infringer, who fails to make a successful case for revocation, is permitted to raise further invalidity defences when later sued for infringement, there can be no principled reason to preclude it from launching a fresh revocation claim, on any new ground. When that failed, the patent-holder would have to initiate a yet further damages claim. To which the alleged infringer could respond with further new defences. And on and on.”
 There is nothing in my judgment that creates carte blanche for parties to institute revocation proceedings on a repetitive basis and should not be understood as doing that. In those circumstances, the considerations of abuse of process may well be relevant and the lower court would be alive to those considerations.
 Notwithstanding the above, this Court in Mkhize, endorsing what was held in Smith, explicated that the “same issue” enquiry is whether an issue of fact or law was an essential element of the judgment on which reliance is placed. Accordingly, each case will depend on its own particular facts and any extension of issue estoppel will be done on a case-by-case basis, with the relevant considerations including questions of equity and fairness to the parties and also to others.
A failure to carefully circumscribe the defence of res judicata and, by extension, issue estoppel may cause hardship and injustice. In my view, what the second judgment purports to do is dramatically widen the scope of res judicata in patent law in perpetuity, while leaving future adjudicators with little wiggle room to invoke the nimble, court-developed determinations of equity to decide whether this really is the “same issue”.
 In this particular case, the respondents did not make out a case before us as to why equity and fairness should preclude the applicant from raising the validity of the patent (as a defence) on grounds of invalidity that have not been finally adjudicated by a court and have crucial differences between them.
 The second judgment relies on an analysis of patent laws in three foreign jurisdictions primarily – Germany, the United States and Japan. In essence, the argument put forward is that in each of these three jurisdictions statutes have been written to prevent a litigant from litigating over the validity of a patent in a revocation application and then subsequently in a separate infringement or damages dispute between the same parties.
I do not agree with the conclusion from this analysis. The argument in the second judgment, that all of these jurisdictions deemed it necessary to legislate to create an explicit rule lends support to the fact that under common law, the applicant’s defence of invalidity based on new causes of action is not prohibited by res judicata. In effect, the second judgment appears to develop the common law in a new direction without the parties having pleaded for this.
 More importantly, these foreign jurisdictions are all examination states. This means that by the time the first revocation proceeding between the two parties begins the State has already tested and at least initially verified the validity of the patent on all of the statutory grounds creating causes of action against validity.
We must not lose sight of the fact that testing the validity of patents is in the public interest because patents create artificial monopolies. Currently, South Africa completely relies on private parties to regulate this artificial monopoly system because the government does not examine a patent’s validity upon registration. Instead of deterring litigants, who are working both in a private capacity and for the public interest, there should be an inclination to encourage them to bring more revocation challenges, not to create extensions in common law that increase the costs and risks of doing so.
 Although the applicant prayed for costs for three counsel, it did not furnish this Court with reasons as to why this Court should award costs for three counsel. Accordingly, I would have awarded the applicant costs of two counsel.
 I would have made the following order:
1. Leave to appeal is granted.
2. The appeal is upheld.
3. The order of the High Court is set aside and replaced with:
“(a) The defendant’s application, dated 15 December 2015, in terms of the provisions of Rule 28 of the Uniform Court Rules is granted.”
(b) The plaintiff’s application, dated 12 January 2016, in terms of the provisions of Rule 28 of the Uniform Court Rules is granted.
(c) Each party is to pay its own costs.”
4. The respondents must pay the applicant’s costs in this Court, including the costs of two counsel.
CAMERON J (Mogoeng CJ, Jafta J, Madlanga J and Mhlantla J concurring):
 The question before us seems highly technical and even obscure. Yet its implications are practical and important. In the High Court, the Commissioner of Patents (van der Westhuizen J) was confronted with applications to amend their pleadings by each of the opposing parties in an action the respondent, Merck, brought against the applicant, Ascendis, for infringing Merck’s patent. The Court barred Ascendis’s amendment and allowed Merck’s.
 Merck’s amendment asserted that the validity of its patent had already been conclusively determined. It was res judicata – not subject to further challenge in judicial proceedings. The Commissioner agreed. It was pointless to allow Ascendis to add defences against the damages claim. Those defences would inevitably fail, since the Supreme Court of Appeal had already ruled Merck’s patent valid. Ascendis at no stage disputed infringement, were the patent to be found valid. So all that remained, the Commissioner ruled, was to assess what damages Merck had suffered.
 Was that conclusion right? Khampepe J (first judgment) says No. The first judgment concludes that Ascendis’s amendment should have been allowed. This is because the Act creates two-track proceedings in patents disputes:
- first, revocation, under Chapter X of the statute, where a challenger (Ascendis in this case) initiates proceedings to remove a patent from the statutory register; and,
- second, infringement, under Chapter VI, where a patent-holder vindicates the rights the statute gives it against an infringer.
The first judgment counsels against concluding, in the face of this statutory duality, that findings in revocation proceedings have final, binding effect on a later infringement action. It warns this might render the dual proceedings nugatory.
 This approach the first judgment grounds in an analysis of the Act that finds that each ground of revocation –
- absence of novelty;
- non-inventiveness or obviousness; and
- lack of usefulness or inutility –
is a separate statutory cause of action entitling the challenger to raise each one, at will, in either revocation proceedings or later in defending an infringement claim.
 The first judgment would therefore set aside the Commissioner’s order that refused Ascendis’s amendment, and, instead, allow it. I respectfully differ. My principal reason is this. Why does Ascendis now want to introduce new defences against Merck’s infringement claim? It is because it previously tried and failed to invalidate Merck’s patent on the ground of novelty. In those proceedings, Ascendis, as the first judgment rightly finds, abandoned its obviousness (or non-inventiveness) defence when it argued, unsuccessfully, for revocation.
The question is this. Should the courts countenance multiple-stage defences in patent disputes – first bite at revocation, second-bite when sued for infringement? I think not. This is not how enforcement of patents should most fairly and efficiently work.
 I differ from the first judgment in what the default position should be in patent proceedings. If an alleged infringer, who fails to make a successful case for revocation, is permitted to raise further invalidity defences when later sued for infringement, there can be no principled reason to preclude it from launching a fresh revocation claim, on any new ground. When that failed, the patent-holder would have to initiate a yet further damages claim. To which the alleged infringer could respond with further new defences. And on and on. The resulting dissonance in the two sets of patent litigation seems calculated to produce not only incoherence, but to affront directly our long-held judicial caution against piecemeal litigation.
 This dissonance would surely arise, should Ascendis be allowed, having previously failed, once more to challenge the validity of Merck’s patent. On the one hand, a successful challenge will create a tension between two sets of courts, first-instance and appellate, reaching two determinations on the same issue, the validity of the patent. On the other, if Ascendis’s second challenge were to fail, then the first judgment entails further validity counter-claims on appeal.
 The Commissioner rightly discerned that the “real issue” between the parties was whether the judgment of the Supreme Court of Appeal, which rebuffed Ascendis’s attempt to revoke Merck’s patent, conclusively determined the patent’s validity. Throughout, the parties’ contest has been about nothing other than whether Merck’s patent is valid. Why else are they litigating? That has been the issue and the sole issue all along. And it is the issue here. In my view, that question – the patent’s validity – has been conclusively determined between these parties. In lawyer speak, it is res judicata.
 In differing from the first judgment, I borrow with appreciation from its thorough overview of the principles. The doctrine of res judicata has ancient roots as an implement of justice. It seeks to protect litigants and the courts from never ending cycles of litigation. Its strict terms applied when a later dispute involves the same party, seeking the same relief, relying on the same cause of action.
 But the doctrine’s roots lay in good sense and fairness. This demanded wider application, that barred repeat cycles of litigation on less stringent exaction of the “same cause of action” requirement. And that happened. First, in Boshoff, in the early twentieth century, and then through a line of more recent Appellate Division and Supreme Court of Appeal decisions.
 And so it has become well accepted that enforcing the requirements of res judicata should yield to the facts in each case. Thus, the doctrine was enforced when a plaintiff demanded the same thing on the same ground, or (which is the same) on the same cause for relief, or further, where the same issue had been subjected to final previous judicial determination. And the breadth of what is required when repeat litigation is barred is still being developed. The Supreme Court of Appeal has recently remarked that there is “no reason in principle why a court cannot relax [even] the same person requirement”.
 The label “issue estoppel” came to refer to instances where the same cause of action requirement was not rigorously enforced. Initially, it was contested whether issue estoppel was part of our law and not some alien innovation. Happily, this debate has been put to rest. Issue estoppel as an extension of res judicata is a firm part of our law. And so it should be, for whatever the sobriquet it was merely an extension of the well-established principles of res judicata. That is why there is no need here to reinvent res judicata.
 The first judgment sustains Ascendis’s contention that issue estoppel should not apply here – for otherwise the amendment has to be disallowed. I cannot agree. Only two reasons could counter-indicate applying issue estoppel here.
- First: does the statute allow continued litigation of already-decided validity disputes?
- Second: do the equities here favour less harshness in enforcing res judicata and issue estoppel?
The answer to both questions is in my view No.
 Ascendis suggested that the statute evidences an asymmetry between challenges and defences. This is because it provides that, where sued for infringement, a defendant may counterclaim for revocation, “and, by way of defence, rely upon any ground on which a patent may be revoked”. The first judgment embraces this view. But this does not square with the statute or its application in the specialist patent courts. It also gives insufficient weight to this case’s procedural history.
 Section 65 envisages damages proceedings that the defendant may turn into a contest about the patent’s validity. If that defence succeeds, this results in the patent being removed from the register. By contrast, if validity is raised as a mere defence, without a counterclaim for revocation, the patent remains on the register and the patent holder is free to pursue infringement claims against any others. But the fact that the Act allows a vigorous defendant to counter-attack to remove the plaintiff’s patent from the register does not mean that, should that defendant earlier have sought the patent’s revocation, but failed, it may repeatedly raise the patent’s validity in successive bouts of litigation. Ascendis here itself initiated revocation proceedings. This set the parties on an adversarial path that would end in the final determination of the patent’s validity in that first round of litigation.
 It is in this way that our patent courts have consistently approached the matter. In Deton the Court refused to entertain an invalidity challenge in an application for amendment that was heard before already-pending revocation proceedings were determined. The amendments were objected to on the grounds of delay, absence of explanation for it and on various grounds challenging the validity of the patent. The Court reasoned that it should avoid arriving at a validity determination that would obviate the invalidity challenge. This would happen through the “usurpation” or “irrelevance” of the revocation proceedings on “the basis of issue estoppel”. The Court questioned whether this could ever be consonant with the statute. And rightly so.
 Here, the Commissioner, in concluding that the Supreme Court of Appeal had finally determined the patent’s validity, relied on Alcatraz II. Rightly so. There, a first revocation challenge failed. A second challenge was then launched.
To this, the patent-holder excepted on the ground that validity had been determined in its favour (res judicata).
To avoid the exception, the challenger contended that the previous court did not adjudicate the evidence relating to the prior art; nor was there a determination on the lack of inventive step. But these contentions were fanciful, since the previous court had dismissed the evidence in question for late filing and there was indeed a final judgment on the merits.
In coming to this conclusion, the Alcatraz II Court, borrowing from the English courts, remarked that—
“[w]hen the question of the validity of a patent is brought for trial by reason of the defendant’s contesting that question, he is bound to put his whole case before the Court; and if he does not do so, then it is his own misfortune.”
 There is good reason for this. To allow a re-run of the validity challenge during the second-stage damages claim would lead to intolerable dissonance. It is true that the revocation determination applies against the whole world (in rem), since the patent is removed from the register, while infringement proceedings apply between the parties only (in personam). But this does not help Ascendis, since the logic and justice of res judicata apply primarily and in the first instance between two parties, regardless of whether an outside-applicable standard like a register makes the determination applicable in rem.
 Nor does section 74 assist. This empowers the patent court in any proceedings in which the validity of any claim in a complete specification is challenged, to certify a finding that the claim is valid. The provision further requires a special costs award against “any party” who later unsuccessfully attacks the validity of the patent. The Court in Alcatraz II concluded that, on a proper construction, “any party” meant any other party, namely a third party. Absurdity would otherwise result.
 Nor does allowing resuscitation of the question of validity in the case of revocation challenges vis-à-vis subsequent enforcement claims square with judicial doctrine in contract enforcement. The courts disallow a contracting party who has tried but failed to invalidate a contract from raising a new defence to the contract’s validity when sued for damages for its breach.
This is not inflexibly applied, nor should it be in patent litigation. Allowing or disallowing a new defence should be done case-by-case. This emerges in the contract context from Prinsloo. There a court in motion proceedings found that Mr Prinsloo had committed fraud. But the conclusion that he had committed fraud arose from the rules applying to affidavit evidence. There was no trial. The Supreme Court of Appeal allowed Prinsloo to defend later proceedings for damages on the basis that he had not in fact committed fraud. Clear injustice would have arisen if full examination of the facts concerning the alleged fraud was forbidden.
 The position should be the same in patent litigation. The default position should be that a previously unsuccessful revocation applicant is precluded from raising the validity of the patent in a subsequent damages claim. And if this is unfair, then the court should allow the later defence to be raised, exceptionally, for that reason.
 The party raising the exception cannot be required to show that upholding it would be just. That would run counter to our jurisprudence and invert the requirements of res judicata itself. It would create an anomaly, in which, despite the grounds for the exception being established, it would not, without more, be upheld. It was Merck that, in this Court and in the courts below, objected to Ascendis’s amendment on the precise ground that allowing it raised the spectre of a cascade of invalidity challenges and contradictory validity determinations.
 That the Act does not sanction endless validity challenges is prudent and squares with the position in the majority of foreign jurisdictions. In Germany, validity challenges and damages claims are strictly bifurcated. Specialist courts are tasked with deciding either issue. This creates the problem that disparate validity determinations may be reached. The solution is this. German law bans validity challenges in the damages proceedings.
 In the United States, similarly to our own, a unitary system applies, in which challenges and damages claims are contested in the same forum. Review inter partes is permitted before the Patent Trademark Office (PTO), which heralded a shift to a more bifurcated system. By creating a separate forum for expedited validity challenges, defendants were shielded from the expense and risk common to patent proceedings. The penalty was, however, duplicated validity challenges. The legislation therefore included a statutory form of estoppel. An unsuccessful challenger is thus barred from raising any ground that was previously relied upon or could have reasonably been raised during the challenge.
 Japan, traditionally a strictly bifurcated system, created its own safeguards. Validity may be raised as a defence in infringement proceedings, but the court awarding damages has no discretion to revoke or declare the patent invalid. Hence the bifurcation between challenges and damages claims is preserved.
 Given that our own system allows for challenge to the patent plus damages determination in the same proceedings, one may raise the pertinence of bifurcated systems abroad. Ascendis has attempted to draw support for its second-stage attack on Merck’s patent because of the bifurcation in our patent litigation. This inference is unreliable. While we do have a hybrid system, one that allows for stand-alone validity challenges, we equally have a home-grown solution to any concern about duplication. This is because a challenger who fails in revocation proceedings, can and should be confronted with a plea of res judicata should it raise validity in the infringement action. Other defences to the damages claim can of course be raised.
 Nor should we impose a bifurcated system onto the Act with all the consequences that might entail. It is true that bifurcation offers the benefit that specialist courts can deal with each issue. But this would come at a price: since bifurcation may result, as would happen here, in asymmetrical patent claim construction. This is because, in the validity challenge, a patent-holder, in warding off challenges on breadth and precision grounds, would contend for the narrowest interpretation of the claim, but, by contrast, in infringement proceedings, when seeking to vindicate the patent, assert its widest possible construction. The patent holder doing this is likened in the literature to an Angora cat – smooth-furred, cuddly and small in the validity challenge, but tiger eyed, bristly-furred, aggressive, when the patent-holder is on the attack. The English Court of Appeal has found this to be a “deeply regrettable” state of affairs that illustrates “yet again the need for a one-stop patent shop”.
 It is correct that, as a “non-examining country”, government does not examine a patent’s validity upon registration. The concerns this triggers suggest we should resist bifurcation, as developments in foreign jurisdictions warn. Imposing bifurcation, in the absence of any firm foundation, magnifies and does not diminish concern about the ease with which validity challenges can be initiated. Bifurcation unnecessarily points us toward an invidious choice: should we accept bifurcation, and with it allow successive validity challenges, or should we accept bifurcation with only single validity challenges, at the risk of detriment to the less-resourced? And if we wish to provide a salve for the exorbitant costs of patent litigation, then to allow multiple validity challenges is a remedy only the ultra-well-resourced could enjoy.
 That need not be the way. An alternative lies in non-bifurcation, matched with equitable application of res judicata. As a result of over-zealous extension, barring a claim or defence may wreak “great hardship and even positive injustice to individuals”.
Recently this Court overrode even res judicata proper, in exceptional circumstances, because justice demanded it.
 The first judgment concludes that the Commissioner’s refusal to allow Ascendis’s amendment has prematurely stifled enquiry about justice and fairness. That is not so. The Commissioner on careful consideration refused Ascendis’s amendment because justice required that the validity of Merck’s patent be considered res judicata. This is precisely what Merck asked it to do and what Ascendis urged it to resist. (That the Commissioner’s reasoning was not in the nomenclature of issue estoppel is neither here nor there).
 In this Court, Ascendis urged that it would be “grossly unfair” if the amendment was barred. The hard facts of the parties’ dealings counter-indicate this. In April 2012 Ascendis (then Cipla) applied for its motion to revoke Merck’s patent to be referred to oral evidence. In May 2012 the parties agreed to stay the infringement action until Ascendis’s application for revocation was finally determined. In August 2012 Ascendis’s attorneys wrote to Merck’s attorneys, expressing the view that oral evidence was needed on its obviousness attack on the patent – but not on its novelty or anticipation attack. They proposed that Ascendis’s revocation application and the application for oral evidence be set down for hearing on the same date. Merck’s attorneys refused: Merck did not agree to a piecemeal hearing, and Merck would deal with any conditional application for a referral to oral evidence if and when it was made.
 The revocation application was heard by Teffo J on 13 March 2013. In the introductory section of their written argument, Ascendis’s counsel sketched various permutations in which Ascendis might ask for referral to oral evidence on the obviousness point. They concluded this section by saying that the present hearing was “confined to the issue of anticipation [novelty] only and no oral evidence is required to decide this question”.
In their written argument, Merck’s counsel expressly averred that Ascendis had chosen to limit their argument to the anticipation point and did not persist with the obviousness point. On oral evidence, Merck submitted that there was no need to have regard to an application that might or might not be made. But they noted that the procedure Ascendis proposed was incompetent – and that Ascendis’s failure to advance argument on the obviousness point amounted to an abandonment of it; any attempt later to resurrect it would be opposed on the basis of res judicata.
 Teffo J delivered judgment on 11 March 2014. She upheld Ascendis’s anticipation (novelty) point and revoked Merck’s patent. On appeal to the Supreme Court of Appeal, Merck’s counsel recorded that Ascendis had not persisted with the obviousness point and that any application for oral evidence on the obviousness point would have been incompetent as violating res judicata and that the appeal was thus confined to the novelty (anticipation) point.
 Ascendis’s counsel merely noted that, in addition to the anticipation point, certain of the patent’s claims were challenged on grounds of obviousness. Nothing more on the subject was said. Merck’s heads on the crucial points were not challenged. The Supreme Court of Appeal overturned the judgment of Teffo J, holding that the anticipation (novelty) point was bad. Soon after, Merck’s attorneys wrote to Ascendis’s attorneys stating that the appellate judgment determined the only matter in issue in the infringement proceedings.
Since it was common cause that, if Merck’s patent was valid, Ascendis was infringing it, they demanded that Ascendis consent to the relief Merck claimed in the infringement action. Failing this, Merck would go to trial and seek an interim interdict plus punitive costs order.
 It was only after this that Ascendis sought to amend its plea in the infringement action by deleting the anticipation (novelty) point and adding the new, inutility, attack on the validity of Merck’s patent. This notice of amendment signalled Ascendis’s intention to persist with its attack on the validity of Merck’s patent, relying on the already-abandoned obviousness point and the (new) inutility point. Still, Ascendis offered no reason as to why it did not raise the inutility ground in the earlier validity challenge. Merck’s response was to seek to amend its pleadings positing the judgment of the Supreme Court of Appeal as rendering the validity of its patent res judicata.
 Meanwhile, Merck obtained an urgent interdict pending the final determination of the infringement action. Louw J found that although only the anticipation (novelty) point had been argued before the Commissioner and the Supreme Court of Appeal, Ascendis had been obliged to advance all its validity challenges when it applied to revoke Merck’s patent. Had it wanted to preserve the obviousness (non inventiveness) point, Ascendis should have applied to separate the issues. Since it didn’t do so, the patent’s validity was now res judicata.
 Just or unjust? In my respectful view justice points to treating the Supreme Court of Appeal judgment refusing Ascendis’s revocation argument as conclusive of the patent’s validity. There can be no reason why res judicata does not apply, appropriately expanded. No doubt this prejudices Ascendis. Of greater consequence is the destabilisation of our patent litigation system, through creating a perilous and novel default. This would be to the prejudice of all. The courts in both the amendment proceedings and the interdict application were alive to this fact. The issue of validity between the parties is long-worn and settled. I would grant leave to appeal, but dismiss the application.