2
(PEPUDA). Section 7 (b) of PEPUDA outlaws “engagement in any activity
which is intended to promote, or has the effect of promoting, exclusivity, based
on race”. Section 7 (c) forbids “the exclusion of persons of a particular race
group under any rule or practice that appears to be legitimate but which is
actually aimed at maintaining exclusive control by a particular race group”.
2 Mr. Ngcaba was, until mid-2017, the executive chairman of Dimension Data.
In that capacity he was party to a service agreement, in respect of which he
received a salary and was entitled to a bonus. Mr. N gcaba says that the
service agreement also entitled him to participate in Dimension Data’s long-
term incentive plan, an aspect of which was a share appreciation rights
scheme. Mr. Ngcaba says that, in breach of his service agreement, and in
breach of its statutory duty not to unfairly discriminate against him based on
his race, Dimension Data failed to procure his participation in both the
incentive scheme and the share rights scheme. He says that eligible White
executives were allowed to participate in those schemes, and that he was
excluded from them because he is Black. He says that if he had participated
in those schemes, as he was entitled to do, he would have been paid just over
R117 million which he did not in fact receive. These, he says, are his damages.
3 Mr. Ngcaba’s action was instituted in June 2018. Since then, both parties have
amended their pleadings more than once, and there have, by my count, been
four interlocutory skirmishes resulting in full opposed argument and written
judgments by four other Judges of this court. Almost eight years later, I have
been asked to decide the fifth such skirmish. It is an opposed application to
separate three special pleas Dimension Data raises from the remaining issues
3
in the action. The first special plea is that this court lacks the necessary
jurisdiction to adjudicate Mr. Ngcaba’s claims, both of which, Dimension Data
says, concern its duty as an employer not to discriminate against one of its
employees. Claims of that type, it is alleged, lie properly within the jurisdiction
of the Labour Court. The second special plea is that Mr. Ngcaba’s claim based
on PEPUDA has prescribed. The third special plea is that Mr. Ngcaba’s
contractual claim – if that is indeed what it is rather than in substance a claim
under statutory labour law – must be arbitrated rather than decided in this
court.
4 Much of the argument before me concerned the merit of the special pleas
themselves, but, as both counsel conceded, the merit of the special pleas is
but one factor to be weighed when considering whether they should be
separated from the other issues in the action. The primary question is not
where the prospects of success lie on the issues sought to be treated
separately, but whether it is convenient, in the relevant legal sense, that those
issues be determined separately.
5 To be capable of convenient separation, the issues to be separated must be
conceptually distinct. They must also generally be capable of determination
without hearing evidence that will later have to be repeated in relation to other
triable issues. It is also desirable that a repetition of witnesses is avoided,
especially if credibility findings will have to be made in relation to particular
witnesses who may have to testify more than once in respect of different
issues.
4
6 Once a court is satisfied that the issues sought to be isolated for separate
determination are conceptually distinct from those arising in the rest of the
action, and that there will be little or no overlapping evidence required to hear
them, the question is whether there is some advantage to be had in ordering
a separation which is not outweighed by any obvious or foreseeable
disadvantage. If the advantages outweigh the disadvantages, the court will
generally order the separation (see S v Malinde 1990 (1) SA 57 (A) at 68C–
E).
7 In this case, the first leg of the test is clearly satisfied. The issues offered up
for separate determination are conceptually distinct, both from each other, and
from the other issues arising in the main action. I need not concern myself with
the repetition of evidence, or with contradictory or embarrassing credibility
findings, since the separated issues can all be decided as points of law arising
on the pleadings alone.
8 It is at the second stage of the test that the separation application runs into
trouble. In balancing the putative advantages of a separation against its
foreseeable disadvantages, I am required to give “careful thought . . . to the
anticipated course of the litigation as a whole” (Denel (Pty) Ltd v Vorster 2004
(4) SA 481 (SCA) at paragraph 3). Here, I cannot say that the parties have
provided much assistance. Dimension Data asks that I consider the principal
advantage it says a separation will bring: the avoidance of a very long and
costly trial on a multitude of issues (there are a further two special pleas
requiring evidence that will not be separated, and, of course, the defences to
the main claims). Yet Dimension Data does not tell me how long it expects
5
that trial to be, how many days it will take, how many witnesses will be called
and, accordingly, how much time, money and trouble will be saved if I grant
the separation it seeks (on the necessity of such information see, for example,
CC v CM 2014 (2) SA 430 (GJ) at paragraph 27).
9 Mr. Ngcukaitobi, who appeared together with Mr. Sisilana for Mr. Ngcaba,
argued that the failure to adduce this information meant that Dimension Data
had failed to discharge an onus to show that the separation would be
convenient. Ms. Cane, who appeared with Mr. Michael and Mr. Mabuda for
Dimension Data, argued that there is no such onus. But, wherever the overall
onus lies, without facts that would allow me to estimate the length of the trial,
I cannot assess with any precision the time that would be saved by ordering a
separation.
10 I will nevertheless assume, in Dimension Data’s favour, that the trial will take
at least three weeks, which places it toward the lengthy end of the civil trials
that are routinely heard in this court. I do not think that I was offered a estimate
of how long the special pleas would take to argue, but I was given a foretaste
of what would be argued. Based on that preview, I think two days would be a
sensible estimate. It follows that, conservatively, thirteen days of court time
might be saved by separating the issues in the way Dimension Data suggests.
I do not know whether a trial of that length falls comfortably within the stamina
and resources of both parties, but since neither party said it did not, and since
both parties have litigated fairly lavishly until now, I think I am bound to assume
that it does.
6
11 Against that potential saving, I must weigh the prospect of a piecemeal appeal.
Trial courts are fairly regularly criticised for ordering separations which turn
out not to commend themselves to courts of appeal. These criticisms are
usually made when the court of appeal is forced to deal with a “piecemeal”
appeal against a trial court’s rulings on the separated issues. Where
successful, these appeals waste the time a separation of issues is often meant
to save (see, for example, Firm-O-Seal CC v Wynand Prinsloo & Van Eeden
Inc 2024 (6) SA 52 (SCA), paragraphs 2 and 10 and Theron and Another NNO
v Loubser NO 2014 (3) SA 323 (SCA), paragraph 18). I think that much of this
criticism is too demanding (see Biologicals and Vaccines Institute of Southern
Africa (Pty) Ltd v Guardrisk Insurance Company Limited [2025] ZAGPJHC
1184 (21 November 2025) at paragraph 12). It is, however, beyond question
that a trial court ought to bear in mind the possibility of an arguable piecemeal
appeal on a separated issue being pursued.
12 Still, a separated issue may conceivably turn on the credibility of one or two
witnesses, the assessment of which is unlikely to be realistically challenged
on appeal. It may also be that one party’s case on a separated issue is so
strong that it hardly seems worth putting anyone to the expense of a lengthy
trial. In those situations, the prospect of an arguable piecemeal appeal against
a judgment on the separated issue will be remote.
13 But neither of these situations applies here. There is no evidence to be led in
respect of the special pleas, so the credibility of witnesses need not be
assessed. More importantly, the issues raised in the special pleas are all
questions of law, and they are of at least moderate complexity.
7
14 The jurisdiction question is not without difficulty. Dimension Data says that,
because Mr. Ngcaba’s claim relates to discrimination against him as an
employee, the Labour Court has jurisdiction, under the applicable provisions
of the Employment Equity Act 55 of 1998 (EEA). Mr. Ngcaba’s retort is that
the discrimination he alleges has more to do with his status as a shareholder
and investor in Dimension Data. In any event, Mr. Ngcaba points out, the
primary claim is for a straightforward breach of contract. There is no pleaded
reliance on PEPUDA or the EEA to advance that claim. It is well-established
that an employment dispute pleaded exclusively in contract may be
entertained in this court (see Baloyi v Public Protector 2022 (3) SA 321 (CC)
at paragraph 41). But the proper characterisation of Mr. Ngcaba’s claim is not
a straightforward question. Certainly, I cannot say that Dimension Data has
overwhelming prospects of showing that Mr. Ngcaba’s claim is not a matter
that engages this court’s common law jurisdiction.
15 The same goes for the special plea of prescription. Dimension Data says that
the claim based on PEPUDA is effectively a delictual claim for damages. That
being so, the damages asked for are a “debt” within the meaning of section 11
(d) of the Prescription Act 68 of 1969. Dimension Data also says that debt
prescribed in 2019, some five years before the PEPUDA claim was introduced
by way of amendment in 2024. By contrast, Mr. Ngcaba says that the PEPUDA
claim is not a claim for a “debt” under the Prescription Act but for statutory
damages under PEPUDA itself.
16 Mr. Ngcaba’s position is not without support. In the context of administrative
law, the Supreme Court of Appeal has held that claims for compensation
8
created by statute and based on the breach of a public law duty are not “debts”
for the purposes of the Prescription Act (see Petersen v SASSA 2025 (3) SA
153 (SCA) at paragraph 21). Both the Constitutional Court and this court have
only reluctantly assumed that they do constitute such debts (see Njongi v
MEC, Department of Welfare, Eastern Cape 2008 (4) SA 237 (CC) at
paragraph 42 and SABC v Motsoeneng 2025 (2) SA 571 (GJ) at paragraph
14). In light of these authorities, it seems to me that whether a claim for
damages under PEPUDA constitutes a “debt” under the Prescription Act
remains very much an open question.
17 The fate of the arbitration special plea is also far from a forgone conclusion.
Mr. Ngcaba says that there is good cause to release him from the arbitration
clause, because only this court, sitting as an Equality Court, can determine
the PEPUDA claim, and because he challenges the constitutional validity of
clause 21.4 of the service agreement. Mr. Ngcaba says that clause 21.4
purports to exclude his reliance on PEPUDA.
18 During argument, Ms. Cane informed me that Dimension Data now accepts
that clause 21.4 does not preclude Mr. Ngcaba’s reliance on PEPUDA. But
even if I were to assume that this renders that point moot, the question remains
whether Mr. Ngcaba should be required to go to arbitration on his contractual
claim while pursuing his PEPUDA claim in this court. Those questions are
entangled with the question of whether Mr. N gcaba’s primary claim really is
contractual in nature at all, and with the question of whether the PEPUDA
claim has prescribed. If it turns out that this court can entertain the primary
claim and the PEPUDA claim, and that the PEPUDA claim has not prescribed,
9
there may well be much to recommend the view that both should be decided
in this court, sitting simultaneously as a High Court and an Equality Court. I
cannot yet say either way. But the answer is unlikely to be so clear cut as to
preclude the possibility of a piecemeal appeal if the question is separated out
from the main action.
19 Accordingly, the disposal of each of the special pleas is likely to involve a
judgment which is appealable in principle. I can think of no reason to discount
in advance the reasonable possibility that such an appeal would be both
pursued and arguable. If any such appeal takes place before the remainder of
the issues in the action are tried, then the time I might save by separating it
out will be wasted if the appeal is upheld and the matter is returned to this
court for the remainder of the trial to proceed. In these circumstances, though
I am prepared to accept that each of the special pleas has genuine prospects
of success, I do not accept that they are so meritorious as to justify attempting
to short-circuit the rest of the trial.
20 Dimension Data offered no other basis on which the separation it proposes
would be convenient in the required sense. Indeed, it seems to me that, given
the elapse of time since these proceedings were instituted, the most
convenient course of action for everyone involved is to press on with the trial
as soon as reasonably possible.