National Union of Metalworkers of South Africa (NUMSA) obo Nganezi and Others v Dunlop Mixing and Technical Services (Pty) Ltd and Others (DA16/2016) [2018] ZALAC 19; [2018] 10 BLLR 961 (LAC); (2018) 39 ILJ 2226 (LAC); 2018 (6) SA 240 (LAC) (17 July 2018)

About SAFLII
Databases
Search
Terms of Use
RSS Feeds
South Africa: Labour Appeal Court
SAFLII
>>
Databases
>>
South Africa: Labour Appeal Court
>>
2018
>>
[2018] ZALAC 19
|

|

National Union of Metalworkers of South Africa (NUMSA) obo Nganezi and Others v Dunlop Mixing and Technical Services (Pty) Ltd and Others (DA16/2016) [2018] ZALAC 19; [2018] 10 BLLR 961 (LAC); (2018) 39 ILJ 2226 (LAC); 2018 (6) SA 240 (LAC) (17 July 2018)

IN
THE LABOUR APPEAL COURT OF SOUTH AFRICA, JOHANNESBURG
Reportable
Case no: DA
16/2016
In the matter
between:
NATIONAL
UNION OF METALWORKERS of SA
(NUMSA)
obo KHANYILE NGANEZI AND
OTHERS
Appellants
and
DUNLOP MIXING
AND TECHNICAL
SERVICES
(PTY)
LTD

First Respondent
DUNLOP
BELTING PRODUCTS (PTY) LTD

Second Respondent
DUNLOP
INDUSTRIAL HOSE (PTY)
LTD

Third Respondent
COMMISSION
FOR CONCILIATION,
MEDIATION AND
ARBITRATION

Fourth Respondent
COMMISSIONER
ALMEIRO DEYZEL N.O.

Fifth Respondent
Heard:
14 September 2017
Delivered:
17 July 2018
The
Order: (Sutherland and Coppin JJA; Savage AJA dissenting)
(1)
The
appeal is dismissed with costs.
(2)
The
Order of the Labour Court is confirmed.
(3)
The
appellant shall bear the costs of the appeal.
Summary:
Derivative
misconduct – employer dismissing its entire workforce for
strike-related misconduct – a category of employees
dismissed
for derivative misconduct because of their failure to provide the
employer with information concerning the identities
of the
perpetrators of
violent
acts
during
the
strike
–
Principle
of derivative misconduct restated -central to this appeal is the
question as to whether employees charged with derivative
misconduct
must be identified by the employer – or whether employer must
prove their presence during the misconduct to be
able to infer that
they have actual knowledge of the perpetrators and that their silence
makes out case with the perpetrator thereby
breaching their duty of
good faith towards the employer – the arbitrator was of the
view that failure to positively place
each employee at the scene
where the misconduct occurred is detrimental to the employer’s
case of dismissing those employees
based on the principle of
derivative misconduct.
The Labour
Court distanced itself from this approach and held that sufficient
evidence on the probabilities placed the dismissed
employees on the
scene of the misconduct and drew an inference that on the
probabilities, the employees had knowledge of the perpetrators
–
Appeal
The Labour
Appeal Court endorsed the Labour Court’s view:
Held:
The
arbitrator did not give consideration to the fact that such presence
and knowledge was capable of proof by means of indirect
evidence, or
by inference, and, accordingly, did not determine whether those facts
had indeed been proven by inference.
Further that
the arbitrator adopted a narrow approach to the evidence by requiring
the individual identification of each employee
as being present as a
sine
qua non
for the employees falling into a category of employees implicated on
the basis of derivative misconduct. On the premise that presence
or
absence had to be established on a preponderance of probabilities, it
must follow that indirect evidence in the form of inferences
drawn
from the whole body of evidence was a necessary category of evidence
to assess. There is nothing in the evidence to gainsay
the inference
that, on the probabilities, they each were present for at least some
of the time, and equally probable, that they
were each present most
of the time, even if not everyone was religiously present on each and
every day. From these circumstances,
the inference can be drawn that
it is improbable that each and every one of them could not have
acquired actual knowledge of the
misconduct perpetrated, more
especially because the misconduct, as described earlier, was so
spectacular.
Labour
Court’s judgment upheld and appeal dismissed.
The
concurring judgment addressed the expanded version of the principle
of derivative misconduct by the Court
a
quo
if not attended to would novate the principle established thus far –
the court is critical of the court
a
quo´s
stance
which found inapplicable in labour disputes the duty to speak
and
the right to silence
applicable to
criminal law - the court held that such
notions
negate what was held
,
namely,
that the duty to speak is triggered by actual knowledge of those
facts. This new notion of the principle also implies that
an
employee, who is otherwise totally ignorant of the misconduct and
identity of any of the perpetrators, but who does not speak
or
exonerate himself, or herself, may be disciplined for some form of
unspecified misconduct.
Further
that t
he
protections in criminal law, which include the right to silence and
the privilege against self-incrimination, were intended to
protect
citizens from unfair police and judicial interrogation. Similar
protections would accordingly not be out of place in labour
relations
where potential tyranny by the police, State and the courts is
replaced with potential tyranny at the hands of employers.
Further
that t
he
right to silence and the privilege against self-incrimination are, in
any event, not protections that are confined to the criminal
law.
I
do not agree with a blanket statement that there is no place for the
right to silence in labour relations.
The court
ended with a word of advice to employers to define derivative
misconduct in their disciplinary codes.
Minority
While
appreciating the duty of good faith owed by an employee to his/her
employer, the dissenting judgment (much as did the concurring

judgment albeit for lack of another word) cautioned the use of the
term derivative misconduct in light of the victimisation that
may
suffer employee willing to come forward to divulge information –
The court emphasised that any judicial development of
the principle
of derivative misconduct which had so far up to
Hlebela
be used only
obiter
in a number of decisions, would include an assessment of the
appreciable risks which may arise for an employee in speaking out,
in
naming perpetrators or for purposes of exoneration and the dangers
inherent which may arise in doing so. The court advocated
for an
inclination towards the policy considerations which
FAWU
suggested
in the development of this complex area of law.
The
minority espoused the view of the arbitrator in finding that the
onus
rests on the employer to prove that the employee had actual knowledge
of the information sought before there is any duty on the
employee to
raise a defence. On the evidence before the arbitrator, actual
knowledge could not, in its view, be inferred on the
facts before the
arbitrator.
Coram:
Coppin JA, Sutherland JA
et
Savage
AJA.
JUDGMENT
SUTHERLAND JA
Introduction
[1]
On
26 September 2012, the first, second and third respondents, all
associated companies (referred to collectively as Dunlop) dismissed

their entire workforce. The context, broadly, was that a protected
strike endured for about a month which was characterised by
serious
violence in defiance of an interdict by the Labour Court.
[2]
The
fairness of that dismissal was challenged by NUMSA on behalf of the
dismissed employees and the matter was arbitrated.
[3]
The
arbitrator concluded that no procedural unfairness was perpetrated.
[4]
In
respect of substantive fairness, the arbitrator concluded that there
were three categories of dismissed employees.
4.1.
First,
a category that had been positively identified as committing
violence. Their dismissal was found to be fair.
4.2.
Second,
a category of employees who were identified as present when violence
took place but who did not physically participate.
Their dismissal
was fair, according to the arbitrator, on grounds of what was called
“derivative misconduct”.
[1]
4.3.
A
third category of employees who were not positively and individually
identified as being present when violence was being committed.
Their
dismissal was found to be unfair. This category of employees was
reinstated and comprises the appellant employees.
[5]
Dunlop
brought a review application in respect of the finding of substantive
unfairness in the dismissal of the third category.
The review
application succeeded. This appeal lies against that outcome.
The
controversy
[6]
Central
to the controversy is the meaning and scope of “derivative
misconduct” and the question of whether the third
category of
employees, ie, the present appellant employees, is culpable in terms
of its prescripts. The concept of derivative misconduct
and in its
application to the facts of this case is addressed in this judgment.
[7]
It
is important to grasp what this case is not about. It is not about
the right to strike, which is Constitutionally guaranteed.
The
exercise of the right to strike presupposes that a strike is embarked
upon within the parameters of prescribed norms and rules.
Striking is
an appropriate choice of collective action if it is functional to
collective bargaining. When a strike is marred by
violence and ceases
to be functional to collective bargaining, the right to strike is not
itself impaired by the employer’s
actions to address the
violence.
[8]
Equally
important to the evaluation of the case, is the need to bear in mind
two basic principles. First, the critical question
and enquiry in
every alleged unfair dismissal case is
are
whether the employer’s decision to dismiss can be objectively
justified; the proper approach to this question is always fact
–
specific. Second, in a review, the critical question is whether the
decision of an arbitrator is one which a reasonable
arbitrator could
reach.
[2]
Accordingly,
in keeping with an approach based on those basic principles, the
critical facts, largely common cause, are narrated,
and the
application of the relevant principles to those facts are thereupon
assessed.
The
relevant facts
[9]
A
protected strike began on 22 August 2012. At once, violence became
its hallmark. An interdict against the violence was sought
and
granted on the same day. The evidence put up to the Labour Court
which led to the interdict included incidents of blockades
in
violation of picketing rules, attacks on cars and on the throwing of
stones.
[10]
In
defiance of the interdict, the violence not merely continued but
escalated over the following month, until the entire workforce
was
dismissed. The catalogue of barbarism included the following:
10.1.
Arson:
setting alight the homes of a manager and of a foreman.
10.2.
Damaging
several vehicles belonging to staff and visitors by stone throwing.,
smashing windows, beating people with sticks and on
one occasion,
throwing a petrol bomb.
10.3.
Blockading
the entrances with cut-down trees and with rubble.
10.4.
Throwing
stones at staff and visitors.
10.5.
Assaults
on staff, by stone throwing, kicking and intimidation.
10.6.
Theft
of a camera being used to record the violence.
10.7.
Scrawling
death threats on a billboard.
10.8.
Violation
of the agreed picketing rules.
10.9.
Defiance
of a court interdict.
[11]
The
dismissal was not preceded by any hearings. Dunlop invoked the
justification for not doing so as contemplated in the code of
good
conduct on dismissals.
[3]
[12]
Calls
to come forward and identify the individuals who were violent were
made by Dunlop to the union and to the strikers during
the strike.
This aspect of the happenings is significant.
12.1.
On
22 August, a letter was sent to the union, the letter described the
acts of violence and demanded that the identities of the
culprits be
given to management. The letter made it clear that the culprits would
be disciplined. Moreover, it declared that failure
to provide the
relevant information would lead to a collective hearing at which all
employees were at risk of dismissal.
12.2.
On
29 August, a further letter to the attorney of the strikers described
more acts of violence, including notably, arson, death
threats, and
theft of the camera. Again, the strikers were called upon to identify
the actual culprits, preparatory to a formal
enquiry.
12.3.
On
12 September, a further list of violent acts was given to the union.
The letter drew attention to the contempt of the court order.
The
union’s intervention was requested.
12.4.
On
26 September, the strikers were told,
via
the
union and by SMS to each, that they had been dismissed. The written
dismissal notice alluded to a list of named culprits of
violence and
to “… other individuals…. dismissed on the basis
of derivative misconduct”.
[4]
The
notice went further to state that the levels of violence inhibited it
from convening disciplinary enquiries. Moreover, it stated:
‘
Your
union and attorneys have been called on numerous occasions to provide
information and details of who in fact are participating
in the
specific acts of misconduct and have failed to do so. The company
therefore believes that the employment relationship has
irretrievably
broken down …. Should any individual wish to appeal the
decision to summarily terminate the contracts of employment,
a
collective appeal will be held … on 3 October…. If the
individuals have any reasons to believe why they should
not be
dismissed for derivative misconduct they should present evidence at
the appeal hearing.’
[13]
The
letters were communicated to the strikers by the union, and their
awareness thereof was not in dispute.
[14]
A
hearing was convened. Save for one employee, not one of the others,
including the appellant employees, attended. The employee
who did
attend accounted for herself and she was reinstated on the strength
of her evidence exonerating herself from any participation
in
violence or knowledge of who were the perpetrators of violence.
[15]
An
unfair dismissal dispute in respect
of
all three categories of employees made its way to arbitration by the
CCMA. Prior to the commencement of the proceedings, the
critical
issues were articulated in an exchange of pleadings and in a
pre-arbitration conference, duly minuted.
[16]
The Statement of Claim, alleged that:
‘
The given reason for the
dismissal of the remaining applicants was that they did not inform
the respondents who w
ere
responsible for the alleged acts of violence, intimidation and damage
to property that took place during the strike’.
[17]
Dunlop’s
plea in answer to this averment reads:
‘ …
.the
remaining [workers] have been dismissed for actual misconduct,
derivative misconduct, or that all the [workers] were
in
common purpose
, on the basis there were
various incidents of violence and intimidation and aggression on the
part of the [said workers]
and after
numerous attempts of calling on [Numsa] to furnish details of these
specific individuals that are guilty of this misconduct
and
this resulted in acts of violence, aggression and intimidation
escalated to such an extent that [Dunlop] dismissed employees
for the
said misconduct’. (sic) (underlining supplied)
[18]
Paragraph
3.3 of the minute of the pre-arbitration conference records that:
‘…
in
the context of the background set out in the dismissal letters. The
specific reason given by [Dunlop] for the dismissal of the
78 union
members ….was derivative misconduct’.
[5]
[19]
In
the arbitration proceedings, the witnesses for the strikers denied
any violence had occurred, or, if it had, none of them were
aware of
it and hence not culpable. This version was held by the arbitrator to
be patently false based on the rebuttal evidence,
including video
footage supporting the employer case. This factual finding was not
challenged on appeal.
[20]
A
dismissal of some employees and the reinstatement others followed,
premised on the category distinctions already described.
The
concept of derivative misconduct
[6]
[21]
It
must be recognised that the phrase “derivative misconduct”
cannot be thought of as more than a label, a term of art
to capture a
rather complex idea. Its genesis is in the
obiter
articulation of
an
example
of a breach of the employee’s duty of good faith by Cameron JA
in
Chauke
and Others v Lee Service Station CC t/s Leeson Motors
[7]
(Chauke
)
.
[8]
The
critical passages are these:
‘
[27]
The case presents a difficult problem of fair employment practice.
Where misconduct necessitating disciplinary action is proved,
but
management
is unable to pinpoint the
perpetrator or perpetrators,
in what
circumstances will it be permissible to dismiss a group of workers
which incontestably includes them?
[28] Two
different kinds of justification may be advanced for such a
dismissal. In Brassey & others The New Labour Law (1987)
at 93-5,
the situation is posed where one of only two workers is known to be
planning major and irreversible destructive action,
but management is
unable to pinpoint which. Brassey suggests that, if all avenues of
investigation have been exhausted, the employer
may be entitled to
dismiss both.
[29] Such a case
involves the dismissal of an indisputably innocent worker. It posits
a justification on operational grounds, namely
that action is
necessary to save the life of the enterprise.
That must be
distinguished from the second category, where the justification
advanced is not operational. It is misconduct.
And no innocent
workers are involved: management's rationale is that it has
sufficient grounds for inferring that the whole group is
responsible for or involved in the misconduct.
[30] The present
case illustrates the second category. Management did not advance an
operational rationale for the dismissal. It
charged the 20 workers in
the paint-shop and cleaning and polishing sections with misconduct -
malicious damage to property - and
concluded that they had all been
guilty of it. Was this unfair?
[31] In the
second category, two lines of justification for a fair dismissal may
be postulated.
The first is that a worker in the group which
includes the perpetrators may be under a duty to assist management in
bringing the
guilty to book
.
Where a worker has or may
reasonably be supposed to have information concerning the guilty, his
or her failure to come forward with
the information may itself amount
to misconduct.
The relationship between employer and employee is
in its essentials one of trust and confidence, and, even at common
law, conduct
clearly inconsistent with that essential warranted
termination of employment (Council for Scientific & Industrial
Research
v Fijen (1996) 17 ILJ 18 (A) at 26D-E).
Failure to
assist an employer in bringing the guilty to book violates this duty
and may itself justify dismissal.
[32] This
rationale was suggested, without being decided, in Food & Allied
Workers Union & others v Amalgamated Beverage
Industries
Ltd (1994) 15 ILJ 1057 (LAC) (FAWU v ABI). There a large group
of workers had assaulted a 'scab' driver, leaving
him severely
injured. The company was unable to prove which of those present at
the workplace at the time actually perpetrated
the assault. All those
who had clocked in and who were thus in the vicinity of the incident
when it occurred were charged with
the assault. None came forward at
the workplace hearings or in the Industrial Court to affirm their
innocence or to volunteer any
evidence about the perpetrators. Nugent
J, sitting with assessors John and Satchwell, suggested at 1063B
that:
'In the field of
industrial relations, it may be that policy considerations require
more of an employee than that he merely remained
passive in
circumstances like the present, and that his failure to assist in an
investigation of this sort may in itself justify
disciplinary
action.'
[33] This
approach involves a
derived justification,
stemming from an
employee's failure to offer reasonable assistance in the detection of
those actually responsible for the misconduct.
Though the
dismissal is designed to target the perpetrators of the original
misconduct, the justification is wide enough to encompass
those
innocent of it, but
who through their silence make themselves
guilty of a derivative violation of trust and confidence.
[34]
In FAWU
v ABI, the court held that, on an application of evidentiary
principles, the failure by any of the workers concerned to
give
evidence, either in the workplace hearings or in the Industrial
Court, justified the inference that all those present at the

workplace on that day 'either participated in the assault or lent it
their support' (at 1064B-C
). There were other inferences
compatible with the evidence. But the inference of involvement was
the most likely since (at 1064E):
'This is
pre-eminently a case in which, had one or more of the appellants had
an innocent explanation, they would have tendered
it, and in my view
their failure to do so must be weighed in the balance against them.'
[35] On the same
basis, the court
rejected the unattested suggestion that the
appellants may have declined to come forward because of intimidation
or from a sense
of 'collegiality'
(at 1064E-F
). The court
concluded, in effect from the absence of evidentiary self-absolution,
that it was 'probable that all the appellants
were indeed present
when the assault took place and either participated therein or lent
their support to it'
(at 1064H).’ (underlining supplied)
[22]
The
notion of derivative misconduct was again addressed by Revelas AJA in
Foschini
Group v Maidi
(
Foschini)
[9]
esp
at [47] where the Labour Appeal Court (LAC) endorsed the decision in
Chauke
without
further comment on the concept.
[10]
Then
in
Western
Platinum Refinery Ltd v Hlebela and Others
(
Hlebela),
[11]
the
LAC again addressed the notion.
[12]
At
paragraph [8], with reference to the cited passages in
Chauke
,
it was stated that:
‘
Several
important aspects of the dicta require qualification. Important to
appreciate is that no new category of misconduct was
created by
judicial fiat. The effect of these
dicta
is to elucidate the principle that an employee, bound implicitly by a
duty of good faith towards the employer breaches that duty
by
remaining silent about knowledge possessed by the employee regarding
the business interests of the employer being improperly
undermined.
And, uncontroversially, and on general principle, a breach of the
duty of good faith can justify dismissal.
Nondisclosure
of knowledge relevant to misconduct committed by fellow employees is
an instance of a breach of the duty of good faith
.
Importantly the critical point made by both
FAWU
and
Leeson
Motors
is
that a dismissal of an employee is
derivatively
justified in relation to the primary misconduct
committed by unknown others, where an employee, innocent of actual
perpetration of misconduct,
consciously
chooses not to disclose information known to that employee pertinent
to the wrongdoing.’
(Underlining supplied)
[23]
As
is plain from the cited passages in
Chauke
,
and in
Hlebela
,
the use of the term “derivative” was used in relation to
the “justification” for the attribution of culpability

for a breach of the duty of good faith. It is not obvious that
Cameron JA intended to coin a new noun-phrase: “derivative

misconduct” but the term, used in that way, has passed into the
lexicon of labour jurisprudence, and has endured.
[13]
Accordingly,
in terms of
Chauke
and
Hlebela
,
the utility of the label “derivative misconduct” is to
succinctly label a species of misconduct evidencing a breach
by an
employee of the duty of good faith comprising a refusal to disclose
information relevant to harm being perpetrated by other
persons
against an employer’s interests. Accordingly, by reason of such
omission, the culpability of the actual perpetrators
of particular
misconduct, is fairly attributed to them too.
[14]
[24]
Importantly,
the foundational passages in
Chauke
must
be read for their substantive allusions to the elements of fairness
about an employee’s dismissal, other than as an “actual

perpetrator”.
[15]
The
question posed by Cameron JA was: “Where misconduct
necessitating disciplinary action is proved, but management is unable

to pinpoint the perpetrator or perpetrators,
in
what circumstances will it be permissible to dismiss a group of
workers which incontestably includes them?” The analysis
in
Chauke
is directed at giving the answer to this question, and is rooted in
the nature of the employment relationship, the critical significance

of which was illuminated by the remarks of Nugent J in
Fawu
and Amalgamated Beverage Industries Ltd (FAWU
[16]
(as
cited in
Chauke
)
about the sound policy rationale within a labour relations context
that an employee’s passivity and silence which denies
help to
an employer to protect its legitimate interests is intolerable.
[25]
The
true enquiry remains whether the facts adduced constitute convincing
evidence of a breach of the duty of good faith, and accordingly,
the
elements of a good faith relationship dictate the scope of the
enquiry. The practical application of the concept of derivative

misconduct must be confined within limits that serve the
sustainability of trust and confidence in an employment relationship.

Accordingly, derivative misconduct finds appropriate application to
proven circumstances in which a number of employees find themselves

potentially implicated in misconduct by reason of their membership of
a relevant group or category, and in respect of which, on
reasonable
grounds, suspicion arises that the persons comprising the group or
category must know of material information relevant
to the
perpetration of harm to the employer by persons within that group or
category. Critically, such knowledge includes knowledge
of facts that
may help to identify the actual culprits within the group.
[17]
Thus,
to illustrate: if acts of misconduct occur at a time when the night
shift are on duty, the employees on the day shift cannot,
logically,
be implicated because they are not members of the relevant group. By
contrast, where a number of employees make common
cause with a
(legitimate) course of conduct over time those participants form a
relevant group. If and when the propriety of acts
carried out
en
passant
that course of conduct is placed into question, all those employees
who are identified being participants in the course of conduct
in
which the relevant group is implicated because they must, in those
circumstances,
on
the probabilities,
be possessed of information relevant to the
en
passant
misconduct.
[26]
Hlebela
went
on to address the issue of negligent ignorance of relevant facts and
held that imputed or constructive knowledge of relevant
facts was not
the kind of circumstances capable of being encapsulated by the
concept of derivative misconduct. In
Hlebela,
at [10] – [11] it was held:
‘
The
undisclosed knowledge must be
actual,
not imputed or constructive knowledge,
of the wrongdoing. Proof of actual knowledge is likely to be
established by inferences from the evidence adduced but it remains

necessary
to prove actual knowledge
.
The moral blameworthiness intrinsic in the non-disclosure implies a
choice made not to tell, which is incompatible with actual
ignorance
of relevant facts as a result of incompetence or negligence. The
non-disclosure must be deliberate….’ (underlining

supplied)
[27]
This
exclusion is justified because the gravamen of derivative misconduct
is, on policy grounds, limited to deliberate non-disclosure.
The
exclusion of these mentioned circumstances ought not to be thought,
in any way, to inhibit the proof of actual knowledge being

established by inferences to be drawn from a matrix of proven facts
and relevant circumstances. What is in the head of any person
is,
logically, unknowable. In real life, proof of knowledge is
established by adducing evidence that, taken as a whole, results
in
proof that the person must have had the relevant knowledge, as the
most reasonable inference.
Prima
facie
proof of facts, from which such an inference can be fairly drawn,
which
remain unrebutted
,
results in proof on a balance of probabilities that the relevant
knowledge was indeed possessed by the employees.
[28]
Because
of the conclusions reached about the evidence adduced and the proper
inferences that are to be drawn therefrom, which are
dispositive of
the factual issues, it is unnecessary, in my view, that this judgment
explore, as did the court
a
quo,
the delicate issue of the obligations, if any, of an employee to
respond to an employer upon being interrogated about alleged or

actual misconduct of others and whether or not the concept of
derivative misconduct as explicated in
Hlebela
may in an appropriate case, be expanded or extended.
[18]
The
application of the concept of derivative misconduct to the facts
[29]
The
decision in
RSA
Geological Services v Grogan NO
[19]
alludes
at [93] to an
onus
on an employer to prove, on the probabilities, that the employees
“knew or must have known about the principal misconduct”.

We agree that this is an appropriate approach. Accordingly, once it
can be inferred from the evidence that the appellant employees

probably were present during the violence, that
onus
shall be satisfied, and absent a positive rebuttal, proof that the
employees “knew or must have known” who perpetrated

violence is established.
[30]
The
reason why the arbitrator held that the employees could not be
culpable on the basis of derivative misconduct was because they
had
not been identified
individually
as being present when violence occurred and therefore, so it was
reasoned, no obligation could arise to offer a response to the

employer’s demand for information about the real villains. In
line with this thinking, the arbitrator held that culpability
in
terms of derivative misconduct could only be triggered when the
employer discharges an
onus
to show that the employees
must
have
the relevant knowledge, and if in respect of each employee, their
presence during violence was not individually proven,
ergo
,
there was no reason to suppose they could have the relevant
knowledge, and thus culpability was unproven.
[20]
[31]
Gush
J in the Labour Court held that this approach was a material
misdirection because the arbitrator did not apply his mind to
the
proper inferences that could be drawn from the evidence as a whole,
which included the inference that the employees were indeed
present
during the perpetration of the violence. Thus, it was in this respect
that the Labour Court and the arbitrator differed
on the appropriate
factual
findings. The basis for the finding by the arbitrator that the second
category of strikers was fairly dismissed was premised on
the factual
finding that they were individually identified as present. That
distinction made by the arbitrator was therefore critical
to the
rationale of the award.
[21]
[32]
It
must be correct that the arbitrator indeed adopted too narrow an
approach to the evidence by requiring the individual identification

of each employee as being present as a
sine
qua non
for the employees falling into a category of employees implicated on
the basis of derivative misconduct. On the premise that presence
or
absence had to be established on a preponderance of probabilities, it
must follow that indirect evidence in the form of inferences
drawn
from the whole body of evidence was a necessary category of evidence
to assess. Inasmuch as it is plain that the arbitrator
did not do so,
as he viewed individual positive identification as the threshold,
Gush J
a
quo
was correct to hold that this was a decision to which a reasonable
arbitrator could not come,
if
a
proper inference could, indeed, be drawn that, on the probabilities,
the employees were present during some, or all, of the violence.
[33]
The
Labour court alluded to several aspects of the evidence to reach the
conclusion that there were indeed inferences to be drawn
from the
whole body of evidence that, on a balance of probabilities, supported
the conclusion that the appellant employees were
present during the
violence.
[34]
First,
Dunlop’s case was that all the dismissed employees were on
strike. This contention went unchallenged; indeed, the evidence
from
the employees accepted that they were all on strike. Accordingly, the
question arises whether strikers would absent themselves
from the
picket line? Put differently, over the course of the month that the
strike endured, it may be asked whether it is at all
probable that
the appellant employees were each present on at least some occasions,
even if it be assumed that they might have
not all been present on
every day? Moreover, the very act of striking, being a collective
activity in which worker solidarity is
a critical dimension, it may
be asked, how likely would it be that strikers would absent
themselves from the demonstrations of
resolve and solidarity which
are the very fibre of strike culture? On this aspect, the employees
chose to be silent. Thus, there
is nothing in the evidence to gainsay
the inference that, on the probabilities, they each were present for
at least some of the
time, and equally probable, that they were each
present most of the time, even if not everyone was religiously
present on each
and every day.
[22]
From
these circumstances, the inference can be drawn that it is improbable
that each and every one of them could not have acquired
actual
knowledge of the misconduct perpetrated, more especially because the
misconduct, as described earlier, was so spectacular.
[35]
Second,
the case advanced on behalf of all the employees was that no violence
occurred, or if it had occurred they were ignorant
of it. This
version was proven to be a palpable lie. That finding is pertinent to
the issue of the inferences to be drawn in respect
of those persons
not individually identified as being present. 37 of their
co-employees were positively identified as perpetrating
violence in
the presence of a throng of employees.
[23]
The
appellant employees, along with all other employees, made common
cause with this lie. An inference adverse was justified.
[36]
Third,
the opportunity both at the appeal hearing (which they spurned, and
at the arbitration hearing where they remained silent)
to distance
themselves from the violence by claiming they were absent, or were
innocent bystanders and, if they could, identify
the culprits was
deliberately not seized.
[37]
It
was argued that it was not the case of Dunlop that the appellant
employees were present during the violence; ie their culpability

would rest simply on their participation in the strike and their
refusal to speak up. Ignoring for the purposes of analysis that
it
was not a defence advanced by the employees that they were absent
during the violence, the contention is at cross purposes with
the
issues that arise for decision. The rationale of the award was that
proof of presence dictated whether an employee was fairly
dismissed
or not; employees in category 2 were held to fairly dismissed on that
premise and have not appealed that outcome and
employees in category
3 who were found guiltless are the subject of this appeal. This is
the crux of the controversy before the
appeal court. The Court
a
quo
addressed that key finding in the award and overturned it. The issue
of the presence of the appellant employees was, at worst for
Dunlop,
always latent in the dispute, even if not articulated in as many
words. The consideration that some employees may not have
been
present on every occasion violence was committed was not left out of
account; however, no one ever advanced a case of absence,
preferring
instead, a cat and mouse approach to the adducing of evidence.
Without doubt, the issue of presence during the violence
was central
to the dispute, regardless of the exact articulation of the issues
before or during the proceedings.
[38]
Gush
J,
a
quo,
held:
[74]
‘By failing to consider whether or not the applicants evidence
created [an]inference that the respondent employees were
present or
under an obligation to exonerate themselves the third respondent does
not consider whether the failure to give evidence
or provide an
explanation was acceptable or whether such conduct constituted
derivative misconduct.
[75] In the
ABI
matter the court when considering the inference to be drawn said the
following:
The inference
which the respondent seeks to draw from the evidence is that all
the appellants were present at the time the
assault took place, and
either actively participated in the assault or at least supported and
encouraged the actual perpetrators.
It is a cardinal rule of logic
when reasoning by inference that the inference sought to be drawn
must be consistent with all the
proved facts. If it is not, the
inference cannot be drawn (
R v Blom
1939 AD 188
at
202-3). In my view all the evidence in the present case is consistent
with that inference.
The appellant's
counsel submitted that the evidence shows no more than that most of
the crewmen were present when the assault took
place, and that this
does not assist in establishing as a matter of probability
the presence of any one of the appellants.
The reasoning is
undoubtedly correct, but in my view, none of the evidence establishes
that it was only a majority of the appellants
who were present. The
evidence is equally consistent with them all being there.
The fact that
the evidence is consistent with the inference sought to be drawn
does not of course mean that it is necessarily
the correct inference.
A court must select that inference which is the more plausible or
natural one from those that present themselves
(
AA Onderlinge
Assuransie Assosiasie Bpk v De Beer
1982 (2) SA 603
(A)). In the
present case however no alternative inferences have been advanced
which have a foundation in the evidence. It
was suggested in
argument that one or more of the appellants may have been absent, or
may have been unwittingly caught up in the
events. This, however, is
no more than speculation, as there is no evidence to suggest that
this is what occurred. In my view this
is pre-eminently a case in
which, had one or more of the appellants had an innocent explanation,
they would have tendered it, and
in my view their failure to do so
must be weighed in the balance against them.
[76] I am
satisfied that the only reasonable and plausible inference that can
be drawn from the evidence is that the respondent
employees were
present during the strike and accordingly during the misconduct. If
they weren’t present or had no information
regarding the
perpetrators they would have said so. They, despite the opportunities
afforded them, did not.
[39]
I
agree. In summary, therefore:
39.1.
Proof
of the presence of the appellant employees during violence has been
proven on a balance of probabilities. The Labour Court
was correct to
find that the arbitrator acted unreasonably in failing to conclude
that the appellants were present at any of the
scenes of misconduct
and had actual knowledge of the misconduct and of the identity of any
of the perpetrators thereof.
39.2.
It
had been implicit in the employer’s case that the appellants
were present and had such knowledge. The absence of direct
evidence
to that effect seems to have persuaded the arbitrator to arrive at
his impugned conclusion. The arbitrator did not give
consideration to
the fact that such presence and knowledge was capable of proof by
means of indirect evidence, or by inference,
and, accordingly, did
not determine whether those facts had indeed been proven by
inference.
39.3.
Circumstantial
evidence relating to the appellants’ presence at the scenes of
misconduct and their knowledge of the misconduct
and/or any of its
perpetrators was placed before the arbitrator. Since it constituted
an important component of the evidential
material in the arbitration,
it was incumbent upon the arbitrator to consider whether to draw the
required inferences, by complying
with, well established, rules of
logic.
[24]
The
failure to do so was not reasonable.
39.4.
The
inference sought to be drawn in this case was whether the appellants
were present at any of the scenes, or incidents of misconduct,
but
more crucially, whether each of them had actual knowledge of any of
the misconduct, or of any of the perpetrators thereof.
All of the
appellants were on strike with the other workers. The inferences that
each of the appellants was present at some or
all of the incidents
where the misconduct occurred, and that they had actual knowledge of
such misconduct and/or of the perpetrator(s)
thereof, are consistent
with the proven facts and are the only plausible inferences that can
be drawn.
39.5.
There
was enough evidence, although not conclusive, that called for an
explanation. The false evidence tendered through the witnesses
called
by the Union, and the failure by the appellants to give evidence
themselves in those circumstances, are factors that could,

justifiably, be placed in the balance against them.
39.6.
A
reasonable arbitrator would not have found otherwise.
The
Sanction of Dismissal
[40]
Accordingly,
applying basic principles, as alluded to earlier, the conclusion must
be reached that the appellant employees’
breach of the duty of
good faith was serious enough to warrant dismissal and was wholly
appropriate in the circumstances. Gush
J articulated the position
thus, which warrants endorsement:
‘
[77]
It is entirely reasonable for an employer to expect protected
industrial action to be accompanied by orderly conduct by those

employees who have embarked on the industrial action. This is
particularly so in circumstances where the employer has not only

entered into a picketing rules agreement with the representative
trade union regulating the conduct of striking employees but has
as a
result of the conduct of the employees been forced to obtain an
interdict restraining the striking employees from committing

misconduct. That strikes are often visited with violence and
misconduct does not justify such acts’.
Conclusions
[41]
Accordingly:
41.1.
The
arbitrator erred in not assessing the evidence for inferences from
which, on the probabilities, the appellant employees were
shown to
have been present during the perpetration of violence;
41.2.
The
evidence supported an inference of their presence during violence;
41.3.
The
Labour Court was correct to conclude that the award ought to be set
aside;
41.4.
The
appellant employees breached their duty of good faith towards their
employer by failing to disclose the identity of the culprits.
41.5.
The
appeal must fail.
Costs
[42]
Both
parties seek costs. Accordingly, costs shall follow the result.
The
Order
(4)
The
appeal is dismissed with costs.
(5)
The
Order of the Labour Court is confirmed.
(6)
The
appellant shall bear the costs of the appeal.
_____________
Sutherland JA
Sutherland
JA (with whom Coppin JA concurs)
COPPIN JA
[43]
I agree my colleague Sutherland JA that the appeal must be dismissed,
but I respectfully do not agree with certain aspects
of the concept
of “derivative misconduct” espoused by the court
a
quo
in this matter, which commentators regard as a radical extension of
the concept.
[25]
They relate particularly to the duty to speak and the right to
silence. I consider this separate judgment to be necessary, lest
my
silence on those aspects is construed as concurrence.
[44] The
background facts have been elegantly narrated by Sutherland JA in his
judgment and the detail of it does not require repeating
here, save
for those facts, or aspects thereof, that require mention in the
course of this judgment.
[45] As pointed
out by Sutherland JA, the review before the Labour Court and
consequently, this appeal, only concern the fate of
those employees
who were not identified by direct evidence as being present in the
crowd at various incidents of misconduct that
marred the strike, and
whose reinstatement was ordered by the fifth respondent (the
arbitrator) in terms of his impugned award.
I shall refer to the
first to third respondents collectively as “the employer”
and by their individual names where
the context requires
it.
[46] The
appellants were dismissed for “derivative misconduct”
because of their failure to provide the employer with
information
concerning the identities of the perpetrators of the various acts of
violence, intimidation and harassment committed
in the course of what
would have been a lawful strike, which started on 22 August 2012 and
endured until about almost the end of
September 2012. The term
“derivative misconduct” might not be ideal, or truly
descriptive of the nature of the misconduct
referred to, but is used
in this judgment for convenience and for want, at this juncture, of a
distinctive and easily useable,
alternative term.
Decisiveness
of the circumstantial evidence
[47] This appeal
in my view is capable of being decided on the, relatively, simple
basis that the arbitrator had not concluded reasonably
that it had
not been proven by the employer that the appellants were present at
any of the scenes of misconduct, had actual knowledge
of the
misconduct and/or any of the perpetrators thereof, and had
deliberately withheld the information.
[48] It is
apparent from the arbitrator’s award that even though it had
been implicit in the employer’s case that the
appellants were
present and had such knowledge, the absence of direct evidence to
that effect seems to have persuaded the arbitrator
to arrive at his
impugned conclusion. The arbitrator did not give consideration to the
fact that such presence and knowledge was
capable of proof by means
of indirect evidence, or by inference, and, accordingly, did not
determine whether those facts had indeed
been proven by inference.
[49]
Circumstantial evidence relating to the appellants’ presence at
the scenes of misconduct, their knowledge of the misconduct
and/or
any of its perpetrators, and their failure to disclose what was in
their knowledge, was placed before the arbitrator. Since
it
constituted an important component of the evidential material in the
arbitration, it was incumbent upon the arbitrator to consider
whether
to draw the required inferences, by complying with, well established,
rules of logic.
[26]
The failure to do so was not reasonable.
[50] The
inferences sought to be drawn in this case was whether the appellants
were present at any of the scenes, or incidents of
misconduct, but
more crucially, whether each of them had actual knowledge of any of
the misconduct, or of any of the perpetrators
thereof, and if whether
their failure to disclose the information of which they had actual
knowledge, was deliberate.
[51] All of the
appellants were on strike with the other workers. There were several
incidents of misconduct that occurred over
the period of the strike.
The arbitrator found that the employers’ witnesses’
evidence “proved an overwhelming
balance of probabilities that
the acts of misconduct testified to, did in fact occur”. My
colleague, Sutherland JA, has listed
these acts of misconduct.
[52] The
employer requested the union, and the striking workers (which
included the appellants), through the union, to assist with
the
identification of the perpetrators of the various acts of misconduct
committed during the course of the strike. There was no
prompt
response to the request and, significantly, no denial that any of its
members were present when the acts of misconduct were
perpetrated, or
of their members’ knowledge of such wrongdoing, or of the
identity of any of the perpetrators thereof. Instead,
the Union, whom
its members, including the appellants, entrusted with their
interests, including their legal interests relating
to this matter,
denied that there were any acts of misconduct committed during the
course of the strike. The evidence presented
at the arbitration, to
that effect, by the union on behalf of its members, including the
appellants, was rejected as false, and
rightly so.
[53] The
inferences that each of the appellants were present at some or all of
the incidents where the misconduct occurred, that
they had actual
knowledge of such misconduct and/or of the perpetrator(s) thereof,
and that their failure to disclose such knowledge
was deliberate, are
consistent with the proven facts and are the only plausible
inferences that can be drawn. There was enough
evidence, although not
conclusive, that called for an explanation. The false evidence
tendered through the witnesses called by
the union, and the failure
by the appellants to give evidence themselves in those circumstances,
are factors that could, justifiably,
be placed in the balance against
them. A reasonable arbitrator would not have found otherwise.
[54] Culpability
for the derivative misconduct in question consisted of no more than
the following: there was a duty on appellants
as individuals to
speak, because they had actual knowledge of one, some, or all of the
incidents of misconduct and/or of the identity
of the perpetrators
thereof; their failure to exonerate themselves, by either disclosing
such knowledge to the employer, or raising
a defence that justified
the non-disclosure of such information, such as for example,
intimidation, or the fear of reprisals and
absence of any effective
protections against the same, was deliberate and therefore culpable.
In my view, this is consistent with
what was held by this Court in
Hlebela.
[55]
The version of the principle seemingly espoused by the court
a
quo
,
in a passage of its judgment
[27]
quoted by Grogan,
[28]
does not articulate that the duty to speak only arises once the
employee has actual knowledge of the primary wrongdoing and/or
the
identity of the perpetrator(s) thereof, and that to be culpable, the
failure to disclose the knowledge must be deliberate,
as was held in
Hlebela,
and
creates the impression that the mere presence of an employee at a
scene where misconduct occurred triggered a duty for him to
exonerate
himself. Read in context, it is my view that it is clear that that is
not what the court
a
quo
meant, but if I am wrong, and that was indeed what the court
a
quo
held, then it was wrong.
[56] To provide
context to my views on the aspects I do not agree with, it is
necessary to briefly trace the development of the
principle of
derivative misconduct.
Brief history
of the development of the principle
[57]
In
FAWU
,
in the course of considering what weight if any ought to be given to
the failure of the appellants in that case to give evidence,
Nugent
J, having referred to the trite approach in civil proceedings stated:
“in the field of industrial relations, it may
be that policy
considerations require more of an employee then that he merely
remained passive in circumstances like the present,
and that his
failure to assist in an investigation of this sort may in itself
justify disciplinary action
”
.
Nugent
J pointed out that even though this had been an issue raised in the
hearing before the Industrial Court, it was not necessary
to deal
with it in view of the conclusion to which he had come, namely, that
the employees’ complicity and involvement in
the assault had
been proved by inference. The quoted
dictum
,
however, contained the seed which eventually blossomed into the
notion of the principle of “derivative misconduct”
that
is the subject of this discussion.
[58]
In
Chauke,
Cameron
JA briefly elaborated on the nature of and rationale for the
principle, but did not find it necessary to apply it in light
of the
court’s view in that matter, namely, that the shared
responsibility for the primary misconduct was properly inferred
by
the employer and that the dismissals were, accordingly, justified.
Whereas, it was merely stated in
FAWU
that there may be “policy considerations” in the field of
industrial relations that required more of an employee than
to remain
passive in circumstances as dealt with there, in
Chauke
,
where a number of workers had been charged with malicious damage to
property, and had, in the absence of direct evidence that
they were
involved, been found guilty of it, Cameron JA elaborated on the
principle, stating:
“…
two
lines of justification for a fair dismissal may be postulated.
The
first is that a worker in the group which includes the perpetrators
may be under a duty to assist management in bringing the
guilty to
book. Where a worker has or may reasonably be supposed to have
information concerning the guilty, his or her failure
to come forward
with information may itself amount to misconduct. The relationship
between employer and employee is in its essentials
one of trust and
confidence, and, even at common law, conduct clearly inconsistent
with that essential warranted termination of
employment (
Council
for Scientific & Industrial Research v Fijen
(1996)
17 ILJ 18 (A) at 26 D-E). Failure to assist an employer in bringing
the guilty to book violates this duty and may itself
justify
dismissal.”
[29]
.
[59]
Cameron JA went on to explain that: “this approach involves a
derived justification, stemming from an employee’s
failure to
offer reasonable assistance in the detection of those actually
responsible for the misconduct. Though the dismissal
is designed to
target the perpetrators of the original misconduct, the justification
is wide enough to encompass those innocent
of it, but who through
their silence make themselves guilty of a derivative violation of
trust and confidence
.”
[30]
.
[60]
In
National
Union of Mineworkers and Others v J Grogan NO and Another,
[31]
(NUM)
this
Court had another opportunity to consider the nature and application
of the concept. It followed and endorsed what had been
said about the
concept in
FAWU
and
in
Chauke
.
McCall AJA, writing for the court, made it clear that “it must
be accepted that an employee may, in appropriate circumstances,
have
a duty to assist management by bringing forward information within
his knowledge about the wrongdoing of other employees….’.
There
the court also appears to have endorsed the view (of the arbitrator
in that matter) that the employer had to prove on a balance
of
probability that each employee was in possession of information that
could have assisted the employer in its investigations.
[32]
[61] In
Foschini
, this Court referred to what was stated in
Chauke
concerning the principle and seemingly concluded in light of it,
although as an alternative, that the employees in that case had
been
fairly dismissed. There the entire team of five employees was found
guilty on a charge of gross negligence. It was alleged
that they
collectively failed to prevent stock losses resulting in financial
loss for the employer and causing an irretrievable
breakdown in the
trust relationship between them and the employer. The employees had
represented themselves at the disciplinary
enquiry and their
explanation for the stock losses had been rejected. The court
confirmed that the dismissals of the employees
were justified on the
basis of collective misconduct. But the court did not say anything
more about the principle than what was
stated in
Chauke
.
[62] In
Hlebela,
this Court put more flesh on what, until then, had largely been
obiter
views concerning the principle. This Court found it
necessary to clarify important aspects of those
dicta.
Sutherland JA, writing for this Court, stated:
‘
[8]
Several important aspects of these
dicta
require clarification. Important to appreciate is that no new
category of misconduct was created by judicial fiat. The effect of

these
dicta
is to elucidate the principle that an employee bound implicitly by a
duty of good faith towards the employer breaches that duty
by
remaining silent about knowledge possessed by the employee regarding
the business interests of the employer being properly undermined.

Uncontroversially, and on general principle, a breach of the duty of
good faith can justify a dismissal. Non-disclosure of knowledge

relevant to misconduct committed by fellow employees is an instance
of a breach of duty of good faith. Importantly, the critical
point
made by both FAWU v ABI and Leeson Motors is that a
dismissal
of
an employee is
derivatively
justified in
relation to
the
primary misconduct
committed by unknown others, where an employee, innocent of actual
perpetration of misconduct consciously chooses not to disclose

information known to that employee pertinent to the wrongdoing.
[9]
…
[10]
The undisclosed knowledge must be actual, not imputed destructive
knowledge of wrongdoing. Proof
of actual knowledge is likely to be
established by inferences from the evidence adduced but it remains
necessary to prove actual
knowledge. The moral blameworthiness
intrinsic in the non–disclosure implies a choice made not to
tell, which is incompatible
with actual ignorance of relevant facts
as a result of incompetence or negligence.
[11]
The non-disclosure must be deliberate. In my view, this too, follows
logically from the value
choices intrinsic in the concept of a duty
of good faith.
[12]
More problematically, whilst the duty to disclose is uncompromised by
the degree of seriousness
of the wrongdoing, i.e. it ought to apply
to late-coming as much as to theft, in my view, whether, in a given
case, the non-disclosure
warrants dismissal would be related, in
part, to the degree of seriousness of the wrongdoing and to the
effect of non-–disclosure
by a person in the position of that
employee on the ability of the employer to protect itself against the
given wrongdoing. . .
.
[13]
…
[14]
…the disclosure of information relevant to the wrongdoing,
pursuant to the duty of good
faith, ought not be dependent upon a
specific request for relevant information; often the wrongdoing per
se might not be known
to the employer. Mere actual knowledge by an
employee should trigger a duty to disclose. Where a request for
information about
known wrongdoing or suspected wrongdoings has
indeed been made, culpability for the non-disclosure is simply
aggravated.
[15]
Furthermore, the anterior premise of these considerations is that an
employee is a witness to
wrongdoing, not a perpetrator. The
misconduct lies within the bosom of a general duty of good faith to
rat on the wrongdoers, not
on culpable participation, even in a
lesser degree than other perpetrators. The employee is thus not a
person who has made common
cause with the perpetrators. A
disinclination to disclose the wrongdoing from a sentiment of worker
solidarity or some other subjective
sentiment falling short of common
purpose is likely to be a typical explanation for non-–
disclosure, but is per se not a
defence to a charge of a breach of
duty of good faith. . . .’
[33]
.
[63]
It is noteworthy that in
Hlebela,
this Court emphasised that “actual knowledge is required to
trigger the duty to speak up, the employer must prove actual

knowledge not merely putative knowledge, and no rule exists for
considerations of negligent ignorance”
.
[34]
A
view was also expressed as to how to deal with an employee with
actual knowledge of the main wrongdoing. Sutherland JA stated:
“An
appropriate way to discipline an employee with actual knowledge of
the wrongdoing of others or who has actual knowledge
of information
which the employee subjectively knows is relevant to unlawful conduct
against the employer’s interests would
be to charge the
employee with a material breach of the duty of good faith,
particularising the knowledge allegedly possessed and
alleging a
culpable non-disclosure. This observation does not mean that the
gravamen of such a charge might not also be articulated
in another
way, provided it is plain what is alleged and why it is alleged to be
culpable.”
[35]
[64]
It
is also noteworthy that in
PRASA,
[36]
in a decision handed down by this Court after we had heard argument
in this matter, the elements of this concept of misconduct,
as
explained in
Hlebela
and
in
NUM,
were confirmed.
[37]
Alleged
Expanded version of the principle
[65]
Poppesquo
[38]
refers to certain
dicta
in the court
a
quo’
s
judgment that is said to suggest that all employees, even those who
were not present at any of the scenes of violence, had a duty
to come
forward to exonerate themselves by providing an explanation. That
view is wrong for the reasons I have already mentioned
earlier. Those
remarks were made by the court
a
quo
in the course of its discussion of the duty of good faith and the
right to remain silent. The court
a
quo
stated,
inter
alia
,
that the “right to remain silent is sacrosanct in criminal
matters where accused persons are presumed to be innocent until
found
guilty. This is not a criminal investigation and the presumption of
innocence does not apply”. In my view, this statement
was
unnecessary, and is in fact wrong.
[66] Such
notions negate what was held in
Hlebela,
namely, that the duty
to speak is triggered by actual knowledge of those facts. This new
notion of the principle also implies that
an employee, who is
otherwise totally ignorant of the misconduct and identity of any of
the perpetrators, but who does not speak
or exonerate himself, or
herself, may be disciplined for some form of unspecified misconduct.
Further, according to this notion,
there is no place in labour
relations for the right to remain silent, and, by implication, for
safeguards, similar to those found
in the criminal law to protect the
liberty of accused, arrested and detained persons. The notion,
seemingly, draws no distinction
between the privilege of
self-incrimination and the right to remain silent that is essential
to give effect to that privilege.
[67] This
expanded notion implies the employer may question any employee whom
it, in effect (based on its subjective suspicions)
chooses to
question, and it is for that employee to exonerate himself or
herself. That would include, establishing that the employer
had no
reasonable basis for assuming that the employee had actual knowledge
of the principal wrongdoing and/or the perpetrator(s)
of it. This, in
my respectful view, besides (effectively) imposing a duty on an
employee to establish his or her innocence, overlooks
or discards
certain fundamental rights of employees, including the right to be
deemed innocent of any wrongdoing. Since it is generally
provided in
disciplinary codes, consistent with the (generally) adversarial
nature of disciplinary proceedings, that the employer
bears the
onus
to prove the misconduct alleged, completely denying the employee the
right to silence and the privilege against self-incrimination,
seems
to be inconsistent with the ethos the Labour Relations Act 66 1995
(LRA) seeks to promote, and to be unfair. An employee
should be able
to freely exercise his or her (hopefully informed) choice concerning
the answering of questions, even though the
choice made might be
detrimental to him, or her.
[68]
While one appreciates that the employer must at least be able to
invite an employee to disclose his or her actual knowledge
(if any)
of misconduct, and warn the employee of the consequences of refusing
to do so, the absence of rules regulating more extensive
questioning
by the employer leaves ample room for abuse. The very notion that an
employee can be sanctioned for not speaking, irrespective
of whether
he or she has actual knowledge of the principal misconduct, or the
identity of any of its perpetrators, is in itself
potentially
tyrannical. The protections in criminal law, which include the right
to silence and the privilege against self-incrimination,
were
intended to protect citizens from unfair police and judicial
interrogation.
[39]
Similar protections would accordingly not be out of place in labour
relations where potential tyranny by the police, State and
the courts
is replaced with potential tyranny at the hands of employers.
[69]
The right to silence and the privilege against self-incrimination
are, in any event, not protections that are confined to the
criminal
law. Section 14 of the Civil Proceedings Evidence Act (CPEA), read
with section 42 of that Act, not only recognises and
seeks to protect
witnesses in civil cases against self- incrimination, but has the
effect of giving that privilege a wider ambit
than in criminal
cases.
[40]
Section 14 provides: “a witness may not refuse to answer a
question relevant to the issue, the answering of which has no

tendency to incriminate himself, or to expose him to penalty or
forfeiture of any nature whatsoever, by reason only or on the sole

ground that the answering of such question may establish what tend to
establish that he owes a debt or is otherwise subject to
a civil
suit”. Notably, in the area of labour relations, it is
recognised that where the conduct of an employee amounts to
a
criminal offence, the employee may assert a right of silence (i.e.
the privilege against self-incrimination) during the internal
hearing
and request that those proceedings be postponed pending the
conclusion of the criminal proceedings, although the employer
is not
obliged to comply with that request.
[41]
Therefore, I do not agree with a blanket statement that there is no
place for the right to silence in labour relations.
[70]
Until
Hlebela,
the
principle of derivative misconduct was merely referred to in passing,
or otherwise not in much detail. It was first briefly
mentioned,
albeit not by the label “derivative misconduct”, in
FAWU.
In
Chauke,
this Court seemingly elaborated on the rule mentioned in
FAWU
,
but once again did not find it necessary to apply it to the facts in
that case, and consequently, no certainty was provided regarding
its
ambit. In
Foschini,
this Court applied the rule merely on the limited basis it was
mentioned in
Chauke,
but
added nothing further to its understanding. In
NUM,
roughly the same occured, although the court added singular
highlights, as mentioned earlier. In
Hlebela,
which was only decided in 2015, this Court found it necessary to
elucidate and further explain this form of misconduct. Most recently,

in
PRASA,
[42]
this
Court confirmed what was held in
Hlebela.
[71]
Even though the principle of derivative misconduct was (arguably) not
created by judicial
fiat
,
as pointed out in
Hlebela,
[43]
employers would be well advised to clearly define this form of
misconduct in their disciplinary codes, so that all interested
parties, including the trade unions and employees, know with
reasonable certainty what is expected of them and what conduct would

be culpable and subjected to discipline.
Conclusion
[72]
For the reasons mentioned at the outset of this judgment, I agree
that the appeal be dismissed with costs.
__________________
P Coppin
Judge of Appeal
SAVAGE AJA
Introduction
[73]
I
have had the benefit of reading the judgment of my colleague,
Sutherland JA, with which I am regretfully unable to agree. In my

view, the appeal should succeed with costs.
[74]
This
matter raises, once again, the difficult problem of strike violence,
appropriate responses to serious misconduct committed
during strike
action and the concept of “derivative misconduct”.
The
appeal is only concerned with those employees who were dismissed for
“derivative misconduct” on the basis that they
had
knowledge of acts of violence, intimidation and harassment but acted
in breach of the trust relationship with the employer
by failing to
come forward and identify the perpetrators, and failed to exonerate
themselves by explaining that they were not present
or could not
identify the perpetrators.
[75]
The
knowledge attributed to the appellant employees included, but was not
limited only to, acts of misconduct for which other employees
were
dismissed.
The
dismissal of those
employees who were identified as having been present when, or who
were shown to have committed serious acts of violence or intimidation

during the course of the strike action, is not in issue in this
appeal. Their dismissals were found at arbitration to be fair and

that finding has been left unchallenged.
Judgment
of the Labour Court
[76]
The
arbitrator, whilst rejecting the evidence of the witnesses led by the
union as lacking credibility and reliability, found that
the
dismissal of the employees for derivative misconduct had been unfair
since the employer had failed to discharge the
onus
to prove that the appellant employees had knowledge of the acts of
violence, intimidation and harassment.
[77]
Aggrieved
with that decision, the employer sought the review of the arbitration
award by the Labour Court. That Court set aside
the arbitrator’s
decision on the basis that, in respect of the employees dismissed for
derivative misconduct, it was one
that a reasonable arbitrator on the
evidence could not have reached. The Court noted there to be a clear
distinction between proving
on a balance of probabilities that the
employees knew who the perpetrators were and failed to disclose this
information and considering
whether the employees were under a duty,
consistent with the essential of trust and confidence to come forward
with an explanation
to the employer. Issue was taken with the
arbitrator’s failure to consider whether the evidence of the
employer’s witnesses
was sufficient to do more than simply
remain silent, with the Court finding that the arbitrator had failed
to consider whether
a reasonable inference could be drawn that the
employees dismissed for derivative misconduct were present during the
strike and
accordingly when strike misconduct occurred. If such an
inference could be drawn, then the arbitrator ought to have
considered
whether “
their
failure to come forward and provide either an explanation exonerating
themselves or providing the names of the perpetrators
constituted
derivative misconduct
”.
[78]
This
led the Labour Court to conclude that –
‘…
the
only reasonable and plausible inference that can be drawn from the
evidence is that the…employees were present during
the strike
and accordingly during the misconduct. If they weren’t present
or had no information regarding the perpetrators
they would have said
so. They, despite the opportunities afforded them, did not.’
Derivative
misconduct
[79]
My
colleague states that “derivative misconduct” cannot be
thought of as more than a label, a term of art to capture
a rather
complex idea. With reference to
Hlebela
,
[44]
he
states that the term creates “
no
new category of misconduct…by judicial fiat
”,
but that “(u)n
controversially,
and on general principle, a breach of the duty of good faith can
justify dismissal”
.
Nevertheless, in
Hlebela,
it was recognised that the concept has been “
elusive
”
given that “
serious
confusion existed among those responsible for instituting
disciplinary process about the concept and how to apply it
appropriately."
[45]
[80]
The
notion of a derived misconduct arising from an employee’s
non-disclosure to the employer in particular circumstances and
in the
context of the employment relationship is not isolated to
circumstances in which there has been collective misconduct alleged,

yet has arisen more often in that context. In
NUM
v Durban Roodepoort Deep Ltd,
[46]
it
was made clear that:
'The
concept of ''collective' guilt is wholly repugnant to our law and any
policy in terms of which all members of any group....
must bear
collective punishment for the wrongdoings of some of the members is
unacceptable to this court because it runs counter
to the tenets of
natural justice and is a violation of the well-known principle that a
person is presumed to be innocent until
proved guilty. There is a
failure of justice even if a single person is presumed to be guilty
and made to suffer with the rest.'
[81]
This
Court in
Chemical
Energy Paper Printing Wood and Allied Workers Union v National
Bargaining Council for the Chemical Industry and Others
,
[47]
made
it clear that:
‘
While
the principle is correct that all employees who have committed
misconduct must be treated similarly unless there is some
justification to treat them differently - in cases of collective
misconduct an employer can only act against those employees it
can
prove to have committed the misconduct complained of. An employer is
therefore obliged, in situations as obtained in this matter,
to
charge only those employees against whom it has evidence. If such
employees are found guilty the employer may impose an appropriate

penalty. An employer cannot, in matters such as this, simply dismiss
all of its striking employees because some from amongst them

committed serious misconduct. As a consequence, some employees who
commit serious misconduct may not be charged or when charged,
the
employer is unable to satisfy the disciplinary enquiry that each of
the employees who is charged is in fact guilty of the misconduct.

Hence, where there has been collective misconduct and the employer
only charges some of the employees because it only has evidence

against them and from amongst those charged some are found to have
committed the wrong and are dismissed and a few acquitted, it
does
not and cannot follow that the dismissal was unfair because of any
selective application of discipline.’
[82]
Within
this framework, a consideration of the judgments which have
considered the scope and application of “derivative misconduct”

on a particular set of facts, illustrates the difficulties which have
arisen with the concept and its application. Importantly,
those
decisions which have been used as the basis on which to develop or
attempt to clarify the concept (upgraded to a “principle”

in some decisions) either have not found derivative misconduct to
exist on their particular facts or have expressed
obiter
views on the issue. Derivative misconduct was not the basis for the
dismissals in
FAWU
,
Chauke
or
Foschini
.
In
NUM,
Hlebela
and PRASA
)
no
derivative misconduct was found to exist and in each case dismissal
on such basis was found to be unfair.
[83]
In
FAWU,
this Court raised, but did not determine, whether an employee’s
failure to assist an employer in an investigation and remain
passive
may justify disciplinary action. The Industrial Court found that
although the identities of the employees who participated
directly in
an assault of a fellow employee were not known, the dismissed
employees had acted with common purpose. On appeal, it
was accepted
that the
onus
of establishing that the appellant employees had associated
themselves with the assault rested on the employer. This Court found

that the evidence was consistent with the inference that all the
employees present when the assault took place either participated
in
the assault or lent it their support; and that it was probable that
all the appellant employees were present and, consequently,
had
either participated in the assault or lent their support to it.
Having regard to the extent to which a party’s failure
to give
evidence may properly give rise to an inference against him (or her),
it was stated that:
‘
What
emerges from the decided cases is that his failure to do so cannot by
itself constitute proof of what is alleged against him.
Nevertheless,
the evidence against him, though not conclusive, may be such than an
explanation would be expected if one was available.
In such cases his
failure to provide an explanation may be placed in the balance
against him. The approach in civil cases is illustrated
by the
following extract from the judgment of Schreiner JA in
Galante
v Dickson
1950(2) SA 460 (A) at 465:
‘
It
is not advisable to set down any general rule as to the effect that
may properly be given to the failure of a party to give evidence
on
matters that are unquestionably within his knowledge. But it seems
fair at all events to say that in an accident case where
the
defendant was himself the driver of the vehicle the driving of which
the plaintiff alleges was negligent and caused the accident,
the
court is entitled, in the absence of evidence from the defendant, to
select out of two alternative explanations of the cause
of the
accident which are more or less equally open on the evidence, that
one which favours the plaintiff as opposed to the defendant.’
[48]
[Own underlying]
[84]
Although
the matter was not determined on the basis of derivative misconduct,
the Court stated
obiter
that:
‘
In
the field of industrial relations, it may be that policy
considerations require more of an employee than that he merely remain

passive in circumstances like the present, and that his failure to
assist in an investigation of this sort may in itself justify

disciplinary action. This was an issue which was raised in the court
a
quo
,
but in view of the conclusion to which I have come it is not
necessary to deal with it in the present case
.
’
[49]
[85]
In
the subsequent decision of
Chauke,
this
Court found that, despite the absence of direct evidence of
involvement, the shared responsibility of paint shop employees
for
the primary misconduct of malicious damage to property was properly
inferred by the employer and that the dismissals were justified.
This
followed a request to the staff to divulge information regarding the
sabotage, which request drew no response. An ultimatum
was then
issued that any further sabotage in respect of which individual
perpetrators remained unidentified would result in the
dismissal of
all employees.
The
employer thereafter dealt with the matter on the basis of common
purpose and not on application of the concept of derivative

misconduct. While in
FAWU
reference was made to the “policy considerations” in the
field of industrial relations that may require more of an
employee
than to remain passive in circumstances as dealt with there, in
Chauke
it was stated:
‘
In
the second category, two lines of justification for a fair dismissal
may be postulated. The first is that a worker in the group
which
includes the perpetrators may be under a duty to assist management in
bringing the guilty to book. Where a worker has or
may reasonably be
supposed to have information concerning the guilty, his or her
failure to come forward with information may itself
amount to
misconduct. The relationship between employer and employee is in its
essentials one of trust and confidence, and, even
at common law,
conduct clearly inconsistent with that essential warranted
termination of employment (
Council
for Scientific & Industrial Research v Fijen
(1996) 17 ILJ 18 (A) at 26 D-E). Failure to assist an employer in
bringing the guilty to book violates this duty and may itself
justify
dismissal’.
[50]
[86]
The
Court continued that:
‘…
this
approach involves a derived justification, stemming from an
employee’s failure to offer reasonable assistance in the

detection of those actually responsible for the misconduct. Though
the dismissal is designed to target the perpetrators of the
original
misconduct, the justification is wide enough to encompass those
innocent of it, but who through their silence make themselves
guilty
of a derivative violation of trust and confidence’
[51]
.
[87]
In
NUM,
the
employer found
inter
alia
that
kimberlite
had been poured down boreholes and interviewed 15 employees who
denied any knowledge of the misconduct. All employees
were charged
with misconduct on the basis that they had acted with common purpose,
alternatively that they had failed to disclose
information about the
misconduct of fellow employees relating to the matter.
The
arbitrator took the view that proof of derivative misconduct was
subject to it being shown that
the
employee knew or could have acquired knowledge of the wrongdoing; and
that the employee failed without justification to disclose
that
knowledge to the employer, or to take reasonable steps to assist the
employer to acquire that knowledge.
On
the facts, it was f
ound
that it was possible that the employees had information relating to
the issue but that the employer’s evidence did not
discharge
the
onus
upon
it to prove its case on a balance of probabilities, with the refusal
to undergo polygraph tests and an un-cooperative attitude
not proving
that they had something to hide.
[88]
The
Labour Court endorsed the arbitrator’s view that the employer
had to prove that each employee was in possession of information
that
could have assisted the employer in its investigations while
accepting that in appropriate circumstances an employee may have
a
duty to assist management by providing information about workplace
wrongdoing. On appeal, this Court accepted that an employee
may have
such a duty in appropriate circumstances but that in the absence of
prima
facie
evidence of such knowledge it could not be concluded that the
employees’ failure to co-operate necessarily meant that they

either did have or must have had something to hide. On the issue as
to whether an adverse inference could be drawn against the
employees
from their failure to give evidence, with reference to
FAWU
,
it was found that this was an issue to be “placed in the
balance against” the employees. This Court found that the

arbitrator’s finding that the employer had not discharged the
onus
fell within the ambit of reasonableness required.
[89]
In
Foschini,
an
entire team of five employees was dismissed for gross negligence due
to their collective failure to prevent stock losses. On
appeal, this
Court considered the issue of collective accountability for stock
losses with reference to
Federal
Council Retail and Allied Workers v Snip Trading
[52]
(
Snip
Trading
)
and SA
Commercial
Catering and Allied Workers Union v Pep Stores
[53]
(Pep
Stores).
In
Snip
Trading,
the arbitrator accepted that collective guilt is repugnant to the
principles of natural justice unless responsibility for the
collective conduct of the group is indivisible. Dismissal was
warranted because, as individual components of the group, each had

culpably failed to protect the interests of the employer and ensure
that the group complies with a rule or attains a performance
standard
set by the employer to protect its assets; each member had been given
an opportunity to explain the team's failure; and
the person to whom
the explanations are given must be objectively satisfied that the
team's failure cannot be blamed on any particular
member of that
team. It was stated that it is therefore unnecessary to prove
individual culpability, “derivative misconduct”
or common
purpose, the three grounds on which dismissal for collective
misconduct can otherwise be justified. With no issue taken
by the
employees with the legal principles,
this
Court accepted that the employees had colluded to keep the stock
losses from their employer and gave unacceptable explanations
for
such losses.  This resulted in an irretrievable breakdown in the
trust and dismissal was appropriate.
[90]
In
Hlebela
the employee was dismissed for failing to disclose information
concerning his personal financial affairs, on request, to his
employer.
On appeal, this Court found the dismissal to have been
unfair in that the employee had not committed misconduct through his
non-disclosure.
T
he
Court elaborated on what, until then, had largely been
obiter
views concerning what constituted derivative misconduct:
‘
[8]
Several important aspects of these
dicta
require clarification. Important to appreciate is that no new
category of misconduct was created by judicial fiat. The effect these
dicta
is
to elucidate the principle that an employee bound implicitly by a
duty of good faith towards the employer breaches that duty
by
remaining silent about knowledge possessed by the employee regarding
the business interests of the employer being properly undermined.

Uncontroversially, and on general principle, a breach of the duty of
good faith can justify a dismissal. Non-disclosure of knowledge

relevant to misconduct committed by fellow employees is an instance
of a breach of duty of good faith. Importantly, the critical
point
made by both
FAWU v ABI
and
Leeson Motors
is that a dismissal of an employee is derivatively justified in
relation to the primary misconduct committed by unknown others,
where
an employee, innocent of actual perpetration of misconduct
consciously chooses not to disclose information known to that

employee pertinent to the wrongdoing.’
[91]
The
Court set out a number of considerations relevant to a determination
of derivative misconduct. The employee must have “
actual,
not imputed destructive knowledge of wrongdoing”
,
otherwise blameworthiness cannot be attributed to him or her. The
Court stated that proof of actual knowledge –
‘…
is
likely to be established by inferences from the evidence adduced but
it remains necessary to prove actual knowledge. The moral

blameworthiness intrinsic in the non–disclosure implies a
choice made not to tell, which is incompatible with actual ignorance

of relevant facts as a result of incompetence or negligence.’
[54]
[92]
Furthermore,
the non-disclosure must be deliberate and the gravity of the
non-disclosure must be proportionate to the gravity of
the primary
misconduct: it is related “in part, to the degree of
seriousness of the wrongdoing and to the effect of non-disclosure
by
a person in the position of that employee on the ability of the
employer to protect itself against the given wrongdoing”.
[55]
The
rank of the employee may affect the gravity of the non-disclosure.
However, it -
‘
ought
not be dependent upon a specific request for relevant information;
often the wrongdoing per se might not be known to the employer.
Mere
actual knowledge by an employee should trigger a duty to disclose.
Where a request for information about known wrongdoing
or suspected
wrongdoings has indeed been made, culpability for the non-disclosure
is simply aggravated...Furthermore, the anterior
premise of these
considerations is that an employee is a witness to wrongdoing, not a
perpetrator. The misconduct lies within the
bosom of a general duty
of good faith to rat on the wrongdoers, not on culpable
participation, even in a lesser degree than other
perpetrators. The
employee is thus not a person who has made common cause with the
perpetrators. A disinclination to disclose the
wrongdoing from a
sentiment of worker solidarity or some other subjective sentiment
falling short of common purpose is likely to
be a typical explanation
for non-–disclosure, but is per se not a defence to a charge of
a breach of duty of good faith…
’
[56]
[93]
Issue
was taken with the "t
he
notion that breach of good faith occurs if an employee 'could have
acquired knowledge of wrongdoing'
"
as
stated in
RSA
Geological Services
,
which was considered to be "
too
broadly or loosely stated
."
[57]
The
Court emphasised that “. . . .
actual
knowledge is required to trigger the duty to speak up, the employer
must prove actual knowledge not merely putative knowledge,
and no
rule exists for considerations of negligent ignorance”
.
[58]
In
this regard, the Court continued that “
no
room exists for considerations of negligent ignorance
”
and that “
the
explanation for non-disclosure may afford, in a given case,
mitigation of the culpability, but it would not stretch to a defence

to the charge
”.
[59]
The
employee needs not to have made common purpose with the perpetrator
and cannot be guilty of derivative misconduct on the basis
of
negligently failing to take steps to acquire knowledge of the
wrongdoing.
[94]
It
was found that -
‘
Even
an unreasonable refusal to disclose the employee’s personal
finances and a reasonable inference that he did so to conceal
the
manner of their acquisition is not capable of being logically linked
to the fact that he has actual knowledge of wrongdoing
by others.
When the employer is thwarted by a non-disclosure to procure
information, it cannot be argued that the employer can
infer proof of
what it suspects
.
[60]
[95]
The
decision of this Court in
PRASA
[61]
was
handed down after argument had been heard in the current appeal.
PRASA
concerned
the fairness of the dismissal of over 700 employees, who were on a
protected strike. After a forensic fire investigation
concluded that
a train burning was “
probably
as a result of riot, strike and/or public disorder”
,
the employer gave the striking employees notice of their contemplated
dismissal following five incidents of train burning which
the
employer indicated it believed had “been carried out by
striking workers and/or persons acting in concert or association
with
striking workers…as a means of putting unlawful pressure to
PRASA Rail to accede to the unreasonable demands of NTM”.
The
employer indicated that despite all reasonable steps taken to
identify the responsible individuals it “has been unable
to do
so and cannot do so unless the striking employees identify the
culprits”. In addition, it gave employees notice that
it “holds
you and all striking employees jointly and severally responsible for
the torching of trains and train coaches and
intends dismissing you
for this reason, unless otherwise dissuaded by your representations”.
A collective response to the
letter was provided by the union in
which it was denied that its members had been responsible for the
train burnings and challenged
the employer’s belief that the
striking employees were responsible. The union distanced itself from
the unlawful conduct
and offered assistance to identify those
responsible. The employer found the representations “
unpersuasive
”
and summarily dismissed the employees.
[96]
The
union challenged the dismissals in the Labour Court, which found the
dismissals both procedurally and substantively fair. This
was so
inter
alia
given that the employees had breached their duty of good faith to the
employer by remaining silent about their actual knowledge
of the
misconduct, failed to dissociate themselves from the misconduct and
failed to help the employer identify the culprits. The
Labour Court
determined that the dismissals were justified on the grounds of
derivative misconduct as the employees had failed,
without
justification, to disclose their knowledge of the culprits, had
failed to provide an innocent explanation or dissociate
themselves
from the misconduct and had preferred general collective
representations which prevented the employer from deciding
who to
dismiss.
[97]
On
appeal, this Court reiterated that the
onus
rests on the employer to prove on a balance of probabilities that the
employee committed the alleged misconduct. It found it unlikely
on
the probabilities that the union had advocated the burning of train
coaches and the forensic report obtained by the employer
had not
concluded that the striking employees were responsible for setting
the train on fire, but rather that it was set on fire
“
probably
as a result of riot, strike and/or public disorder
”.
[62]
Furthermore,
no credible evidence had been presented that the employer “
had
taken reasonable steps to identify the culprits
”;
[63]
and
the employer had not called on the employees to disclose evidence
about the actual perpetrators of train burnings. In the
circumstances,
this Court found the Labour Court’s reliance on
derivative misconduct to be misplaced, with its essential elements
not proved.
[98]
This
Court stated that “
in
the decisions of Hlebela and
[ABI]
,
where the principle of derivative misconduct was applied
”,
the facts were materially distinguishable and that “(i)
n
both those cases, the employer succeeded in showing that all of the
dismissed employees had knowledge or ought to have had knowledge
of
the primary misconduct, owing to their physical presence at or near
the place, and time, of the occurrence of the primary
misconduct
”.
[64]
Factually,
this statement is not correct given that in
FAWU
“derivative misconduct” was not the basis for the
dismissal and in
Hlebela
the dismissal on that basis was found to be unfair.
[99]
The
Court in
PRASA
found
that the employer’s “
reliance
on principle of derivative misconduct
”
was misplaced and unjustified and that “(i)
n
essence, the striking employees were dismissed not for derivative
misconduct but rather for “collective misconduct”,
a
notion which is wholly repugnant to our law, not only because it runs
counter to the tenets of natural justice but also because
it is
incompatible with the established principle of innocent until proven
guilty
.”
[65]
The
dismissal of the employees was therefore found both procedurally and
substantively unfair.
[100]
With
regard had to the cases considered above and to the development of
the concept of “derivative misconduct”, I consider
it
both appropriate and in the interests of fairness not to encourage
further use of the term to name a species of misconduct which
is so
often plagued by a lack of clarity. Clarity is served by calling
things what they are in plain language, whether a breach
of a duty of
good faith, of trust or confidence or a failure to disclose material
information to the employer. It
em
4(1) of Schedule 8 to the LRA requires as much when it provides that
the
employee
be notified
of
allegations of misconduct “
using
a form and language that the
employee
can
reasonably understand
”
.
Notice to an employee of a complaint of, or dismissal for,
“derivative misconduct” fails, in my view, to achieve
this and I consider that use of the term should, therefore, be
discouraged.
Duty
to disclose information
[101]
We
live in
a complex society, still suffering from the devastating consequences
of a system of entrenched racial discrimination, plagued
by pervasive
and deep-rooted inequality and, very often, reflecting the deep
divisions which continue to exist between people.
The workplace
closely mirrors these challenges and complexities. What constitutes
sound policy within a labour relations context
must reflect
appropriate and careful regard to this broader context. Developing
our labour jurisprudence to include an expansive
duty upon on
employee to act in good faith or with trust and confidence towards
his or her employer, with a duty to “
rat”
,
as is suggested by this Court in
Hlebela
,
on fellow employees must therefore be a careful process, one which
ensures that there is appropriate regard to the context and
tensions
inherent in the contractual relationship between the employer and
employee, the position of the employee and the circumstances
and
conditions under which employees work and live.
[102]
The
record indicates the high level of fear experienced by the one
employee who did attend the disciplinary hearing in order to

exonerate herself. In the context of a violent and dysfunctional
strike, this clearly illustrated the extent of personal risk faced
by
the employee even after her dismissal, when she sought only to
exonerate herself and when her attendance at the hearing was
not
aimed at the disclosure of any material information to the employer
concerning other employees. While I accept that the relationship

between employer and employee has as one of its essentials, trust and
confidence, and conduct clearly inconsistent with it may
warrant
termination of employment, in considering the scope of such duty, it
is as well to have regard to the fact that the emergence
of an
expansive duty on an employee to act in good faith towards the
employer has been the subject of some academic criticism in

reflecting the vulnerable position of an employee within an
inherently asymmetric contractual relation with the employer.
[66]
The
“
policy
considerations
”
referred to in
FAWU
which
require consideration in determining the
scope
of an employee’s duty
to
assist
an employer protect its legitimate interests must, therefore, in my
view, reflect appropriate regard for the position of
both parties in
the relationship. This would include an assessment of the appreciable
risks which may arise for an employee in
speaking out, in naming
perpetrators or for purposes of exoneration and the dangers inherent
which may arise in doing so.
[103]
In
Hlebela, this Court took the stance that an employee is “bound
implicitly by a duty of good faith towards the employer”
and
that he or she “breaches that duty by remaining silent about
knowledge possessed by the employee regarding the business
interests
of the employer being properly undermined”. It continued that
“(u)ncontroversially, and on general principle,
a breach of the
duty of good faith can justify a dismissal. Non-disclosure of
knowledge relevant to misconduct committed by fellow
employees is an
instance of a breach of duty of good faith.”
[104]
I
consider a less unequivocal recordal of such duty to be appropriate,
one which seeks to strike a balance between the rights and
duties of
both employers and employees in the workplace. While a harsh view may
be taken of an employee’s passivity and silence
when the
employer’s best interests could be advanced by disclosure, in
determining the fairness of a dismissal account must
be taken of all
relevant factors which include the risk of serious mortal or other
danger to the employee.
[67]
I
am consequently unable to align myself with the view expressed in
Hlebela
that “(a) disinclination to disclose the wrongdoing from a
sentiment of worker solidarity or some other subjective sentiment

falling short of common purpose is likely to be a typical explanation
for non-disclosure, but is per se not a defence to a charge
of a
breach of duty of good faith…’
[68]
.
In my view, this statement fails to encourage an appropriate regard
for the policy considerations which
FAWU
suggested
would require consideration in the development of this complex area
of our law.
Decisiveness
of the circumstantial evidence
[105]
It
is actual knowledge by an employee that in
Hlebela
has
been said to trigger a duty to disclose
.
I
turn now to consider whether the Labour Court was correct in finding
that the arbitrator had not concluded reasonably that it
had not been
proven by the employer that the appellants were present at any of the
scenes of misconduct, or that they had actual
knowledge of the
misconduct and/or any of the perpetrators thereof and thus were under
a duty to disclose the information sought
by the employer.
[106]
It
is apparent from the arbitration award that it was the absence of
direct evidence against the employees that led the arbitrator
to the
conclusion reached that their dismissals had been unfair. My
colleague takes the view that the arbitrator did not give
consideration to the fact that the presence of the employees and
their knowledge of the misconduct and/or its perpetrators was capable

of proof by means of indirect evidence, or by inference, and,
accordingly, did not determine whether those facts had been proven
by
inference. The inferences sought to be drawn were whether the
appellants were present at any of the scenes or incidents of
misconduct; and whether each of them had actual knowledge of any of
the misconduct, or of any of the perpetrators thereof.
[107]
The
facts relied upon by my colleague in order to draw the inferences
arrived at are that:
i.
all
of the appellant employees were on strike with the other employees;
ii.
there
were several incidents of misconduct that occurred over the period of
the strike, which the arbitrator found on the employer’s

evidence had proved on an overwhelming balance of probabilities;
iii.
the
employer requested the union, and the striking employees, including
the appellant employees, through the union, to assist with
the
identification of the perpetrators of the various acts of misconduct
committed during the course of the strike. There was no
prompt
response to the request and, significantly, no denial by the union
that any of its members were present when the acts of
misconduct were
perpetrated, or of their members’ knowledge of such wrongdoing,
or of the identity of any of the perpetrators
thereof and no later
attempt made to exonerate themselves; and
iv.
the
union, whom its members, including the appellant employees, entrusted
with their interests, including their legal interests
relating to
this matter, denied that there were any acts of misconduct committed
during the course of the strike; and the evidence
presented at the
arbitration, to that effect, by the union on behalf of its members,
including the appellants, was rejected by
the arbitrator as false,
and rightly so.
[108]
Given
these proved facts, my colleague asks “how likely would it be
that strikers would absent themselves from the demonstrations
of
resolve and solidarity which are the very fibre of strike culture?”
He takes the view that the inferences drawn that each
of the
appellant employees were present at some or all of the incidents
where the misconduct occurred, and that they had actual
knowledge of
such misconduct and/or of the perpetrator(s) thereof, are consistent
with the proved facts and are the only plausible
inferences that can
be drawn. Since there was enough evidence, although not conclusive,
that called for an explanation, the false
evidence tendered through
the witnesses called by the union, and the failure by the appellants
to give evidence themselves in those
circumstances, are factors that
could, justifiably, be placed in the balance against them.
Furthermore, “there is nothing
to gainsay the inference, on the
probabilities, that they were present for at least some of the time,
and equally probable, that
they were each present most of the time,
even if not everyone religiously present on each and every day”;
and “it is
improbable that each and every one of them could not
have acquired actual knowledge of the misconduct perpetrated, more
especially
because …[it]
was
so spectacular.
”
My colleague therefore considers that the arbitrator adopted too
narrow an approach to the evidence in requiring individual

identification of each employee present and that a reasonable
arbitrator would have found otherwise.
[109]
I
am not persuaded that this is so.
S192(2)
provides that if the existence of the
dismissal
is
established, the employer must prove that the
dismissal
is
fair.
The
view I take of the matter is that the arbitrator concluded reasonably
on the facts before him that it had not been proven by
the employer
that the appellants were present at any of the scenes of misconduct
and had actual knowledge of the misconduct and/or
any of the
perpetrators of it.
[110]
As
was stated in
Govan
v Skidmore,
[69]
the
Court
must
“…
.
by balancing probabilities select a conclusion which seems to be the
more natural, or plausible, conclusion from amongst several

conceivable ones, even though that conclusion be not the only
reasonable one”.
[70]
In
my mind the more plausible or natural inference to be drawn from
those that present themselves in this matter is not that the

employees were present at some or all of the incidents where the
misconduct occurred, and that they had actual knowledge of such

misconduct and/or of the perpetrator(s) thereof. This is all the more
so since a number of employees were dismissed for the misconduct

committed, having been identified committing it, and still others
were dismissed for being on the scene when misconduct was committed

by way of eyewitness testimony and photographic evidence.
Furthermore, the arbitrator found the dismissal of the two employees

who testified at arbitration for the union to have been fair given
their identification on the scene.
[111]
The
fact that the employees did not exonerate themselves, by either
disclosing any knowledge to the employer, or raising a defence
such
as intimidation, or the fear of reprisals and absence of any
effective protections against the same does not lead me to a

different conclusion; nor does it, in my mind, allow a finding in the
circumstances that the employees can as a result be inferred
to be
culpable.
[112]
Whilst
inferential reasoning may obviously be used in considering whether
misconduct has been proved, appropriate regard must be
had to the
cautions sounded by the Labour Court in
SATAWU
v Collet Armed Security Services CC:
[71]
‘
It
is, however, critical that even such derivative misconduct cannot
diminish the standard of proof an employer must still comply
with to
establish the existence of misconduct. The fact is that employees
that attract culpability must still be identified. To
illustrate this
with a simple example – assuming
an
employer has 100 employees and during a strike some employees
participated in unlawful behaviour. Does this now mean that all

employees, just because they are employed by the same employer and
may have participated in the strike, can now be held accountable
for
this misconduct by certain individuals on the basis of derivative
misconduct just because they are all employed by the same
employer
and participated in the same strike? Surely not. What if a particular
group of employees were not even present when the
unlawful behaviour
took place and never witnessed or was aware of the same?’
[113]
To
explain this away simply on the basis that the employees are required
to exonerate themselves in my mind has the effect of shifting
the
onus
.
The employer must prove that the employee had actual knowledge of the
information sought before there is any duty on the employee
to raise
a defence. On the evidence before the arbitrator, actual knowledge
could not, in my mind, be inferred on the facts before
the
arbitrator. While the trade union should properly have acted at all
relevant times in a manner which was both responsible and
displayed
leadership, it is patently apparent that it failed to do so.
Following the dishonest testimony of union witnesses, numerous

employees were nevertheless found on the evidence before the
arbitrator to have been fairly dismissed for the misconduct
committed.
[114]
The
dishonesty of the union witnesses did not however, in my mind, allow
an inference to be drawn that all employees charged with
misconduct
as a result of their silence held actual knowledge of misconduct and
were consequently culpable by virtue of such silence.
If this were
so, it raises the obvious question: what of those employees who were
on strike but chose not to be on the picket line
and knew nothing of
the misconduct committed; or those employees who were on the picket
line but did not witness strike misconduct?
I am not persuaded that
there was an obligation on those employees to testify individually to
exonerate themselves, whether at
the disciplinary hearing or the
arbitration hearing, in the manner suggested by the employer given
the burden which rested on the
employer to prove the existence of the
misconduct alleged and the fairness of their dismissals.
[115]
For
all of these reasons, I would find that the decision of the
arbitrator fell within the ambit of reasonableness required and
would
grant an order that the appeal succeeds. There is, in my mind, no
reason in law or fairness why costs should not follow the
result.
________________
Savage
AJA
APPEARANCES
:
FOR THE
APPELLANTS:
Adv P Schumann
Instructed by
Brett Purdon Attorneys
FOR THE
RESPONDENT:
Adv A J Dickson SC
Instructed by
Farrell Inc.
[1]
The arbitrator did not invoke the doctrine of common purpose,
although that concept had been alluded to in pleadings by Dunlop

along with the invocation of the idea of derivative misconduct.
(Cited in paragraphs 18-20 of this judgment).
[2]
Sidumo
and Another v Rustenburg Platinum Mines Ltd and Others
[2007] 12 BLLR 1097 (CC).
[3]
Paragraph
4(4) schedule 8 to LRA; read also with paragraph 6. Also, the
arbitrator held, at [143] of the Award, that the failure
to hold a
pre-dismissal hearing was justified, a finding that has gone
unchallenged on appeal.
[4]
Appeal
Record:
Vol
24, p2434.
[5]
The 78 employees comprised both category 2 and category 3. It was in
the award that the two categories were distinguished by
reference to
whether the employees had been individually identified as being
present during the violence, on the facts as found
by the
arbitrator.
[6]
John
Grogan
analysed the case law on this topic in “
Derivative
misconduct” Not a trap to snare the (apparently) innocent
2015
(31) Employment Law journal and again in “
Derivative
misconduct- drawing in the net”
2018 (34) Employment Law, including dealing with the judgment
a
quo.
Tarryn
Poppesqou
has addressed the topic and the judgment of the court
a
quo
in
“
The
sounds of silence and the concept of derivative misconduct and the
role of inferences
”
2018 (39) ILJ 34-50.
[7]
(
1998)
19 ILJ 1441 (LAC).
[8]
In
Chauke
the issue was the pattern of continual sabotage to vehicles in the
employer’s paint shop. Ultimately, the court held the

employees culpable without resorting to the concept of derivative
misconduct.
[9]
(2010) 31 ILJ 1787 (LAC).
[10]
The case involved a stock shrinkage of 28% in a branch workshop
which had only five employees. Tolerable shrinkage was 2%. They
were
all dismissed for “team misconduct”. Despite the
opportunity none of the five offered any testimony to rebut
the
accusations of complicity in the theft. (see paragraphs 41– 49
of that judgment on the approach to proof of their culpability).
[11]
(2015) 36 ILJ 2280 (LAC).
[12]
In
Hlebela
,
the issue was whether the employee was withholding information about
the theft of platinum ores which the employer was engaged
in
refining, and how it might be proven that he had such knowledge. In
National
Transport Movement v M Tau and Others
(2018)
39 ILJ 560 (LAC
),
a decision heard before the hearing of this appeal and in respect of
which judgment was delivered afterwards, Kathree- Setiloane
AJA
dealt with the topic of derivative misconduct, and endorsed
Hlebela.
In
that case, it was found that the employer had failed to make out a
case to invoke derivative misconduct to justify a dismissal
and had
rather conceived its case on the basis of collective misconduct, an
illegitimate premise.
[13]
The question arises whether the label “derivative misconduct”
is one that deserves to be retained, or whether another
formulation
might be attempted. Notwithstanding the temptation to theorise the
proposition and clad it in fresh linguistic garments,
in my view, no
real usefulness can come of that exercise, and I am content that it
lives on as a term of art. My colleague, Savage
AJA is more critical
of the use of the term; See para99 of her judgment.
[14]
This species of misconduct is not to be conflated with criminal
doctrine of common purpose which addresses the notion that persons

who identify with a course of action by their acts of association
with its perpetration are equally guilty of the deed perpetrated.
[15]
The significance of the point that persons guilty of derivative
misconduct are not ‘actual perpetrators’ must be

emphasised. The breach of the employment relationship is rooted not
in the act itself but in the failure to render support to
the
employer.
[16]
[1994] 12 BLLR 25 (LAC).
[17]
See
Chauke
at para 31.
[18]
See:
Grogan (2018) Supra, and Poppesqou, supra,
at 44. In both articles the writers deal with the notion of an
extension being evidenced in the judgment
a
quo.
[19]
(2008) 20 ILJ 406 (LC).
[20]
The award at paras 78, 79 and 81.
[21]
Judgment
a
quo
at paras 55 and 61.
[22]
See the
dictum
of Nugent JA in
Fawu
at
1064E, cited in
Chauke
(Supra).
[23]
This finding did not mean that only the 37 persons, so identified,
were the only culprits of violence. The
locus
of the controversy is the identity of the other culprits not caught
on camera or pointed out by managerial staff.
[24]
Minister
of Safety and Security v Madikane and Others
(2015) 36 ILJ 1224 (LAC) at para 4
;
P J
Schwikkard
and S E van der Merwe
Principles
of Evidence
4
ed (2015) p 23
.
[25]
See,
inter
alia
J Grogan “
Derivative
Misconduct - Drawing in the net
”
2018 (34) part (February) ; and T Poppesquo:“
The
sounds of Silence: The Evolution of the Concept of Derivative
Misconduct and the Role of Inferences”
2018
(39) ILJ 34-50.
[26]
Minister
of Safety and Security v Madikane and Others
(2015) 36 ILJ 1224 (LAC) para 4
;
P J
Schwikkard
and S E van der Merwe
Principles
of Evidence
4
ed (2015) p 23
.
[27]
See:
Dunlop
Mixing & Technical Services (Pty) Ltd and Others v National
Union of Metalworkers of SA obo Khanyile and Others
(2016) 37 ILJ 2065 (LC) at para 79.
[28]
See
above.
[29]
Ibid
at para 31 at p1447.
[30]
Ibid
at para 33 at 1447.
[31]
National
Union of Mineworkers and Others v J Grogan NO and Another
(JA 30/08)
[2010] ZALAC 12
(1 June 2010) (‘NUM’).
[32]
See
National
Transport Movement (NTM) and Another v Passenger Rail Agency of
South Africa Ltd (PRASA)
[2018] 2 BLLR 141
; (2018) 39 ILJ 560 (LAC) at para 31.
[33]
Hlebela
(
above)
at paras 8-15 at 2284-2285.
[34]
Hlebela
(above)
at para 17 at 2286.
[35]
Hlebela
(above)
at para 20 at 2287D-E.
[36]
PRASA
(above)
at paras 30 and 31.
[37]
NUM
(above) at para 62.
[38]
See
above.
[39]
See
generally, P J Schwikkard ‘
Silence
and Common Sense’
(2003)
Acta
Juridica 92;
Henry
Lerm ‘
The
right to silence under siege’
(May
2010)
De
Rebus
30.
[40]
P
J Schwikkard and S E van der Merwe
The
Principle of
Law
of Evidence
4 ed (Juta 2015) at 158; DT Zeffert and A P Paizes
The
South African Law of Evidence
5
ed (
LexisNexis
2009) at 579.
[41]
See:
Davis
v Tip NO
1996
(1) SA 1152
(W);
Straub
v Barrow NO and Another
[2001] 6 BLLR 679
(LC);
Fourie
v Amatola Water Board
(2001)
22 ILJ 694 (LC).
[42]
See
above
[43]
See
(above) para 8.
[44]
(2015) 36 ILJ 2280 (LAC).
[45]
At
at para 4.
PRASA
at) at para 31.
[46]
(1987)
8 ILJ 156 (IC) at 162H-I.
[47]
(
2010)
31 ILJ 2836 (LAC) at para 20.
[48]
1062H-
1063B.
[49]
At
1063B.
[50]
At
para 31.
[51]
Ibid
at para 33 at 1447.
[52]
(2001)
22 ILJ 1945 (ARB); [2001] 7 BALR 669 (P).
[53]
(1998)
19 ILJ 939 (CCMA).
[54]
At
para 10.
[55]
At
para 12.
[56]
At
paras 8-15.
[57]
At
para 17.
[58]
At
para 17.
[59]
At
para 17.
[60]
At
para 28.
[61]
At
paras 30 and 31.
[62]
At
para 37.
[63]
At
para 40.
[64]
At
para 44.
[65]
At
para 46.
[66]
See
Maloka, TC ‘
Derivative
Misconduct and Forms thereof: Western Refinery Ltd v
Hlebela 2015 36 ILJ 2280 (LAC)’
in (Vol 19) [2016] PER 36 at footnote 25 with reference to
Klare
1981
Berkeley
J Emp & Lab L
450-482
;
Klare
1985
Md
L Rev
731-840
and
Klare "Countervailing Workers' Power" 63;
Epstein
1983
Yale
LJ
1357
;
Stone
2001
UCLA
L Rev
519
;
Collins 2001
ILJ
(UK)
17;
Mitchell
Redefining
Labour Law
;
Deakin and Wilkinson
Law
of the Labour Market
;
Arup
et
al
Labour
Law.
[67]
In
Hlebela
it was recognised that the gravity of the non-disclosure must be
proportionate to the gravity of the primary misconduct; and
that the
rank of the employee may affect the gravity of the non-disclosure.
[68]
Hlebela
at paras 8-15.
[69]
1952
(1) SA 732
(N).
[70]
See
too
AA
Onderlinge Assuransie Assosiasie Bpk v De Beer
1982 (2) SA 603 (A).
[71]
[
2013]
ZALCJHB 301 at para 61.