About SAFLII
Databases
Search
Terms of Use
RSS Feeds
South Africa: South Gauteng High Court, Johannesburg
SAFLII
>>
Databases
>>
South Africa: South Gauteng High Court, Johannesburg
>>
2019
>>
[2019] ZAGPJHC 539
|

|

S v Ngubane and Another (SS123/2016) [2019] ZAGPJHC 539; 2021 (2) SACR 158 (GJ) (18 June 2019)

REPUBLIC
OF SOUTH AFRICA
IN
THE HIGH COURT OF SOUTH AFRICA,
GAUTENG
LOCAL DIVISION, JOHANNESBURG
CASE
NO:
SS123/2016
In
the matter between:
THE
STATE
And
NGUBANE,
MTHOKOZISI
BHEKI
ACCUSED 1
MAJOZI,
NKOSIYETHU
MUZIMANDLA
ACCUSED 2
SUMMARY
This
case concerned – from the version of the state -
that
the two accused, acting in a group of five originally, acting in the
execution of a common purpose to murder the security guard
and to rob
him and then to rob the store, did murder and rob the security guard,
and, when they then encountered resistance from
the store owner (Mr
Khote), exchanged gunfire with Mr Khote. Two of their own members
were killed in the shootout and another was
wounded. Mr Khote was
also wounded. The two remaining perpetrators surrendered, were taken
into custody, and together with the
third wounded perpetrator, were
transported to hospital. The wounded perpetrator died in hospital and
the remaining two were then,
upon their discharge from hospital,
transported to a police station, where they were charged. In
addition, the version of the state
is that each of them was in
unlawful possession of a firearm and ammunition.
The
court addresses itself to the method of reasoning by which a court
may operate. It considers whether a deductive or an inductive
model
of reasoning is inappropriate and in addition considers the role of
intuition.
Ultimately
the court concluded that the most appropriate model for resolving
questions of fact is a model of reasoning known as
abduction:
reasoning to the best explanation.
The
Court took the view that abduction was the most appropriate model for
dealing with all questions of fact, including circumstantial

evidence.
The
Court observed that the reasoning adopted by Watermeyer J in
R
v Blom
1939 AD 188
was
erroneous in that it was circular and, in addition, was based on a
‘reasonable possibility’ or, worse, a mere possibility

for an issue which was of ultimate significance, and which ought to
have been determined on the standard of beyond a reasonable
doubt.
Furthermore,
the Court observed that the reasoning of Watermeyer J seemed to
require piecemeal reasoning whereas our courts seem
to both require
that the entire conspectus of the evidence is accounted for and that
our courts permit themselves to consider evidence
as a ‘mosaic’
or as a cable or rope, made of strands, instead of demanding that
every fact to be proved must be proved
beyond a reasonable doubt, as
it would, following the analogy of a rope or cable, rather than as
the links in a chain.
The
Court proposed that this is consistent with the abductive model
proposed above.
Finally,
the court turned to deal with the various issues that arose in this
case - reasoning by abduction.
Held:
the accused had acted in a common purpose to rob the store which
included the premeditated murder of the security guard.
Held:
Accused were convicted of
premeditated
murder in respect of the death of the security guard, two counts of
murder – in respect of their two co-participants;
attempted
murder- in respect of the shots fired at Mr Khote; robbery with
aggravating circumstances; possession of an unlicensed
firearm and
the unlawful possession of ammunition.
J U D G M E N T
GRANT,
AJ
:
INTRODUCTION
[1]
In what
follows
I will first set out the
charges against the accused. I then address the critical issue of
proof
-
how it is that, based on pieces of information we may discern about a
past event, we may claim adequate knowledge of that event.
[2]
In addition, I address the issue of
circumstantial
evidence – since I will be adopting a model of reasoning that
relies through-out on inferences - and how and when we may
rely on
inferences as information that may or must be accounted for in an
attempt to discern a past event.
[3]
Thereafter I consider the evidence in this matter and whether it
permits for a version of a past event - the criminal conduct
in
question - to be discerned with adequate certainty.
[4]
Of course, adequate certainty in this context - that of a criminal
trial - must be certainty beyond any reasonable doubt.
CHARGES
[5]
This matter concerns an attack on the Bread-bin and Milky Bar
(hereinafter referred to as ‘the store’) in which
the
security guard and two of the perpetrators were killed. Both accused
were charged as follows:
[5.1]
COUNT 1: MURDER
- lN THAT on or about the 10th day of May 2016 and at or near the
Bread Bin and Milky Bar, Maraisburg in the district of Johannesburg

Central, the accused did unlawfully and intentionally kill an adult
male person, Victor Nkoluleko Dlamini;
[5.2]
COUNT 2: MURDER
- lN THAT on or about the date and at or near the place mentioned in
Count 1, the accused did unlawfully and intentionally kill
an adult
male person, Somandla Nkanyiso Dladla; and
[5.3]
COUNT 3: MURDER
- lN THAT on or about the date and at or near the place mentioned in
Count 1, the accused did unlawfully and intentionally kill
an adult
male person, Lidokuhle sonny-boy Mkhize, -
All three counts of
MURDER read with the provisions of section 51(1) and (2) as well as
schedule 2 of the
Criminal Law Amendment Act 105 of 1997
and further
All (counts of murder) – as read with the provisions of
sections 92(2), 256 and 258 of the Criminal Procedure
Act 51 of 1977
(hereinafter referred to as the ‘CPA’); and
whereas the State’s
summary of substantial facts included the following:
“
The murder of
Victor Nkoluleko Dlamini (the deceased in respect of Count 1) was
planned or premeditated.”
[5.4]
COUNT 4:
ATTEMPTED MURDER
– lN THAT on or about the date and at or
near the place mentioned in Count 1, the accused did unlawfully and
intentionally
attempt to kill an adult male person, Essop Mohamed
Khota by shooting at him with a firearm -
read with the provisions
of sections 92(2), and 258 of the CPA further read with the
provisions of section 51(2)(c) and schedule
2 of the
Criminal Law
Amendment Act 105 of 1997
;
[5.5]
COUNT 5: ROBBERY
WITH AGGRAVATING CIRCUMSTANCES
- lN THAT on or about the date and
at or near the place mentioned in count 1, the accused did unlawfully
and intentionally assault
Victor Nkoluleko Dlamini and did then and
there and with force take from his possession a 9mm Parabellum
calibre FN model hi-power
semi-automatic pistol with serial number
77C49073 and/or 6 x 9mm rounds of ammunition being his property or
property in his lawful
possession and did thereby rob him of the
same, aggravating circumstances as defined by
section 1
of the
Criminal Procedure Act 51 of 1977
being present in that the accused
inflicted grievous bodily harm and/or wielded a firearm during the
robbery -
as defined in
Section 1
of the criminal procedure act 51 of 1977 – read with the
provisions of
section 51(2)
and schedule 2 of the criminal law
amendment act 105 of 1977 and further read with the provisions of
sections 92(2)
,
256
and
260
of the
Criminal Procedure Act 51 of 1977
[5.6]
COUNT 6 -
POSSESION OF A PROHIBITTED FIREARM
- lN THAT on or about the date
and at or near the place mentioned in Count 1, the accused,
unlawfully and intentionally had in
their possession firearms, to
wit:
·
a 9mm Parabellum Calibre Vekto/Lew model 288 semi-automatic pistol;
and/or
·
a 9mm Short Calibre FEG model AP-63 semi-automatic pistol; and/or
·
a .38 Special Calbre Smith Wesson Model 60 revolver; and/or
·
a 9mm Parabellum CZ Model 75 semi-automatic pistol, -
the serial numbers
thereof having been obliterated –
in contravention of
section a(l)(f)(iv) read with the provisions of
sections 1
,
103
,
117
,
120
(1)(a) and
121
, read with schedule 4 of the firearms control act
60 of 2000 and further read with the provisions of
section 250
of the
Criminal Procedure Act 51 of 1977
as well as
section 51(2)
of the
Criminal law Amendment Act 105 of 1997.
ALTERNATIVLEY (to
COUNT 6): POSSESSION OF AN UNLICENSED
FIREARM - IN THAT on or
about the date and at or near the place mentioned in Count 1, the
accused, unlawfully and intentionally
had in their possession
firearms, to wit:
·
a 9mm Parabellum Calibre Vekto/Lew model 288 semi-automatic pistol;
and/or
·
a 9mm Short Calibre FEG model AP-63 semi-automatic pistol; and/or
·
a .38 Special Calibre Smith Wesson Model 60 revolver; and/or
·
a 9mm Parabellum CZ Model 75 semi-automatic pistol, -
without being the holders
of a licence, permit or authorisation issued in terms of the Act for
the said firearms -
in contravention of
section 3 read with the provisions of sections 1, 103, 117, 120(1)(a)
and 121, read with schedule 4 of the firearms
control act 60 of 2ooo
and further read with the provisions of
section 250
of the
Criminal
Procedure Act 51 of 1977
as well as
section 51(2)
of the criminal law
amendment act 105 of 1997.
[5.7]
COUNT 7:
POSSESSTON OF AN UNLICENSED FIREARM
- lN THAT on or about the
date and at or near the place mentioned in Count 1, the accused,
unlawfully and intentionally had in
their possession firearms, to
wit:
·
a 9mm Parabellum Calibre CZ Model 75 semi-automatic pistol with
serial number J1465; and/or
·
a 9mm Parabellum Calibre FN Model Hi-Power semi-automatic pistol with
serial number 77C49073; and/or
·
a firearm the calibre and exact details unknown to the State -
without being the holders
of a licence, permit or authorisation issued in terms of the Act for
the said firearms –
in  contravention of
section 3 read with the provisions of sections 1, 103, 117, 120(1a)
and 121, read with schedule 4 of the
firearms control act 60 of 2000
and further read with the provisions of section 250 of the criminal
procedure act 51 of 1977 as
well as section 51(21 of the criminal law
amendment act 10s of 1997.
[5.8]
COUNT 8:
UNLAWFUL POSSESSION OF AMMUNTION
lN THAT on or about the date and
at or near the place mentioned in Count 1, the accused, unlawfully
and intentionally had in their
possession at least:
·
29 x 9mm Parabellum Calibre rounds of ammunition; and/or
·
3 x .38 Special Calibre rounds of ammunition; and/or
·
3 x 9mm Short Calibre rounds of ammunition -
without being in lawful
possession of: a licence in respect of a firearm capable of
discharging the ammunition; a permit to possess
ammunition; a
dealer's licence, manufacturer's licence, gunsmith's licence, an
import or export or in-transit permit or transporters
permit issued
in terms of the act; or were otherwise authorised to do so –
in
contravention of section 90, read with the provisions of section 1,
103, 117, 120(1), 121 and schedule 4, of the
Firearms Control Act 60
of 2000
, and further read with the provisions of
section 250
of act
51 of 1977.
[6]
The two accused (referred to as ‘A1’ and ‘A2’
respectively) entered a plea of not guilty on all the
above charges
and elected to exercise their right to remain silent and to not
provide any plea explanation.
[7]
It was confirmed that they understood that they were charged under
the minimum sentencing legislation:
Criminal Law Amendment Act 105 of
1997
.
[8]
In addition, the state confirmed, and the accused were warned, that
the state sought to rely on the doctrine of common purpose.
The
accused acknowledged this, and their Counsel indicated that he had
prepared on that basis.
[9]
The accused made various formal admissions in terms of
section 220
of
the CPA which reduced the issues in dispute. These admissions
included a photo album which contained photos of the scene. The

accused confirmed the admissions made.
[10]
In cross-examination the versions of the accused were put to the
witnesses. Both accused agree that they were present at the
scene at
the relevant time, however, they deny any involvement in the robbery
or the possession of firearms or ammunition. In this
way they placed
identity in issue – that it was them who robbed the shop.
Perhaps more accurately, since the two accused
did not dispute being
in the store at the relevant time, the accused denied what it is that
the state alleges they did while in
the shop.
[11]
Later, the defence placed in issue, if the possession of any firearm
were to be attributed to the accused:
[11.1] knowledge of the
accused as to the semi-automatic nature of the firearm (or any
firearm so attributed) in question; and
[11.2] the failure of the
state to positively prove that the accused were not lawful holders of
the relevant firearm licence.
PROOF
[12] Zeffertt and
Paizes,
[1]
in what is arguably
the leading work on the law of evidence in South Africa, pose the
question as to the basis on which a trier
of fact (who I shall refer
to as the ‘decision maker’) decides any matter of fact
given that s/he will only be exposed
to pieces of information
relating to a historical event – particularly where the
information is different in nature. Some
of this information might
appear to support deductive logic, while other information appears to
require the application of intuition.
[2]
They pose the question: how do decision makers decide? They are
daring enough, rightly it seems, to suggest that it is possible
that
a decision maker may not even realise that he or she has chosen one
mode of reasoning over another. Their challenge on ‘proof’

deserves to be repeated here at length:
The fact that these
questions are complex and taxing does not mean that they should
simply be avoided. If the above examples
[
[3]
]
demonstrate
anything at all it is that an election between the two approaches
must
be
made, either to govern the whole field of forensic proof or, as seems
to be more likely, to deal with particular problems on
an
ad
hoc
basis.
The problems that make such an election necessary are neither
uncommon nor artificial. They are everyday situations that
form the
daily staple of our judicial fare.
…
To take but one more
example to illustrate this, consider the question of what kind of
quantum of evidence is necessary to satisfy
a particular standard of
proof. Assume that it is known that of a crowd of 10 000 that
attended a football match, 5 001 gained
admission by knowingly
presenting forged tickets. It could be argued that, since
statistically the probability that any one person
gained admission in
this way is greater than the probability that he or she did not, the
owners of the stadium could succeed, if
this were the only available
evidence, in a civil action against any one of the spectators. And,
if the number of dishonest spectators
were to be increased to a
number far closer to 10 000 (say, 9 900 or, if one felt that this was
still not enough, 9 999), a similar
argument could be raised for the
conviction of any randomly selected spectator on a criminal charge of
fraud
…
It is crucial, however,
that the courts understand that such an election is unavoidable and
that they appreciate the theoretical
nature and practical
consequences of that election in general and in each case. To that
end
they will need to know
more about the extent to which mathematical principles may
appropriately be invoked in a forensic context
and to be able to
understand and evaluate the relationship between these principles and
the conventional rules of evidence.
[4]
[13] I think these are
fair questions and that I must be able to account for how it is that
I operate. I do not think it is an overstatement
to say that far too
often, the decision-making process is utterly unfamiliar to ordinary
people and in some cases, appears quite
strange. It is for this
reason, no doubt, that suspicions arise that the reasons a decision
maker offers for his or her decision
are only ex post-facto
justifications.
[5]
[14]
It may be expected that my attempt to answer this question will
require the adoption of and adaption to some new way of thinking
or
some new set of rules. On the contrary though, the analysis and the
conclusion are, I believe, liberating. It proposes that
decision
makers embrace the manner of reasoning by which they navigate their
everyday lives and eschews almost entirely the imposition
of
operating by “special rules” for particular kinds of
evidence. It allows for exactly the kind of reasoning that
is so
critical to making an informed decision on any set of facts: that it
is all taken into account, and all accounted for.
[15]
Let me begin by saying that I do not think I have access to knowledge
about anything for certain. There is a common misconception
–
if I may – that there is such a thing as definitive ‘hard’
evidence of something. Popularly, for instance,
it is supposed that
cases of ‘direct evidence’ are easy and more compelling
than those involving ‘indirect’
or circumstantial
evidence.
[16]
The first hurdle in understanding the process – or what is
possible
[6]
– is to
understand that this distinction, between cases of ‘direct’
and ‘circumstantial evidence’
does not hold. There are
many cases in which circumstantial evidence is more compelling than
so-called direct evidence.
[7]
A
fingerprint, after all, is circumstantial evidence.
[17]
Direct evidence might seem more forceful in terms of proof
(
probative
), but it is not necessarily so. It can be less
prone to logical error than less direct forms of evidence, but those
kinds of evidence
(such as eyewitness testimony) may have problems of
their own – for instance – whether the witness is
reliable
(a witness may simply be mistaken). Alternatively,
the question arises whether the witness is credible – that is,
whether
the witness is telling the truth or lying.
[18]
Even if provided with what might be considered the most probative of
all evidence – perhaps
real
evidence in the form of a
firearm – the murder weapon, registered to the accused, found
at the murder scene with his fingerprints
on it, where the victim was
found dead – already we are relying on various sources of data:
fingerprints are circumstantial
evidence and are only useful because
an expert has linked an almost invisible pattern on the gun to the
ridges on the accused’s
hands. And then of course, none of this
is of any use if the victim was stabbed to death. Or, worse, was shot
(by the accused’s
gun) but may have succumb to a stab wound
first – inflicted by some other third party. All we are left
with is scattered
fragments of data – which I shall call
information, evidence, or data points – so that we can begin to
create an image
– something comprehendible.
[19]
So, I suggest, as Einstein insisted, one must create a mental
picture. The notion of data points is to shift one into thinking
of a
piece of graph paper – or better still, a canvass, on which the
pieces of information that one receives may be plotted
or marked. The
question becomes whether an image appears on the canvass –
whether, when one assembles all of the pieces of
the mosaic
presented, one may discern a picture. The reference to a mosaic is
quite deliberate and is an analogy often referred
to by our Courts in
explaining how they account for all the evidence – at least
when dealing with circumstantial evidence.
[8]
[20] The model of
reasoning proposed here is entirely compatible and consistent with
how one ought, properly, to deal with circumstantial
evidence. This
is a coincidence, but no accident. The model proposed here recognises
that in all that we do we are drawing inferences,
some inferences are
compound or multi-layered, but every decision is based on inferences,
and is, ultimately, in itself an inference.
There are natural
implications for how to deal with circumstantial evidence –
simply as one deals with all inferences. Nevertheless,
this will be
discussed below from the perspective of circumstantial evidence –
because it is critical to observe the morass
that has developed in
attempting to create special rules to apply to what we might want to
otherwise identify as circumstantial
evidence.
[9]
As one will note from that discussion, the argument there will be
that special rules should be abandoned and what may otherwise
qualify
as circumstantial evidence, ought again, to be treated as we do all
evidence.
[21] Thus, the argument
for treating circumstantial evidence as all other evidence is
bi-directional. Whether one looks at it from
the perspective of how
we should deal with evidence generally, or whether one considers it
from how we ought to treat circumstantial
evidence, one seems to be
driven to the same conclusion. All evidence is, in a relevant respect
“circumstantial” in
that it requires inferential
reasoning and looking backward, one cannot escape that what we might
otherwise treat as circumstantial
evidence is only an instance of
inferential reasoning.
[22]
However, even in proposing this, I have left over a most fundamental
issue. On what basis do I propose that the pieces of information
may
be treated as marks on a canvass? Does this not assume an approach to
reality and knowledge – and, at the very least,
a mode of
reasoning.
[23]
I regret that it does. I say regret because this discussion is often
absent from many claims to knowledge or to have reconstructed
what
happened at a particular moment, at a particular place. Do I, for
instance, rely, as Holmes claims to have on
deduction
(although wrongly so). Can I say – with absolute certainty -
that, say, because:
[10]
Y1 was shot dead; and
The bullet is traced to
the gun of X1;
_____________________________
that it was X1 who shot
Y1.
[24]
No, I cannot. It seems that while the application of law to fact may
follow a deductive form of reasoning, the determination
of the facts
cannot. And even so, there is controversy here – not least
because the idea of law may be regarded as inherently
normative and
not reducible to a mathematical equation.
[25]
Do we operate by induction? That is, by considering the behaviour of
a sample of people in all material respects the same as
the accused,
and infer his or her guilt from that? Certainly not – although,
a hint of this may appear in the form of similar
fact evidence, where
one relies on a sample of the accused’s own conduct to
generalise to what he or she might have done
at the relevant moment.
But otherwise, induction is unhelpful. It belongs to research
sciences or other activities where one seeks
to extrapolate something
about a population, based on a sample of that population.
[26] But then what are we
doing? It would seem that whatever we might claim to be doing, or
whatever we may wish we could do, we
are inevitably engaged in a
mixed process of logical reasoning – both deductive (reliant on
a mathematical model of the world
[11]
and non-deductive (reflecting a more intuitive approach).
[12]
[27]
It seems that the mode of reasoning by which we choose between
versions of facts, or indeed, how we navigate in this world,
is what
has come to be known as ‘
abduction
’.
[28]
In most part it seems that, the mixed or dual mode of reasoning that
we operate on in attempting to decide between versions
of facts, we
engage in an
abductive
exercise
[13]
– as
described, to a large extent, by
Charles
Sanders Pierce
.
[14]
We look at the facts on which we think we can rely and formulate an
explanation for them. In particular, we are looking for the
best
possible explanation
[15]
– or, in criminal law, the only feasible, or, to be more
precise, the only reasonable explanation.
[16]
[29]
It is an advantage of abduction that it admits of uncertainty.
[17]
In this way, it is at least honest. As Tuzet explains:
[A]bduction never yields
certainty. It is a probable inference, that is an inference
determining conclusions whose truth does not
necessarily follow from
the truth of the premises. But this is not a check: it is indeed a
principle of responsibility. The knowledge
of the uncertainty of
abductive conclusions, means the responsibility for their inference.
The knowledge of the lack of certainty
of a certain piece of
reasoning, means the impossibility of concealing an arbitrary
decision under the shield of logic. The shared
knowledge of the
hypothetical nature of a certain conclusion, means the impossibility
of claiming it as necessary.
[18]
[30]
Pierce
[19]
argued that we
approach everything we do with preconceptions and prejudgments. He is
famously credited with saying:
Any novice in logic may
well be surprised at my calling a guess an inference. It is equally
easy to define inference so as to exclude
or include abduction. But
all the objects of logical study have to be classified; and it is
found that there is no other good class
in which to put abduction but
that of inferences. Many logicians, however, leave it unclassed, a
sort of logical supernumerary,
as if its importance were too small to
entitle it to any regular place. They evidently forget that neither
deduction nor induction
can ever add the smallest item to the data of
perception; and, as we have already noticed, mere precepts do not
constitute any
knowledge applicable to any practical or theoretical
use. All that makes knowledge applicable comes to us viâ
abduction.
Looking out of my window this lovely spring morning I see
an azalea in full bloom. No, no! I do not see that; though that is
the
only way I can describe what I see. That is a proposition, a
sentence, a fact; but what I perceive is not proposition, sentence,

fact, but only an image, which I make intelligible in part by means
of a statement of fact. This statement is abstract; but what
I see is
concrete. I perform an abduction when I so much as express in a
sentence anything I see. The truth is that the whole fabric
of our
knowledge is one matted felt of pure hypothesis confirmed and refined
by induction. Not the smallest advance can be made
in knowledge
beyond the stage of vacant staring, without making an abduction at
every step.
[31]
It falls then to the decision maker to declare his prejudices and
preconceptions – rather than pretending that s/he has
none.
[20]
[32]
As
Pierce
noted, it is the direction of the justification which validates a
reason.
[21]
In
this way, the suspicions of Dewey – referred to above
[22]
- are addressed.
[33]
In
addition,
it
is a feature of abduction that it permits one to take account of new
information. In this way it crosses the divide between those
who
might think that intuition and mathematics cannot be reconciled.
[23]
The ability to account for new information and the updating of
beliefs is foundational to Bayes Theorem. Bayes theorem, in
particular,
Bayesian probability,
[24]
acknowledges (as does inductive reasoning)
[25]
that one commences with an initial estimation of the probabilities of
an event – known as the ‘prior probabilities’.
They
are ‘prior’ because they have not yet taken into account
new or additional information. The theorem permits for
very accurate
calculations to be performed based on the integration of new or
additional information to one’s (prior) beliefs
as to the
(prior) probabilities. Once updated, the beliefs represent the
‘posterior probabilities’ – relative
to the
new/additional information. This process may be repeated to
accommodate all information. It represents a clearly attractive

method for establishing the probability that a particular event took
place – which is especially valuable because it requires
that
one declare one’s initial (subjective) beliefs.
[34]
Bayes
theorem
is, regrettably, exceedingly difficult to utilise on a practical
level – and it is, of course, sensitive to one’s initial

estimation. One must also be cautious to only feed into the equation
estimates of variables that are truly independent. A variable
is a
piece of information which can assume different values – for
example, the colour of a car. In the infamous case of
Collins
[26]
– the use of mathematics failed spectacularly because, at
least, the statisticians failed to separate out the variables
properly.
[35]
So, for
instance
, the probability that
the male perpetrator (of a male and female) had both a beard and a
moustache – whereas the fact that
one has a beard implies that
one also has a moustache. This might have been an obvious error –
adopted for the sake of illustration.
In real life, this is not a
simple exercise.
[36]
In
addition
, assigning a probability to
a variable may be readily available in some spheres, such as in
medicine but, for the sort of probability
that is a function of
social research, the necessary data may not be. For instance, in a
rape case, one may need to know the probabilities
of anyone engaging
in casual sex at a public venue. Alternatively, in a murder case, one
may need to know what the chances are
of being attacked and stabbed
in two independent attacks in the same evening. These are the sort of
difficult probabilities that
a decision maker would need available to
process some of the factual issues which confront our courts.
[37]
Thus, I look to the ‘canvass’ to see whether the ‘data
points’ reveal a picture. In the language of
ropes and cables
relied on in discussing circumstantial evidence,
[27]
I consider whether the data has created strands which together may
hold the nominal weight (whatever that is) of a guilty or innocent

explanation.
[38]
In settling on the best and only reasonably possible version one must
accommodate all of the evidence and be alert to the fact
that no
version can be constructed which includes a fact which is
inconsistent with that version. This is, of course, in line with
the
general principle that while
confirmation
of any particular
hypothesis, depending on the degree to which its presence would
discriminate between that hypothesis and any
other, can never
guarantee the truth of any particular hypothesis.
[39]
This is not the case with facts or evidence which is
inconsistent
with a version or hypothesis. One single fact can
disconfirm
a
hypothesis which is consistent in all other respects with the
hypothesis. One need only consider the well-known example of the

black goose to observe the functioning of confirmation and
disconfirmation. The hypothesis that all geese are white can be
confirmed
by the presentation of numerous white geese and then
further white geese. At this point one may amass perhaps millions of
white
geese but nevertheless not have proved one’s hypothesis.
[40]
In contrast, one need only present one single black goose to disk
confirm the entire hypothesis. The power of disconfirmation
is
therefore overwhelming and any attempt at reasoning must be deeply
sensitive to whether or not any fact exists which would disconfirm
a
hypothesis or the version which is proposed.
[41]
As indicated above,
[28]
adopting this style of reasoning it is submitted, allows for what
Zeffertt and Paizes may term a hybrid model where one adopts
a
mathematical model in some respects and otherwise an intuitive model
in other respects. As they say it would seem impossible
to
definitively determine which of these two should prevail. It does not
appear that there is any resolution to the problem of
which style of
reasoning (deductive or intuitive) one should adopt so that perhaps
the answer is that one might need to adopt a
little mathematics here,
and a little bit of intuition there.
[29]
[42]
Does this model of abduction assist in answering the classical
dilemma referred to above
[30]
– of the football stadium? I think it does – because the
model obliges one to consider the alternative possible versions.
It
allows the maths to do substantial work, while reminding one that
each individual accused may, nevertheless, be able to present
a
(reasonably) plausible explanation as to their innocence.
[43]
In exactly the same way as the flaw in what has come to be known as
the “prosecutor’s fallacy”,
[31]
is that more data is required, the abductive model requires that one
must take account of the “big picture”. An abductive

model requires one to consider whether the accused can construct –
perhaps even out of what may be regarded as strands in
a rope or
cable, an alternative hypothesis. One needs more to decide the issue.
[44]
This does not eschew a deductive/mathematical model – it allows
it, as mentioned, to do substantial work, but it allows
one, at the
same time, to give expression to the “intuition” that
there is something wrong in relying on the maths
alone. On an
abductive model, the decision maker would, I expect, remain sensitive
to the presentation, by the accused, of other
information, such as
the accused’s ticket stubs, or proof of payment.
[45]
Thus, I have set for myself the question of what hypothesis could
possibly explain to the exclusion of all other hypotheses,
the facts
before me in this matter. The evidence presented by the state
comprises various different forms of evidence including
eyewitness
testimony and admissions. The defence presented the testimony of the
two accused to the effect that they were present
at the scene of the
robbery innocently. Thus, in this case I must consider not only the
so-called direct evidence, but also the
circumstantial evidence
presented by the state.
FROM
THEORY TO APPLICATION
[46]
In adopting an abductive model of reasoning to establish the facts,
what must one do? The implications are onerous.
[47]
In my view, being faithful to the underlying precepts of abduction
from Pierce and Bayesian theory, one must begin by being
alert to any
preconceptions – preconceptions in the sense that one will
likely have formed estimates of the probabilities
of certain events
having occurred as the information is being presented. These may
relate for instance to the appearance of the
accused, he or she may
walk with a limp or appear intimidatingly large and strong. One may,
for instance, hear that a vehicle drove
through a stop street without
stopping – and form a prima facie view of the impropriety of
doing so.
[48]
One's attention may also be drawn inevitably to the fact that an
accused is in custody and not on bail and that this in itself
will
mean that someone has considered the various factors, including,
whether this person will stand trial or be a menace to the
witnesses
and that it is possible, of course, to draw an adverse inference
against a person for this reason. We are cautioned not
to do so by
case law to the effect that in terms of the appearance of the
accused, s/he may not be presented with any hint of a
previous
conviction.
[32]
[49]
This is the second manifestation of one's commitment to intellectual
honesty. The first, of course, is to admit what it is
that one is
doing in the sense of what mode of reasoning one is adopting.
[50]
The significance and value of setting out one's prior beliefs and
prejudices is that you force yourself into an exercise of
bringing
what might otherwise be unconscious into one's consciousness where
one can work with it or at least acknowledge it.
[33]
[51]
Next it appears sensible to sketch out the various versions which one
will be required to adopt – of course, in the alternative.
Each
version will represent a mosaic or picture, perhaps a moving picture
but a picture nonetheless. It may be helpful - indeed
it would seem
to be helpful - to consider what each party proposes presenting to
sustain each picture that it proposes –
it's version.
[52]
At the very core of each and every version will be a number of “data
points” or “dots” on the “canvass”.
The
exercise requires that one sketch out the dots and then look to see
in what way they can be joined. In fact, in constructing
any version
one is required to join the dots. However, one will likely recall
that – in doing this sort of exercise - the
dots may be joined
in a variety of different ways. On occasion – in doing this
sort of exercise - the “picture”
the author wants you to
see will be numbered. But in forensic fact-finding there is no such
numbering and no particular reason
to attempt to be faithful to
anyone's construction of what it is that the dots show.
[53]
It may be helpful to consider what it is that the prosecution, for
instance, in a criminal case proposes to prove as the data
points
from which one must see the picture it proposes emerge. This is
helpful for two reasons:
[53.1] Firstly, because
one will be alert to what it is that the prosecution in a criminal
case, or perhaps the plaintiff in a civil
case sets up for themselves
as what it is that they intend to prove. This reveals the basis on
which they see a particular version
or picture, so that when the
prosecution or plaintiff fails to prove or set down a particular data
point, the question must arise
as to whether the picture or version
it proposes can truly be drawn. On some occasions, it will remain
possible - some data points
will be redundant, but in others they
will be essential and in their absence, there will be no picture. The
version will have failed
to be established.
[53.2] Secondly though it
will allow one to consider whether, even if the plaintiff or
prosecution is able to prove these data points,
this must necessarily
lead to the conclusion proposed.
[54]
Even at what might be considered an early stage, it will be helpful
to consider whether the picture proposed in fact amounts
to anything
or whether such a picture can be seen or can otherwise emerge from
the various data points. Against this the opposing
versions must be
considered and to the extent to which they overlap the facts are
common cause – however, the picture that
one is being asked to
see is very different.
[55]
The standards of proof applicable in civil and criminal trials will
also have implications for the work one must do at this
stage. While
it is trite that the standard of proof in a civil trial is proof on a
balance of probabilities, one may easily overlook
that this could be
satisfied by one of two mutually exclusive propositions.
[55.1] The first is
simply that the plaintiff's version must be more probable than the
defendant’s version so that for instance
if the plaintiff
proves his case at, say, a 40% probability whereas the defendant
proves his or her case at a probability of only
30%, then the
plaintiff has prevailed.
[55.2] Another
alternative – which appears to be adopted by our courts is the
model by which a plaintiff must prove that his
or her case (version)
is more likely than not, so that s/he must prove at least an excess
of 50%.
[34]
[56]
Nevertheless, in a civil trial one need consider only the version
proposed by the plaintiff and the defendant. One may be in
an
entirely different situation in a criminal trial where, while the
prosecution must of necessity propose one particular guilty
version,
the possibility may arise (such as for instance where an accused
elects to remain silent), that the presiding officer
must, of his own
accord, consider the alternative possibilities – alternative
versions.
[57]
Each alternative version raises the prospect that there are other
pictures that may be consistent even with what the state
proposes to
prove. Of course, since our law can only but take account of the
possibility that an innocent accused person may have
forgotten
certain facts and can only propose several different alternatives,
one will need to consider these alternatives up against
whether the
hypothesis of the state is the only reasonable explanation.
[58]
In this way one may see the prospect of an accused's alternative
versions presenting more force than a single version of the
same
probative value alone. Schreiner J famously commented
[35]
– regarding the reasoning of Davis AJA in the case of
R
v Du Plessis
[36]
that he (Schreiner) fails to see or understand – as Davis AJA
proposed
[37]
- how discrete
alternatives can somehow combine to present, as a composite, a
reasonable doubt against the state’s version.
We see here how
this is indeed possible.
[59]
In
R v
Du Plessis
Davis AJA allowed for the prospects of alternative innocent
explanations operating together in favour of an accused.
[38]
The facts of that case concerned the theft of a motor vehicle
where the critical piece of evidence against the accused having

somehow been involved in the theft was the discovery by the police of
a fingerprint on the underside of the front bumper. The rest
of the
car had been wiped clean presumably to clear it from any possible
fingerprints.
[60]
The accused in his defence explained that it may have been that he
was indeed near the car which may have been driven by friends
of his
and that he had perhaps stood near the car leaning against it and
lent down touching the bumper to steady himself, alternatively
he may
have lent down to tie his shoelaces and in doing so again touched the
bumper.
[61]
The significance of these explanations is that they are mutually
exclusive and yet they were taken together. Thus, an accused
person
has in his or her favour that he or she may propose alternative
possibilities except that, as the directive from Schreiner
J appears,
at its core, to caution against, is that these versions must at least
be compatible.
[62]
Thus, the question for consideration was, on the one hand, for the
state, whether the fingerprint had been placed there when
the accused
made contact with the vehicle during the course of its theft or
thereafter, knowing it to have been stolen. Alternatively,
for the
defence, whether his fingerprint had come to be on the bumper by
virtue of some innocent explanation such as that he had
lent on the
car to steady himself either when he was leaning on the car or
perhaps as he might have lent down to tie his shoelaces
near the car.
[63]
If our criminal law required of an accused person to be able to
account for his or her whereabouts at every moment of everyday
things
may well be different.
[64]
I turn now to consider the law of circumstantial evidence and the
implications, if any, of the model proposed here –
of
abduction.
[65]
Thereafter I discuss the law of common purpose and whether it permits
for a conviction of the unlawful possession of a firearm.
[66]
I
turn then to an application of the model of abduction to the facts of
this case.
[39]
CIRCUMSTANCIAL EVIDENCE
[67]
Our law on circumstantial evidence is governed by what have come to
be known as the two cardinal rules of logic in
R
v Blom.
[40]
These “rules” appear from the judgment of Watermeyer JA,
as follows:
In
reasoning by inference there are two cardinal rules of logic which
cannot be ignored:
The
inference sought to be drawn must be consistent with all the proved
facts. If it is not, the inference cannot be drawn.
The
proved facts should be such that they exclude every reasonable
inference from them save the one sought to be drawn. If they
do not
exclude other reasonable inferences, then there must be a doubt
whether the inference sought to be drawn is correct.
[41]
[68]
Despite the reverence with which this judgment is cited, it appears
to be conceptually flawed and also appears to be at odds
with what
our courts do.
[69]
What is crucial to notice – as Zeffertt and Paizes have –
that there is a logical flaw built into the ‘rules’.
The
rules propose that every inference sought to be drawn must be:
[69.1] Consistent with
all the
proved facts
; and
[69.2] The only
reasonable inference given the
proved facts
.
[70]
Zeffertt and Paizes ask, rightly, what are these ‘proved facts’
– and on what standard must they have been
proved.
[42]
In my
view the entire exercise could grind to a halt here. However, I will
let it be for the moment because there is possibly an
even greater
problem ahead: if every inference is to be tested against the
standard of reasonable doubt – the ‘rules’
in
Blom
are a specific injunction to treat evidence in a piecemeal fashion –
utterly contrary to what our courts in fact do.
[71]
The reasoning adopted by Watermeyer AJ is, itself, contrary to his
own rule.
[72]
To understand the reasoning adopted by Watermeyer AJ, it is helpful
to understand the basic facts of the case. The case concerned
whether
the accused was responsible for the death of a woman whose dead body
had lain across a railway line and had been struck
by a passing
train. The train had done such damage to the body that it was
difficult to know why the woman had died – and,
most
importantly, whether the accused was responsible. There was only
so-called circumstantial evidence against the accused.
[43]
[73]
For Watermeyer AJ, the evidence fell into three categories:
(1) Evidence of the
conduct of the accused before the event, showing that he had (
a
)
a motive to kill the deceased, (
b
) a design to kill someone.
(2) Evidence to show that
accused had an opportunity to kill the deceased.
(3) Evidence of the
conduct of the accused after the event showing a guilty
conscience.
[44]
[74]
As to my submission that the reasoning is flawed - one may see that
in
R v Blom
itself, logical errors were committed in the
application of the very rules which went on to become regarded as
critical to reasoning
by inference. Watermeyer JA, in
R v Blom
,
from whose judgment the rules appear, agreed that nothing less than
proof that the victim in that case was murdered in a particular
way
(by chloroform poisoning) would suffice to prove that the accused had
killed her.
[75]
However, in what may qualify as, with respect, a strange line of
reasoning, he proceeded to arrive at that conclusion by accepting,
by
the operation of a
weak inference
– that the accused had
killed the victim – and so in turn that the victim had been
killed in the required way. What
is critical to note is that his line
of reasoning, as impressive as his logic appears, operates on the
assumption of the answer
to the ultimate question: that the accused
had killed the victim.
[76]
It may be helpful to strip this of its content so as to reveal the
underlying logical form. What Watermeyer JA set for himself
was that,
to resolve the ultimate enquiry, whether the accused killed the
deceased, say “D”, he must find “E”
–
whereas “E” is not the ultimate issue. The then reasoned
as follows:
D is “reasonably
possibly” true
E is ‘possibly’
true.
_________________________
[Conclusion]
[77]
The reasoning of Watermeyer JA is as follows:
If the case be considered
without regard to the evidence of Dr. Naude and Inspector Talken,
then it would, in my opinion, have been
insufficient. It would have
related entirely to facts and circumstances antecedent to and
coincident with the girl's death, and
even if the accused's motive
and unlawful design connected with chloroform and opportunity were
given their fullest weight, an
inference that he killed the girl
would not have been justified because in no way could they be
regarded as excluding the hypothesis
of murder by someone else.
But the evidence of a
guilty conscience after the event supplies additional material for
inference which when considered together
with the other evidence
makes it reasonably possible to infer that the accused killed the
girl, and that leads to a possible further
inference that he did it
by means of chloroform, and excludes the hypothesis of murder by some
other person.
[45]
[78]
Thus we see that ‘evidence of a guilty conscience’ is
added to the other material to conclude that it was ‘reasonably

possible’ the accused killed the victim – and that, from
that, it became ‘possible’ to infer that the accused
had
done this by an application of chloroform.
[79]
This extract reveals several problems:
[79.1] At times, the
court operates by agreeing that the accused can only be regarded as
the murderer if the victim was killed by
chloroform – but then
relies on what is, ultimately in question (whether the accused killed
the victim), to conclude that
the victim was killed by the
administration of chloroform.
[79.2] The issue, which
the court agreed it must establish, that the victim was killed by the
administration of chloroform, was
proved on the basis of a ‘possible’
inference.
[80]
The reasoning is at least circular
[46]
and reveals that the court was satisfied to convict on what was only
a
possible
inference.
[81]
In addition, it appears to be possible to read the second rule as
requiring that each and every inference sought to be drawn
must –
for a criminal case – be the only reasonably possible
inference.
[47]
This
interpretation would, in the context of a criminal trial, demand that
a court, in drawing an inference, considers each and
every proposed
inference to be drawn, individually against the standard of proof
beyond a reasonable doubt, and if found to fail,
must be discarded –
and so on for each and every inference sought to be drawn.
[82]
It is helpful to distinguish two kinds of reasoning: chain type
reasoning versus what may pass under different names as rope,
cable
or mosaic type reasoning.
[82.1] Where one adopts
‘chain type’ reasoning – the name arises from the
analogy of a chain- which is only as
strong as its weakest link.
Thus, in a case where one adopts “chain type” reasoning”,
every link must be –
in a criminal case – as strong as
whatever “proof beyond a reasonable doubt” would demand.
[82.2] Where one adopts
“rope/cable style” reasoning, one permits items of
evidence, which would not in themselves carry
the weight of ‘proof
beyond a reasonable doubt’, but one may rely on several such
“strands” which, when
weaved together, may well carry
such weight. We see our courts adopting this style of reasoning –
but it is expressly prohibited
by the second of the
Blom
rules.
[83]
As indicated, Watermeyer AJ himself relied on rope/cable style
reasoning – contrary to his own rule – the
second rule in
Blom
: every inference must be the only reasonably possible
inference.
[84]
For some, this interpretation is neither sensible nor even possible.
Yet it is the interpretation given to the circumstantial
evidence
presented in the infamous Australian “dingo case” of
R
v Chamberlain
.
[48]
This approach has since been rejected in Australia.
[49]
[85]
This interpretation would also require of our Courts that they
consider evidence on a piecemeal basis, whereas, clearly, they
do
not. Our Courts have shown that they are prepared to aggregate
discrete facts and inferences together to arrive at a sensible

answer.
[50]
[86] That they do so is
also evident in the ‘mosaic’ to which our courts refer.
In
State
v Hadebe
,
[51]
the Supreme Court of Appeal followed the approach set out in
Moshephi
& Others v R
,
[52]
where the following was said:
The
question for determination is whether, in the light of all the
evidence adduced at the trial, the guilt of the appellants was

established beyond reasonable doubt. The breaking down of a body of
evidence into its component parts is obviously a useful aid
to a
proper understanding and evaluation of it. But, in doing so, one must
guard against a tendency to focus too intently upon
the separate and
individual parts of what is, after all, a mosaic of proof. Doubts
about one aspect of the evidence led in a trial
may arise when that
aspect is viewed in isolation. Those doubts may be set when it is
evaluated again together with all the other
available evidence. That
is not to say that a broad and indulgent approach is appropriate when
evaluating evidence. Far from it.
There is no substitute for a
detailed and critical, examination of each and every component in a
body of evidence. But, once that
has been done, it is necessary to
step back a pace and consider the mosaic as a whole. If that is not
done, one may fail to see
the wood for the trees.
[53]
[87]
Each piece of a mosaic may be unclear or even vague in itself, but,
when placed with enough other pieces, it reveals a ‘picture’

of what occurred.
[88]
Thus, following the general model of abduction proposed above –
to the extent to which it is consistent with what may
sensibly be
extracted from the ‘cardinal rules’ in
R v Blom –
the following seems to follow:
[88.1] There can be
nothing wrong with a rule of logic that reminds one that one cannot
hold two mutually exclusive beliefs simultaneously.
One may also not
draw an inference from facts where one of those facts is inconsistent
with the inference. However, this must be
true of all inferences –
primary (from ‘direct evidence’) or secondary (from
‘circumstantial evidence’).
[88.2] We must decide
what standard of proof ought to be applied to any particular
inference or issue and consider whether, given
the evidential
material
considered together
, including any inferences
that are at least permissible, the appropriate standard is met.
[88.3] The standard of
proof required will be determined by the importance of the fact in
question. Where the standard of proof
to be satisfied is proof beyond
a reasonable doubt, it will be perfectly acceptable to ask whether
any other reasonable explanation
could account for all the evidence.
This is simply a different way to ask whether another reasonable
possibility exists that would
explain the evidence in question.
[89]
There is nothing new or startling about this approach. In fact, it is
the suggestion only that we continue to do this where
we do and that
we are not distracted by thinking that we need to switch into some
other special mode of thought that is simply
a distraction. It really
is nothing more than to suggest that we should adopt – as I
believe we do in any event, when we
do not get distracted: abductive
reasoning.
[90]
One
may find that inevitably the following
question is posed:
What are the chances
that:
Q (an inference) is not
true (the null hypothesis);
[54]
Given
[55]
that:
A;
B;
C; [and] ...
[91]
In the classic example of the man (X) carrying a blood-stained sword
leaving the home of a person (Y), who, on immediate inspection,
is
found to be dead for having been run through with a sword, the sort
of question we seem naturally inclined to ask is: what are
the
chances that X was not responsible for the death of Y, given that:
The victim was killed
with a sword;
X was with Y immediately
before Y was found to be dead;
X had a sword; and
X's sword was blood
stained.
[92]
Given that the answer to this question takes on the significance of
answering whether we regard X as innocent or guilty, the
degree of
certainty that is required is beyond a reasonable doubt.
[93]
There will be instances in which a fact, or secondary fact, from
which an inference, which carries the ultimate significance
(of guilt
or innocence), may only attract a degree of certainty as to its
existence of far less than the ultimate degree of certainty.
[94] On a variation of
the analogy of a mosaic,
[56]
a
more helpful analogy may be that of a rope or cable - where each
strand of the rope or cable is, in itself, weak, and unable
to ‘carry
the load’ at the standard of proof required of the inference
required. But - as the analogy goes - the strands
may, when weaved
together, carry the ‘required load’. This is, of course,
exactly what is required by abduction.
[95]
In contrast, by analogy, one may note that a chain carries the weight
required of it to support, by the operation of a completely
different
mode. Each link in a chain must be capable of carrying the load.
[96]
This mode of reasoning (by analogy), is the sort that is appropriate
to facts or inferences which must all be proved at the
required
degree of proof to sustain the ultimate conclusion. Thus, on a charge
of murder, where proof is required that the accused,
acting with the
required
mens rea
, committed the wrongful conduct. Each of
these must be proved beyond a reasonable doubt, and each component
(the elements (conduct,
unlawfulness, capacity and fault) must be
proved beyond a reasonable doubt. Thus, a chain of at least four
links must be established,
where each link is proved beyond a
reasonable doubt.
[97]
Then, if, in any case, whether the accused committed the conduct in
question, turns on whether, and only whether, the accused
was, say,
in possession of a 9mm firearm, then proof that the accused was in
possession of such a 9mm gun, assumed the significance
of the
ultimate conclusion, and must be proved beyond reasonable doubt.
[98]
Other circumstances may arise that require that a conclusion of
ultimate significance is, as usual, proved beyond a reasonable
doubt,
but it may be that this conclusion can be proved by the combination
of proof of any, say, 5 of any 100 other facts or inferences.
So, for
instance, if the question is whether the accused killed the victim,
this will depend on innumerable facts or inferences
which will
support the conclusion.
[99]
One may be satisfied that, for instance, the accused had threatened
to kill Y, had attempted to do so before, has stood to
gain
financially or otherwise from the death of the victim, had
opportunity or was seen in the vicinity of the victim around the
time
of death, said when he woke up that day that he was going to kill
someone that day, and that blood was found on his clothes,
his
clothes were torn, he had scratch marks on his face and the victim
had skin under his/her nails, and finally that the accused
had asked
the local reverend to forgive his ‘sin’ that night -
whereas he had never sought redemption before. If one
is not
satisfied, there will be innumerable facts or inferences which would
implicate the accused.
[100]
Of course, anything which is inconsistent precludes the required
inference. This is the work done, by, say, an alibi.
[57]
An alibi is straightforwardly inconsistent with the guilt of an
accused and cannot be forged together with other facts to draw
the
ultimate conclusion of guilt.
[101]
A view of circumstantial evidence as creating alternative possible
hypotheses which may be constructed of strands in a rope
or cable can
easily and well be accommodated in the abductive model proposed. A
further benefit of recognising this is that it
permits for the
construction of alternative hypotheses which are not linked in the
same way as the scales which we often have reference
to assist us
with the analogy of balancing.
[102]
The difficulty with the notion of a scale is that whatever goes in
one pan or taken from that pan effects the relative position
of the
other pan. This is difficult to conceive where the presiding officer
needs to construct, at least conceptually, all
independent
versions or hypotheses while acknowledging at the same time that
there may remain an extent to which the truth of what happened
will
remain unknown and – if I may be so bold – to some
extent, unknowable.
[103]
It follows from what I have said, that a court is engaged in the
determination of, in a criminal case, whether the states
version has
been proved beyond reasonable doubt and in the case of the civil
trial whether the version of the plaintiff has been
proved on a
balance of probabilities.
[104]
The question that is addressed on this model is whether, treating the
guilty hypothesis as a rope or cable, all of the necessary
strands
are in place – treating the wrongful conduct and wrongful
mental state as all part of ultimately one composite event.
COMMON PURPOSE
[105]
Common purpose is the doctrine according to which, understood
properly, with reference to groups who act in what would qualify
as a
common purpose,
[58]
what
everyone in a group does, is to be regarded as what each person in
the group does.
[59]
The
doctrine has been approved as constitutionally compliant.
[60]
[106]
It therefore closes the gap that may otherwise exist between the
conduct of any member of a group and the killing of another
member of
the group – by another member of the group or anyone responding
in private defence to the attack - where the surviving
member did not
actually cause the death of the other member of the group.
[61]
[107]
In
S v
Nkombani
[62]
a party (‘X’) to a common purpose to rob, which included
the prospect of killing, who dispatched two co-perpetrators
to rob a
filling station, was convicted of the murder of one of the
co-perpetrators who, during the course of the robbery, was
shot and
killed by the other co-perpetrator. Significantly, the causing of
death of the co-perpetrator was imputed to X.
[63]
[108]
In
S v
Nhlapo
[64]
the Appellate Division convicted co-perpetrators of the murder of a
guard, killed during the commission of a robber, by a fellow
guard.
Van Heerden AJA (Wessels JA and Muller JA concurring) stated:
[T]he robbers knew that
they would have to attack and overpower guards who were armed for the
specific purpose of using their fire-arms
to thwart any attempted
robbery. It may be conceded that they hoped to overpower the guards
without a shot being fired by the latter,
but they must have known
that the guards would endeavour to use their firearms when attacked.
It follows that they must have known
that their attack on the guards
could lead to a gun battle during which
anybody,
be it a guard, one of the robbers
or
an innocent bystander, might be killed in the envisaged cross-fire.
Consequently they also foresaw the possibility of one guard
being
killed by a shot fired in the direction of the robbers by another
guard or, for that matter, a person such as a staff member
of [the
store] witnessing the attack. In sum, the only possible inference, in
the absence of any negativing explanation by the
appellants, is that
they planned and executed the robbery with
dolus
indeterminatus
in
the sense that they foresaw the possibility that
anybody
involved in the robbers' attack
,
or in the immediate vicinity of the scene, could be killed by
cross-fire.
[65]
[109]
Defence counsel relied on an article
[66]
which I wrote in which I am critical of the doctrine and application
of common purpose. Counsel referred me, in particular, to
part of my
conclusion
to the effect that the principles relating to mistake of law
[67]
ought to be applied to the doctrine of common purpose so that a
mistake of law where the crime concerned requires dolus, would

exclude the operation of common purpose.
[110]
Defence counsel however failed to explain the argument or pursue the
argument in any substantive form. None of the premises
or
intermediary steps engaged with in the argument were argued before
court and I find myself unable to follow the conclusion to
which I am
referred.
[111]
I must treat the accused as they would have been treated had this
argument been presented before any other court. I expect
that at the
very least in the form that was presented it would have been rejected
out of hand as I must do.
COMMON PURPOSE AND
POSSESSION OF AN UNLAWFUL OR ILLEGAL FIREARM
[112]
What remains then only is what may become of the charges relating to
the unlawful possession of firearms or of firearms where
the serial
number has been obliterated. The pertinent question in this respect
is whether possession of these items may be imputed
by common
purpose, or imputed at all. This is the question of whether
possession is, of its very nature, an “eiehandgige
misdaad”
– in English “own-hands” crime – a crime
which one can only permit oneself – with
one’s own hands.
[113]
The best-known example of an own hands crime is the former crime of
rape as it was defined previously under the common law.
Under the old
common law definition only the actual perpetrator who had himself,
unlawful sexual intercourse with the victim could
be convicted of the
crime of rape. A person who assisted – to any extent –
even by holding the victim down, could not
be convicted of rape
because it was an own hands crime.
[114]
Examples persist today such as bigamy and driving under the
influence. It is only the person who is is already married under
the
civil law who then knowingly engages in a second marriage, who
commits bigamy. A marriage officer who facilitates even knowingly,

this second civil marriage, is at most an accomplice to the bigamy
cannot him or herself be convicted of the offence of bigamy.

Similarly for driving under the influence. It is only the individual
who him or herself is intoxicated and takes control of a vehicle
who
commits the offence – no other person, however enabling, can be
convicted of the offence – but may be convicted
as an
accomplice to driving under the influence.
[115]
From these examples one may notice that there appears to be something
related to the status of the individual that is specific
to him or
her which makes it the case that only that person can commit the
offence in question. In the crimes described the positive
status in
one respect, such as already being married under civil law, or being
intoxicated, qualifies one for a possible conviction
of the crime in
question.
[116]
According to Snyman:
Possession
consists of two elements, namely a physical and a mental. The
physical element is objective in nature. It is referred
to as corpus
or detentio, and entails the physical control over the article. The
second element is subjective in nature. It is
referred to as animus,
and describes the intention with which X exercises control over the
article. Before X can be said to possess
an article, both corpus and
animus must be present, and they must be present simultaneously
.
[68]
[117]
The
leading
case
appears to be
S
v Nkosi
[69]
where the following was said:
The
issues which arise in deciding whether the group (and hence the
appellant) possessed the guns must be decided with reference
to the
answer to the question whether the State has established facts from
which it can properly be inferred by a Court that:
(a) the group had the
intention (animus) to exercise possession of the guns through the
actual detentor and
(b) the actual detentors
had the intention to hold the guns on behalf of the group.
Only
if both requirements are fulfilled can there be joint
possession involving the group as a whole and the detentors, or

common purpose between the members of the group to possess all the
guns.
[70]
[118]
This
has
subsequently
been
adopted by the Supreme Court of Appeal in
S
v Kwanda
[71]
and endorsed and confirmed by the Constitutional Court in
S
v Makhubela and Another
.
[72]
It is helpful to have reference to what the unanimous Court
[73]
said there – an extensive quote is included for its importance:
What
remains are the applicants' convictions for the unlawful possession
of firearms and ammunition, that is, counts four and five.
It is
common cause that they did not have any firearms in their possession.
They were, however, convicted of these charges in the
trial court on
the basis of the doctrine of common purpose.
In
convicting the applicants for unlawful possession of firearms and
ammunition on the basis of the doctrine of common purpose,
the trial
court departed from settled jurisprudence. The test for establishing
liability for the possession of firearms and ammunition
was
established in S v Nkosi as follows:
'The issues which arise
in deciding whether the group (and hence the appellant) possessed the
guns must be decided with reference
to the answer to the question
whether the State has established facts from which it can properly be
inferred by a Court that:
(a)
the group had the intention (animus) to exercise possession of the
guns through the actual detentor and
(b)
the actual detentors had the intention to hold the guns on behalf of
the group.
Only if both
requirements are fulfilled can there be joint possession involving
the group as a whole and the detentors, or common
purpose between the
members of the group to possess all the guns.'
This test has since been
cited with approval in numerous judgments of the High Court and the
Supreme Court of Appeal.
In
these judgments, the courts have found perpetrators guilty of a crime
involving the use of firearms on the basis of the doctrine
of common
purpose, but nevertheless found that the perpetrators could not be
found to be guilty of the unlawful possession of firearms
on the
basis of this doctrine
.
The test takes into account the fact that the application of the
doctrine of common purpose differs in relation to 'consequence

crimes', such as murder, and in relation to 'circumstance crimes',
such as possession.
[74]
[119]
The
last
(reported)
word
seems to have come from Binns-Ward J in
S
v Jordaan and Others
,
[75]
who endorsed the reasoning in
Nkosi
,
stating:
As
discussed, accused 4 and 5 were probably aware that accused 1
was armed and that he intended to use the firearm against
any member
of the Hard Livings gang that they might encounter. That would not be
sufficient to convict them of unlawful possession
of the firearms and
ammunition. As explained by Marais J in
S v Nkosi
1998
(1) SACR 284
(W)
, joint possession of firearms and ammunition is
proved only if the state establishes beyond reasonable doubt
(a)
that the
company of which the actual detentor was part intended as a whole to
exercise possession of firearms through the actual
detentor and
(b)
that the actual detentor intended to hold firearms on behalf of the
others.
[76]
[120]
These cases all
appear
to turn on the fact that – in line with
fundamental principles of common purpose - the mental requirement of
p
ossession
cannot be attributed. This in itself will limit or virtually
eliminate the attribution of “possession” – as it

has in all of the cases in which it has arisen for decision.
RIGHT TO SILENCE
OBLIGATION TO PRODUCE
LICENCE
[121] The state relied on
section 250(1)
[77]
of
the CPA which provides, in summary, that where an accused is charged
with an offence which involves an object in respect of which
the
accused person must hold a valid licence to possess that object, an
obligation or duty falls to the accused person to produce
the
licence. This reliance was disclosed in the indictment.
[122]
It would seem improper not to indicate some concern as to whether the
section relied upon does not impermissibly intrude upon
an accused
person’s presumption of innocence and or his or her right to
silence. This might arise from the natural observation
that the
provision permits a court to convict an accused of an offence where
the state has not proved one of the elements. However,
in
S
v Fransman
[78]
the court rejected a constitutional challenge to this provision.
[123]
I remain concerned however as to the answer to what I understand to
be the relevant question: whether the provision permits
for the
conviction of an accused person despite a reasonable doubt as to his
or her innocence.
[79]
[124]
I am mindful however that I was not asked to consider this question
nor was the constitutionality of the provision challenged
in any way.
However, even if it had been, I expect that a ready and accurate
answer would be to the effect that there are moments
during a
criminal trial when one may continue to remain silent, only at one's
peril.
[125]
MosenekeJ remarked in
Thebus
:
[80]
'The
fact that she or he is not obliged to testify does not mean that no
consequences arise as a result. If there is evidence that
requires a
response and if no response is forthcoming, that is, if the accused
chooses to exercise her or his right to remain silent
in the face of
such evidence, the Court may, in the circumstances, be justified in
concluding that the evidence is sufficient,
in the absence of an
explanation, to prove the guilt of the accused. This will, of course,
depend on the quality of the evidence
and the weight given to that
evidence by the Court.'
[81]
[126]
In my view once it has been proved that one was in possession of a
firearm and one is charged with the unlawful possession
of that
firearm, the moment arises where one must break one's silence and
reveal one's licence – quite constitutionally.
FAILURE TO PROVIDE A
VERSION
[127]
It is necessary to note an argument raised by the state in response
to questions put to the accused as to why they had not
given their
innocent version previously.
[128]
In this respect the state referred me to the Constitutional
Court case of
S
v Thebus
[82]
of 2003.
[129]
However, on my reading of the case a majority - four
[83]
against three
[84]
- took the
view that it would be improper to draw a negative inference on the
basis of silence.
[130]
To draw an inference in these circumstances from their silence
is, on the view of the four members of the court, to
punish them for
exercising their right to silence.
[131]
In my view, with respect, this must be correct, in particular because
when an accused is warned that they may elect to remain
silent, he or
she is not also warned at the same time that if they elect to do so –
to remain silent - they may suffer negative
consequences.
[132]
With all of this in mind, I now turn to an application of the
principles set out above.
APPLICATION
[133]
If the state is to be allowed to prevail, it's version must be the
best explanation for the evidence and leave no room for
an
alternative reasonable version. There can also, of course, be nothing
inconsistent in the version opted for - or, in recognizing
an
alternative version for the accused, there can be nothing
inconsistent in this version.
[134]
At its core, of course,
there
are, in this
case, essentially two competing versions (hypothesis - which I will
refer to as versions).
[135]
In what
follows
I set out the respective
versions for consideration as supported by the evidence presented.
[136]
I also set out, as a consequence of deliberately challenging my own
reasoning, the issues which arose during the presentation
of the
evidence and which posed the danger of preconceptions doing the work
of reasoning. Having forced them into my awareness,
I consider
whether they could be justified or otherwise inform my ultimate
decision.
HYPOTHESIS - VERSIONS
COMMON CAUSE:
[137]
It
appears
to be common cause - and the
versions of all overlap to the effect that:
[137.1] the store was
indeed robbed that day;
[137.2] That two of the
attackers were killed (one dying on the scene and the other later in
hospital);
[137.3] That the security
guard of the store (Mr Dlamini) was shot dead by one of the
attackers;
[137.4] That an employee,
Ms Masango, witnessed the shooting of Mr Dlamini;
[137.5] That Mr Khote,
the owner of the store, intervened and exchanged gunfire with the
perpetrators and was responsible for shooting
the two who died;
[137.6] That a fifth
member of the group of attackers managed to escape;
[137.7] That an unknown
police officer attended at the scene shortly after the attack and
that he assisted Mr Khote in apprehending
the attackers in the store
- who had surrendered and were standing with their hands raised when
the officer and Mr Khote reentered
the store; and
[137.8] That the accused
were at the store throughout the robbery.
[138]
Following on observing where the versions overlap - what is common
cause, it becomes clear that it is only the question of
what the
accused did in the store that is in question.
STATES VERSION
[139]
The states version is that the two accused, acting in a group of five
originally, acting in the execution of a common purpose
to murder the
security guard and to rob him and then to rob the store, did murder
and rob the security guard, and, when they then
encountered
resistance from the store owner (Mr Khote), exchanged gunfire with Mr
Khote.
[140]
Two of their own members were killed in the shootout and another was
wounded. Mr Khote was also wounded. The two remaining
perpetrators
surrendered, were taken into custody, and together with the third
wounded perpetrator, were transported to hospital.
The wounded
perpetrator died in hospital and the remaining two were then, upon
their discharge from hospital, transported to a
police station, where
they were charged. In addition, the version of the state is that each
of them was in unlawful possession
of a firearm and ammunition.
NECESSARY IMPLICATIONS
[141]
The version of the state carries the following necessary implications
- there must be evidence (information) from which it
may be inferred
that:
[141.1] the accused were
in a common purpose - as described;
[141.2] one or other of
them, or of their group (assuming a common purpose is established -
two of whom died in the shootout), did
murder and rob the security
guard and proceeded then to engage in a shootout with the owner, and
did, in fact, shoot him.
[141.3] They will be
unable to account satisfactorily for their joint presence at the
scene at the relevant time; and
[141.4]
They will be unable to account satisfactorily for their conduct in
the shop.
[142]
The identification of the accused as the relevant two perpetrators
from the day in question, relied partly on dock identifications
(by
Mr Khote and Ms Masango). In addition, the identification relied on
the logical inference that three men were arrested at the
scene, that
the accused admit being arrested at the scene, that one of their
number died from wounds sustained in the attack, and
that they –
the remaining two – were presented in Court.
[85]
MS MASANGO.
[143]
Ms Masango testified that, on the day in question, she had been
talking to Mr Dlamini when a man approached him and
shot him at
point-blank range. She testified that this man then took Mr Dlamini’s
weapon and shot him again.
[144]
Ms Masango was asked in evidence in chief whether, since the day of
the attack, she had seen the accused again. She responded
that she
had not. Her answer, as I understood it, begged the further question
whether she meant that she had not seen either perpetrator
until then
(when she testified), or including then – including the time at
which she was testifying. The state did not clarify
the position and
proceeded with its questions.
[145]
This left a critical question unanswered for the court and the legal
representative of the accused sought to place an interpretation
on
her response to exclude an identification of both accused as
perpetrators from the day.
[146]
At that point, given the lack of clarity on what it was that Ms
Masango had meant, I was obliged to recall her to clarify
her
testimony on this point.
[86]
[147]
Upon being recalled and questioned for clarification on what she had
meant, she immediately identified accused 2 as the person
who had
shot Mr Dlamini – in front of her.
[148]
The evidence of Ms Masango is that A2 was the actual cause of death
of Mr Dlamini in that he, A2, shot Mr Dlamini several
times at close
range directly in front of her.
MR KHOTE
[149]
The state's version is supported by the testimony of Mr Khote
who testified that after hearing gunfire, he opened the
door of his
office, after picking up his firearm.
[150]
He explained that he noticed an individual standing in front of his
office between his office and the counters in the shop.
This
individual was holding a firearm and he (Mr Khote) called to one of
the cashiers, asking for the cashier to confirm whether
this is one
of the responsible perpetrators. This individual turned and he and Mr
Khote exchanged gunfire.
[151]
Mr Khote testified that this individual then exited the shop. It
transpires and is common cause that this individual was on
of the
perpetrators, presumably the first one to be engaged in an exchange
of gun-fire with Mr Khote, who apparently died on the
scene and his
body was found by Capt. Alvers and Sgt. Naidoo. Mr Khote testified
that he closed the door to his office and reloaded
his firearm.
[152]
He then reopened the door and noticed a second individual standing in
front of his office - again in the space between his
office and the
counter - and they too exchanged gunfire. This individual fell to the
ground and was apparently incapacitated.
[153]
A third perpetrator entered his field of view when he ran to the
second perpetrator (now lying incapacitated on the floor
in front of
his office) and attempted to collect that perpetrator’s firearm
from him, which was now lying on the floor. Mr
Khote identified this
perpetrator as accused number two (A2) before court.
[154]
Mr Khote indicated that he exchanged fire with this perpetrator and
testified that he had wounded this perpetrator in the
left upper
thigh. It transpires that A2 - it is common cause – was not
shot in his left upper thigh.
[155]
Mr Korte continued to testify that this individual then retreated
back into the store. Mr Khote continued in his testimony
to indicate
that a forth perpetrator had appeared behind the counter from where
he had been standing at the door to his office
– in a position
just to the right of where the cashier’s would sit or stand –
and he and this perpetrator also
engaged in an exchange of gunfire.
[156]
At this point there was the possible suggestion that Mr Khote could
not have possibly seen a person standing at that position.
However,
it was shortly conceded by the defence when it was clarified that
reference was being made to the cashier's counter which
is fairly low
to the ground, at perhaps the height of 1 to 1.2 m.
[157]
Mr Khote's evidence was that he observed this individual clearly
during the exchange of gunfire and identified this person
as accused
number one (A1) before court.
[158]
Mr Khote continued in his testimony to explain that after having
engaged with the fourth perpetrator who he identified as
accused one
(A1) before court, he had heard customers, and, in particular
children, or perhaps babies, screaming and that he had
attended to
them and had led them outside. He was explicit in that he limited the
description of the customers he assisted outside
to woman and
children or babies. He made no mention of assisting any men from the
shop.
[159]
It was never completely clarified whether on the version of Mr Khote
or of Sgt. Naidoo, any other customers had remained inside.
In
response to defence counsel Sgt Naidoo said that he did not see
anyone left in the shop. It also seemed implied in the testimony
of
Sgt Naidoo that there was no-one left in the store – so that it
seemed unnecessary to ask him whether any other customers,
in
particular, male customers had remained in the store following on his
testimony.
[160]
However, the state conceded that the possibility remains that certain
male customers had indeed remained behind in the store.
The
significance of this would seem to be that it would open the
possibility of confusion to arise as to who amongst those who

remained in the shop, were perpetrators and who were customers. I
bear this possibility in mind, in favour of the accused as I
consider
the evidence to the effect that the perpetrators were indeed the two
accused presented before court.
[161] Mr Khote continued
in his evidence to the effect that shortly after that a police
officer arrived on the scene. He does not
know the name of this
police officer and significantly the state was unable to identify
this person or to present him as a witness.
The testimony of Mr Khote
describes a very circumscribed role to this officer who departed from
the scene - on the evidence of
Mr Khote and of the other two officers
(Alvers and Naidoo) who arrived shortly thereafter. No one was able
to identify the officer
and the defence rightly pointed to this as a
weakness in the state's case. I discuss this further below.
[87]
For present purposes it is worth noting that because the state was
unable to present the ‘unkown officer’, some of
the
evidence of Mr Khote is, to an extent, uncorroborated – as that
of a ‘single witness’.
[88]
[162]
The evidence of Mr Khote proceeds on the basis that once in this
unknown officer had arrived on the scene he accompanied this
officer
back into the store. As they entered the store they noticed, although
Mr Khote can only really speak for himself, that
upon entering he
noticed that two of the perpetrators when he recognised from having
engaged in a firefight, were standing towards
the rear of the store
but in front of what has come to be called the two large shelves at a
distance of perhaps 5 meters from the
door.
[163]
Mr Khote testified that he and the unknown officer approached the two
and took hold of them, brought them to the front of
the store and
made them lie down with the second (now incapacitated) perpetrator
who had fallen to the ground in between Mr Khote's
office and the
cashier's counter.
[164]
Mr Khote testified that shortly thereafter another two police
officers arrived on the scene including Sgt. Naidoo - in the
company
of Capt. Alvers.
[165]
Mr Khote concluded his evidence in chief by noting that after the
three were made to lie down in the front of the store they
were
attacked and assaulted by members of the community.
[166]
Capt. Alves remained outside to attend to the individual who had been
injured and was now lying outside.
SEGEANT NAIDOO
[167]
Sgt Naidoo approached Mr Khote and the unknown police officer who was
still on the scene. He testified that he was directed
to the three
individuals now lying motionless on the floor in the space between Mr
Khote's office and the cashier's counter.
[168]
He testified that these three individuals were taken into custody
that one subsequently died in hospital that the other two
had also
been transported to hospital but had been arrested on their discharge
and that, as he understood it, these were the two
individuals
presented before court as accused one and two.
ACCUSED’S
VERSION(S)
[169]
The version of the accused – to which they testified - is that
neither of them knew anything about the planned attack
in the store
on the day. They were not in any common purpose to kill or rob
anyone; and did not, individually do anything connected
with the
murder or robbery of anyone.
NECESSARY IMPLICATIONS
[170]
The version of the accused carries the following several necessary
implications, including:
[170.1]
there must be evidence (information) from which it may be inferred
that they were innocently in the store - which would
require that
they can account satisfactorily for their presence and conduct in the
shop.
[170.2]
It would not fall to them to disprove that they were in a common
purpose, that they did not rob or shoot anyone, nor that
each
unlawfully possessed a firearm and ammunition - since this would
require that they prove a negative - which they cannot do.
It will
only fall to them, if it appears that the state has presented a prima
facie case, to assume the risk that this case may,
upon final
evaluation, be regarded as proof beyond a reasonable doubt, or, if
they do not accept this risk, to place reliable information
before
the court from which a discernible version - hypothesis or picture -
emerges, which is consistent with innocence.
[170.3]
A1
needs it to be that Mr Khote is mistaken
as to what he saw A1 do in the store, and A2 needs it to be that both
Mr Khote and Ms Masango
are so mistaken; and
[170.4]
As discussed below,
[89]
A1 and
A2 were not the people who surrendered to Mr Khote (and the unknown
officer) and were taken to the front of the store and
made to lie at
the front of the store - instead, they were both innocent shoppers
who had somehow become unconscious while shopping,
and, in a frenzied
attack by the community, were mistakenly switched for the real
perpetrators.
[171]
The question which arises is whether this is
at all feasible – or, reasonably feasible?
BOTH ACCUSED
[172]
Both accused deny being part of any group which had robbed the Bread
Bin shop, including that neither one had killed anyone,
intended to
kill or rob anyone, nor was either ever in possession of an
unlicensed firearm which he was not permitted to possess.
[173]
In closing the defence pointed out that no attempt had been made to
establish in evidence that either of the accused - as
a matter of
fact - were not the holders of the required licensed.
[174]
The defence, on behalf of both accused, sought to rely on the
“failure” of Masango to identify either of them
in court.
[175]
Being as charitable as I can with the versions put by the accused -
because they were not put clearly and were contradicted
by themselves
or by other evidence admitted by them or otherwise relied upon –
I consider the versions of the accused to
the extent to which they
could be discerned.
ACCUSED ONE
[176]
Accused one (A1) testified that he works for a taxi transport
organisation as a security officer. He had been at work on the
day in
question and left work at around 16:00 hours in the afternoon.
[177]
Originally – on his evidence – he was due to go home.
However, he explained, that instead, he was then given the
task of
going to the area where the store is (the bread bin) to check whether
certain taxis were not operating in contravention
of the
organisation's rules.
[178]
He therefore travelled on one of the organisation’s taxis to
the road in which the bread than is situated and alighted
from the
vehicle.
[179]
Shortly thereafter he realised that he was short of mobile phone
airtime and decided to go and purchase airtime from the store.
[180]
He testified that once he had completed his task and called, he would
be collected by another of the taxi association’s
taxis which
would then transport him home from there.
[181]
He testified that he entered the store and stood in the queue to
purchase airtime from a cashier.
[182]
. Accused one then proceeded to describe how he was allegedly robbed
as he stood there waiting to purchase airtime. He testified
that a
man in a black jacket approached him and asked him if he was the
security guard. He denied that he was and the man then
set about
searching his pockets and removed from him his cell phone from one
pocket and R230-00 from his other pocket.
[183]
The accused was clear throughout that the man had made no threats nor
produced any form of weapon. The man then proceeded
to - in the words
of the accused - undress him of his jacket and set about stealing or
robbing him of his jacket.
[184]
In response to this he put up no resistance. He testified that after
removing his jacket the man instructed him to lie on
the floor and in
response to this he complied and lay on the floor. He thereafter
heard a bang or popping sound and lost consciousness.
[185]
The accused was, however, not consistent on the cause for him lying
down on the floor – he switched at one point to
explaining
that, he had lain down because of the loud bang or pop, instead of
being in response to the instructions of the man
who robbed him.
[186]
I understand from his testimony that he is alleging that the mark on
his forehead was incurred either at the moment that he
lost
consciousness or shortly thereafter.
[187]
He testified that he thereafter woke up in hospital. He testified
that once discharged from hospital a day or so later he
had indeed
been free to go and was surprised when the police arrived and took
him into custody.
ACCUSED TWO
[188]
The version of accused 2 (A2) begins with an explanation of his
receipt of some presumably second-hand clothing which
he was desirous
of two pass on to a person whose relationship with him or to him was
never made clear.
[189]
At best from his evidence it would seem that, in answer to the
question why he was passing on the clothing to this person,
he was
simply following orders which it seemed emanated somehow from his
family.
[190]
On his evidence he had agreed, with this unknown person, to meet at a
garage near the store at around 4 to 5 o'clock in the
afternoon on
the day in question. This rather extended period for meeting was
indeed the testimony of the accused that they were
due to meet
anytime between 16h00-17h00 in the afternoon.
[191]
The accused testified that after waiting for this other person for
quite some time he had become thirsty and decided to buy
a cool
drink. He decided then to cross the road - as he first put it - to
the bread bin to purchase a cool drink from there. He
subsequently
altered his version to explain that the store was not simply across
the road, but instead, across the road and some
distance down the
other side of the road.
[192]
To proceed with his version, he explained that he walked across the
street and went into the store to buy a cool drink.
[193]
He testified that when inside the store, he located the fridges and
took hold and uplifted the cold drink he sought to drink
and was then
about to turn to go and pay for his cool drink when  - he too -
heard a loud noise and lost consciousness.
[194]
He awoke in hospital and was also surprised to be subsequently
arrested. In response to all allegations against him he met
them with
a denial by virtue of the fact that he was intent merely on
purchasing a cool drink. He explained that since he had lost

consciousness he could not account for what had happened and was
therefore not accountable for what happened in the store that
day.
[195]
Insofar as the facts and versions are concerned, it is notable that,
in the same way as A1 was unable to account for anything
because he
explains that he had been unconscious at all relevant times, and
cannot explain what caused his unconsciousness, similarly
for A2.
Accused 2 is unable to account for anything because he too was
unconscious at all relevant times, and he too is unable
to explain
what caused his unconsciousness.
ANALYSIS
[196]
Thus, in essence, A1 explains that he (A1) was not part of any group
who attacked the store, but was instead innocently waiting
in line to
buy airtime;
[197]
A2 explains that he (A2) was also not part of any group who attacked
the store but was instead innocently buying a cooldrink.
[198]
I turn now to consider the feasibility of the versions put, with
particular reference to issues in dispute or otherwise of

significance.
CREDIBILITY AND
RELIABILITY OF THE WITNESSES
FOR THE STATE
[199]
Both Mr Khote and Ms Masango were utterly forthright in their
testimony. They were careful only to answer what had been asked
of
them
[200]
They did not hesitate in giving their answers and appeared cautious
to only speak of what they knew.
[201]
Although defence counsel attempted to make much of the fact that Ms
Masango had initially not identified either accused in
Court when she
first testified, it is undeniable that she had not been asked, and
that, when recalled and asked, she did not hesitate
in identifying A2
as the person who shot Mr Dlamini in front of her.
[202]
It was a pertinent moment when Ms Masango was challenged as to how
she could be certain that she is correctly identifying
A2 as the
perpetrator who shot Mr Dlamini. In response, she explained that she
was certain because A2 shot Mr Dlamini directly
in front of her. She
went on to state that his face is burned into her memory and she sees
his face when she closes her eyes.
[203]
I regard the testimony of Ms Masango and Mr Khote as credible and
reliable in all material respects.
[204]
In addition, I regard the testimony of the two officers who testified
about the events on the scene, Capt. Alves and Sgt Naidoo,
as
credible and reliable in all material respects.
[205]
The failure of the state to identify and to call the unknown police
officer must attract comment.
[206]
The state led the evidence of Capt. Alves to the effect that it is
not uncommon for an officer to assist and then disappear
- possibly
even because she or he wants to avoid the ‘admin’ that
would be required of him.
[207]
The question must arise whether he disappeared and was not called to
testify because his conduct was somehow improper. However,
what that
improper conduct could have been or what other nefarious purpose his
disappearance may serve was never seriously pursued
by the defence.
[208]
It must attract the sentiment that this is unacceptable conduct for
an officer, and it is a weakness in the state’s
case. It leaves
space in the state’s case for an interpretation to be placed on
the facts which is at odds with the version
which the state urged
upon this court. The court must be sensitive to the fact that there
is an aspect of the state’s version
that, if the possibility
arises, could and should be interpreted against it.
[209]
However, I can make no more of it. How he could have affected the
scenarios sketched and what it is that he may have done
wrong is left
to pure conjecture. In the end this court could find no way in which
to reconfigure the facts presented by the state
to include something
that this officer might have done, which could materially affect its
version.
[210]
The point here is to acknowledge it as a problem for the states
version and that this was considered.
[211]
Nevertheless, it is found to lead nowhere - even if I sought to
combine the possibility it raises with other possibilities
raised on
behalf of the accused.
[212] It also creates the
possibility that the identification by Mr Khote of A1 is single
witness testimony – and so must
attract caution. In some
respects the, the evidence of the state is reliant only on that of Mr
Khote, as a single witness.
[213]
However, our law permits for a conviction based on the evidence of a
single witness – or rather – a single ‘piece’

of evidence.
[90]
Nevertheless,
our jurisprudence has cautioned against unconsidered reliance on the
testimony of a single witness.
[91]
The point seems to be – again – not to suddenly shift
into a different mode of reasoning or to adopt a formalistic
approach
– but to exercise caution as will be dictated by common sense.
As reiterated in
S
v Sauls and Others
:
It
has been said more than once that the exercise of caution must not be
allowed to displace the exercise of common sense.
[92]
[214]
In any event – there does not appear, from the facts, to be any
facts on which the court is presented with
only
the evidence of Mr Khote. Even the identification of A1 is supported
by the logical inferences
[93]
– unless he was innocently lying unconscious in the shop and
was switched with a perpetrator, and as is required by this
logic,
switched back again, in a frenzied attack by the community – he
simply must be the relevant perpetrator.
THE ACCUSED
[215]
The Testimony of the accused was contrived and deliberate. Their
responses were slow and tortured – as they appeared
to attempt
to consider the implications of one answer over another.
[216]
They were determined to disagree with anything put to them by the
state to the point that, on occasion, was in line with their
own
testimony. Thus, they contradicted themselves – in a desperate
attempt to disavow anything the state suggested.
[94]
[217]
In
addition
, there was the attempt by both
A1 and A2 to suggest that they were not the people arrested on the
scene on that day - despite their
own admission to the contrary.
[218] I reject their
evidence as false and mendacious. I am mindful though that this does
not translate into a conviction for the
state. Mendacity
alone
,
on some issues does not – or ought not – to permit a
court to conclude that an accused must therefore be guilty.
[95]
An innocent accused may be a terrible witness and may even lie on
occasion for reasons other than that she or he is guilty.
[219]
It certainly does not assist the accused that their testimony does
not create reasonable doubt, but the question remains,
whether the
state has proved its case – beyond reasonable doubt.
[220] Mendacity may,
however, feature as a factor from which an inference may properly be
drawn – which together with other
considerations - may
establish guilt beyond a reasonable doubt.
[96]
SPECIAL CONSIDERSTIONS
FAILURE TO EXPLAIN OR
PROVIDE A VERSION PREVIOUSLY
[221]
As discussed above,
[97]
the
state urged this Court to take an adverse view of the failure of the
accused to have previously – before trial –
explained or
otherwise disclosed their version to the state.
[222]
As concluded, I understand our law to be that this would be an
infringement of an accused’s right to silence –
and that,
with respect, this is inevitably correct.
[223]
I cannot therefore make anything of the failure or refusal of the
accused to disclose their respective version before trial.
They were
thoroughly entitled to this silence and it is respected.
[224]
In any event I have found it unnecessary to take this failure
of theirs into account in arriving at my conclusion.
The evidence of
Mr Khote and Ms Masango together with the admissions made and other
probative material put to the court are utterly
adequate.
[225]
As much as the testimony of the accused is unpersuasive, several
points were raised which require attention.
THE ‘MARK’
[226]
A1 made much of the reference, in the testimony of Mr Khote, to the
mark on his forehead - in identifying him.
[227]
The accused claimed that any identification of him which relies on
the appearance of a ‘mark’ on his right upper
temple. He
disputed this because he claims that the ‘mark’ was only
inflicted on the day in question - and so, it is
not a basis on which
the owner could identify him.
[228]
Several observations must be made of this line of reasoning:
[228.1] The only basis
for the claim is the claim itself - made by the accused. Nothing
supporting the claim was placed before court,
in particular, no
medical report of an injury suffered on the day.
[228.2] The
identification by Mr Khote of the accused by reference to the scar,
was in response to a follow up question regarding
by what features he
identifies the accused.
[228.2.1.]
Not only was it a follow up question - but it, after having left it
off of his account of the basis on which he recognised
the accused,
the defense counsel asked him if he had any marks. Mr Khote then
agreed that the accused had a mark on his face -
on his forehead.
[228.2.2.]
Before this - when asked how he can identify A1, he said: "by
his size, his face, everything." Asked by what
features he can
identify A1, he said: "by his round head, he is bald, and is
fat".
[228.2.3.]
Only after that was he asked the question whether he knew of any
marks on the accused.
[228.2.4.]
He was then asked whether it was new or old - to which he responded
that he does not recall.
[228.2.5.]
He was asked when he noticed it - to which Mr Khote responded that he
had noticed it when the accused was lying down
in the front of the
store.
[229]
At its best (for the defence), Mr Khote indicated that, one of the
features by which he knows the accused is the mark on his
head. He
was not claiming, in his testimony that he is only able to recognise
the accused because of the mark.
[230]
But there doesn't seem to be any sensible construction on the basis
of which it might be suggested that Mr Khote did only
or could only
recognise A1 because, and only because of the mark.  It had, in
fact, to be extracted from the witness by suggestion
- that the
accused (A1) does have a mark on his forehead and that it was there
on the day.
[231]
He was unequivocal in his testimony that he recognised the accused
because he saw his face as he engaged with him in an exchange
of gun
fire. In my view it is mischievous to attempt to reduce the
identification by Mr Khote of the accused to a complete reliance
on
the mark.
COMPLEXION
[232]
For A2 it is critical that we distinguish him as light in
complexion whereas Miss Masango was incorrect in her identification

because she said that he was dark in complexion.
[233]
I know of no basis on which to distinguish validly and reliably
between degrees of darkness or lightness of complexion. I
can imagine
that what for one person amounts to a dark complexion could easily
amount for another to be light in complexion.
THE LEG
[234]
A2 submits that Mr Khote is wrong to the extent to which he suggests,
in his testimony, that he shot A2 in the left upper
thigh.
[235]
He is indeed correct in this assertion - which was conceded by the
prosecution: A2 was not shot in his left upper thigh.
[236]
However, this error does not appear to carry the weight which A2
places on it. A2 argues that since he was not shot in the
leg, it
cannot have been him - in so far as Mr Khote relied on having shot
him in the leg in order to identify him. Regrettably
for the accused,
(A2), this does not follow.
[237]
The ready explanation to this is only that Mr Khote must have been
mistaken that he had in fact struck the accused (A2) in
the leg when
firing at him. If we expected of people to be able to observe the
travel of a bullet which is discharged from a gun,
there may have
been something in this. But we do not expect this of anyone.
[238]
It was also not put to Mr Khote that he is incorrect in this respect
and he was not given an opportunity to explain why he
testified that
he had struck the accused (A2) in the leg.
[239]
This failure can, in itself, be fatal - except that I do not think
anything can be made of this anyway.
THE BEANIE
[240]
A2 points out that he cannot be guilty because Ms Masango said
whoever killed Mr Dlamini - who she says is him - was wearing
a
beanie.
[241]
We are informed by the accused (A2) that he was not wearing a beanie
on the day.
[242]
Perhaps in an alternative universe, this sort of defence may carry
some weight, but I regret that his insistence about the
accessories
he wore must count for nothing.
[243]
Even so, and without meaning to imply that this claim could ever
amount to anything, several ‘beanies’ were found
on the
floor of the store - abandoned or lost. These appear in the photo
album admitted into evidence by both accused.
THE SUBSTITUTION
[244]
In a scarcely discernible submission, defence counsel argued that it
is not clear that the two people taken into custody by
Mr Khote (with
the assistance of the unknown officer), are the two men who were
presented in court as accused 1 and 2.
[245]
In this way, defence counsel is understood to have argued that
somewhere between Mr Khote taking control of the two, and their

presentation in court, the real perpetrators were somehow mixed up
with the two presented in court.
[246]
There appears to be two independent suggestions here - or the
suggestion of two distinct opportunities for this ‘substitution’.
[246.1] That between
taking control of the perpetrators (as they surrendered), being made
to lie down in from of the store (with
their wounded co-perpetrator),
and when they were taken into the custody of the police and
transported to hospital, the real perpetrators
had been mistakenly
switched for others from amongst the men who may have remained in the
store or other community members - and
that this switch would have
occurred during the attack by the community on the perpetrators and
when they lay at the front of the
shop; and
[246.2] That between when
they were taken into custody (from when lying in the front of the
shop) to their presentation in court,
the perpetrators had been
switched with the accused so that the accused are not the men taken
into custody in the front of the
store.
[247]
There are several problems with the submission, and it reveals a
desperate attempt to distract this court. It is so strained
that it
extends over the line of being an absurd and improperly founded
suggestion, to being in direct contradiction to a
section 220
admission recorded by the accused.
[248]
Starting backward - there can be no merit in the suggestion that the
accused were mixed up and are not the people who were
arrested on the
scene - simply because they have admitted it. But the pursuit of this
line of argument does damage to the credibility
of everything they
have put to this court. They have shown that, in an attempt to evade
responsibility, they will do whatever they
can, including changing
their version and even contradicting their very own admissions.
[249]
Also, even putting aside the preposterous suggestion of a frenzied
community attack in which the real perpetrators were mixed
up with
other innocent people present, the suggestion requires that the
eyewitness testimony of Mr Khote that these two are indeed
the two
perpetrators who he took under control when they surrendered, and the
eyewitness testimony of Ms Masango confirming that
A2 is the person
who shot and killed Mr Dlamini - both of whose testimony is beyond
reproach.
[250]
But there is another dimension which one is prompted to consider when
considering hypothesis which must be complete and internally

consistent to be feasible.
[251]
If it is common cause that the perpetrators were taken under control
by Mr Khote and the unknown officer, and made to lie
at the front of
the store, but the accused version is that:
[251.1] They fell
unconscious, A2 at the fridge and A1 at the counter, then one must
wonder how they were also conscious members
of the community -
perhaps a remaining customer - who was now, in the attack by the
community, mixed up with the ‘real perpetrators’.
[251.2] The suggestion -
when subjected to the abductive model - must include the submission
that the fury of the community fell
on the accused, as they lay
unconscious in various parts of the store, and that the community who
set upon the unconscious perpetrators
who were lying at the front of
the store, missed their target and attacked these two unconscious
innocent shoppers instead, dragging,
at least A2, from the cooldrink
fridge (when he fell unconscious) to the place at the front of the
store - ejecting the real perpetrator
for A2. This implication was
not even considered, nor any feasible suggestions made as to the how
these necessary implications
may be explained.
[252]
For A1, the implications of his version and the submission that he
was originally - on his own account - lying unconscious
at the front
of the store - then the community attacked the perpetrators lying at
the front of the store, and switched them out
for innocent
bystanders. But on this account - his own account - he would have
been switched out for an innocent member of the
public. This is
because, the accused requires that the people who were lying at the
front of the store - were switched. However,
he wants it to be the
case that he was both switched and not switched. Of course, this
cannot be. Alternatively, he needs it to
be that the frenzied members
of the community committed a double switch: they switched him out for
the real perpetrator, or, perhaps,
someone else, then, somehow,
switched him back again.
[253]
These suggestions must be rejected as mischievous fabrications.
FEASIBILITY OF VERSIONS
[254]
In light of the above, I turn now to consider which version, as
between the state and the respective accused, offers the best

explanation of the facts.
STATE
[255]
One may recall that the state’s version is as follows:
[255.1]
The states version is that the two accused, acting in a group of five
originally, acting in the execution of a common purpose
to murder the
security guard and to rob him and then to rob the store, did murder
and rob the security guard, and, when they then
encountered
resistance from the store owner (Mr Khote), exchanged gunfire with Mr
Khote.
[255.2]
Two of their own members were killed in the shootout and another was
wounded. Mr Khote was also wounded. The two remaining
perpetrators
surrendered, were taken into custody, and together with the third
wounded perpetrator, were transported to hospital.
[255.3]
The wounded perpetrator died in hospital and the remaining two were
then, upon their discharge from hospital, transported
to a police
station, where they were charged.
[255.4]
In addition, the version of the state is that each of them was in
unlawful possession of a firearm and ammunition.
[256]
The evidence presented makes this version clear and consistent
and it appears to indeed represent the best explanation of the facts

that may be discerned.
[257]
I turn now to consider whether the respective versions of the accused
are at all feasible – at least to the point that
they raise a
reasonable doubt as to their guilt.
ACCUSED 1
[258]
In considering the veracity of the version of A1, putting aside for
the moment the states version and the direct evidence
against the
accused, one may ask, engaging in what may – if we wanted to
attach a label – inferential reasoning: what
are the chances
that A1 would happen to stop at that store to buy airtime, given
that:
[258.1]
His work for the day, at Clearwater mall was complete - and he was
due to go home;
[258.2]
He was then assigned the duty of going to inspect whether taxis were
adhering to the association’s rules at a place
near the shop;
[258.3]
That he travelled there by taxi but needed to call for a taxi to be
transported home; and
[258.4]
That he would presumably need airtime to communicate his findings
regarding the taxis he observed in the vicinity; and
[258.5]
Then only noticed that he needed airtime?
[259]
I doubt this. But it is, in itself, inconclusive. However, that is
not the full hypothesis – it is not what the accused
needs the
court to believe. Instead, we must consider the chances that A1 was
robbed as he stood innocently in the queue to purchase
airtime, given
that:
[259.1]
There is no other evidence of a robbery of any individual on the day
- other than the security guard - and only of his firearm;
[259.2]
A1 is a ‘security guard’ for a taxi association;
[259.3]
A1 stood silently and passively as he was literally undressed by the
alleged robber;
[259.4]
That he would be confused as to what made him lay down on the day -
an instruction from the robber, the sound of gunfire,
or a
combination of both.
[260]
This is the version to which the accused testified, but it strains
credulity.
[261]
Ultimately, we must consider whether the
information we have may be explained by the version of the accused –
as a reasonably
possible version. For this we must believe his
version of:
[261.1]
innocently standing in the store to buy airtime,
[261.2]
standing passively while he is robbed;
[261.3]
being confused about why he lay down;
[261.4]
being unable to account for how he lost consciousness; and
[261.5]
being unable to account for anything that happened after that;
[261.6]
Mr Khote is mistaken in his identification of the accused with whom
he exchanged gunfire and who he took control of once
the accused
surrendered; and
[261.7]
He was mixed up with an actual perpetrator during a frenzied attack
by the community upon the real perpetrators.
[262]
This explanation is incredulous, and, when considered with the
finding that his testimony was generally mendacious
[98]
and lacked credibility, I am unable to entertain it as reasonably
possibly true.
ACCUSED 2
[263]
Similarly, for
A2
in respect of his
presence in the shop. One may ask - what are the chances that:
[263.1]
He would have received a bundle of clothes the quantity of which he
is unclear;
[263.2]
That he would be required by someone - whose identity or basis for
authority over him is unknown to him - to deliver the
clothes to
someone;
[263.3]
That the person to whom the clothes were to be delivered was and is
unknown to him;
[263.4]
That a meeting time of between 16h00-17h00 would be specified;
[263.5]
That he would choose to cross the road and walk some distance down
the opposite side of the road to buy a cool drink when:
[263.5.1.]
Doing so may have meant missing the person he was there to meet; and
[263.5.2.]
There was a store at the petrol station which was appointed as the
meeting destination. Although the accused explained
that in his
experience the cool drinks were more expensive at petrol stations -
he made no attempt to establish the price; and
[263.5.3.]
There was a second store, a ‘Muchachos’, directly next to
the filling station and he could not explain why
he did not even
consider purchasing a cooldrink from that store.
[263.6]
He experienced the terrible misfortune of inexplicably losing
consciousness at the critical time – and being then
unable to
account for what happened in the store.
[263.7]
Ms Masango is mistaken in identifying him as the person who shot Mr
Dlamini;
[263.8]
Mr Khote is mistaken that he was one of the perpetrators; and
[263.9]
He was also mixed up with the real perpetrators in a frenzied attack
by the community.
[264]
Again,
this
explanation strains credulity, and when considered with the finding
that his testimony was generally mendacious
[99]
and lacked credibility, again, I am unable to regard it as raising a
reasonable doubt as to his guilt.
IDENTIFICATION
[265]
Our courts have cautioned that one must be circumspect in relying on
the identification of an accused by a witness –
in virtue of
the natural fallibility of human observation.
[100]
[266]
The identification of the two accused as perpetrators from the day in
question and more particularly, what it is that they
did on the day,
does depend, to an extent, on the dock identification of A1 by Mr
Khote, and of A2, by Mr Khote and Ms Masango.
[267]
As indicated, I find the evidence of Ms Masango and Mr Khote to be
credible and reliable in all material respects.
[268]
As discussed above, it is immaterial whether A2 was indeed wounded in
the leg and whether he was wearing a beanie at some
point. It is
equally immaterial whether A1 already had the ‘mark’ on
his forehead, or not.
[269]
But the evidence against them goes far beyond the eyewitness
identification by Mr Khote and Ms Masango. It relies on their
own
admission that they were arrested on the scene, that Mr Khote
testified that he took these two accused and made them lie down
in
the front of the store, and the incomprehensible ‘switch’
which they propose.
[270]
On a conspectus of all the evidence, and in particular the extreme
improbabilities inherent in what A1 & 2 require to
be true for
their version to raise a reasonable doubt as to their guilt, I cannot
accept their version as reasonably possibly true.
Instead, I accept
the version of the state in all material respects and am of the view
that it establishes guilt on the part of
the accused beyond any
reasonable doubt.
COMMON PURPOSE
[271]
The most straightforward conduct that may be attributed to every
member of the group is the conduct of the accused in the
killing of
Mr Dlamini, the security guard. Significantly, the inference is
inescapable, given that it was the way in which the
attack was
initiated, that the common purpose included the express mandate that
the security guard would be killed. This makes
the murder of Mr
Dlamini premeditated – and since it was part of the express
mandate, this premeditated murder is attributable
to all in the
common purpose.
[272]
The second is the taking by force and thus the robbery of Mr Dlamini
of his firearm. At the very least this conduct is attributable
and is
attributed to accused one and two. Both accused one and two engaged
in a firefight with Mr Kothe and by the operation of
common purpose
they may have the conduct of each other attributed to each one so
that while both attempted to kill Mr quarter,
on an application of
common purpose, properly, they both attempted to kill Mr Kothe on two
separate occasions.
[273]
Furthermore, by an application of common purpose, they are liable for
the murder of their co-perpetrators, whose deaths they
must have
foreseen.
FIREARMS AND AMMUNITION
[274]
In the current matter, there is ample evidence – referred to
above - that each of the participants were in actual possession
of a
firearm. It is on this basis that a finding that each was in unlawful
possession of a firearm is made and it is, in my view,
a failsafe
method of reasoning.
[275]
It is not clear who was in possession of the firearm from which the
serial number had been obliterated. However, since such
a firearm
cannot have been licenced, being in possession of that firearm, would
also, necessarily, make whoever was in possession,
guilty of being in
possession of an unlicensed firearm.
[276]
It should be recalled that defence counsel argued that the state had
failed to prove that the accused did not hold the required
licences
for any firearm they may be found to have possessed, and that, in
addition, it was asserted on behalf of the accused that
they did not
have the requisite knowledge – for sentencing purposes –
that any such firearm, as they may be found to
have posed, was a
semi-automatic.
[277]
On the argument that the state had failed to prove that the accused –
if found to have possessed firearms – were
not licenced –
the state relied on
section 250
of the CPA (discussed above).
[101]
[278]
The accused did not produce any licences in accordance with the
evidentiary burden placed on them by
s 250.
It follows that this
court must conclude that they were in unlawful possession of the
firearms.
[279]
The proposition that they did not know that these firearms were
semi-automatic requires that the court gives them the benefit
of any
doubt when they have denied even possessing a firearm. The defence
raised is one which is peculiarly within the knowledge
of the accused
– and requires that the accused take the court into his/her
confidence and explain how he or she could be
mistaken.
[280]
In
S v Pistorius
the unanimous SCA made it clear that one
cannot prevaricate. On the accused’s defence of putative
private defence, it said:
The
immediate difficulty that I have with the accused’s reliance
upon putative private defence is that when he testified,
he stated
that he had not intended to shoot the person whom he felt was an
intruder. This immediately placed himself beyond the
ambit of the
defence, although as I have said, his evidence is so contradictory
that one does just not know his true explanation
for firing the
weapon.
[102]
[281]
In addition, the suggestion would give rise to an absurd situation on
the facts: the accused would both – if they laboured
under the
misapprehension that the firearms were not semi-automatic –
that is, single shot, self-reloading firearms, would
have presumably
stopped after firing a shot to reload their firearms. Yet, there is
nothing in the facts to support this –
more especially, as
indicated, nothing from the accused themselves on this point.
CONCLUSION
[282]
The testimony of the accused is simply not credible and the
respective versions they present are incredulous.
[283]
I am aware that strange things happen, but I am utterly unable to
even imagine a universe in which the versions required by
the accused
to be accepted as reasonably possibly true - to the exclusion of the
state’s case - could ever obtain.
[284]
Even adopting the greatest caution to be sensitive to a feasible
innocent explanation, I am compelled instead to adopt the
version of
the state – as the only feasible explanation for the
information that may be discerned. In this context the only
feasible
explanation translates into proof beyond a reasonable doubt against
the accused.
[285]
In conclusion the question arises whether the accused have done
anything to disturb the version of the state being accepted
as the
best possible explanation or otherwise that they have, at least,
raised a reasonable hypothesis which must be acknowledged

irrespective of the State's version.
[286]
The answer to this must be that the state has indeed sketched a
coherent version which explains all the evidence and
in respect of
which, no inconsistencies appear. It is internally consistent,
utterly feasible, and is supported in all material
respects by the
evidence.
ORDER
[1]
IT
IS THEREFORE ORDERED THAT:
[1.1]
Both accused one and two are guilty as charged, of:
[1.2]
One count of premediated murder in respect of the death of Mr
Dlamini;
[1.3]
Two counts of murder – in respect of their two co-participants;
read
with the provisions of
section 51(1)
and (2) as well as schedule 2 of
the
Criminal Law Amendment Act 105 of 1997
and further read with the
provisions of sections 92(2), 256 and 258 of the Criminal Procedure
Act 51 of 1977 (hereinafter referred
to as the ‘CPA’);
[2] Attempted murder- in
respect of the shots fired at Mr Khote - read with the provisions of
sections 92(2), and 258 of the CPA
further read with the provisions
of section 51(2)(c) and schedule 2 of the
Criminal Law Amendment Act
105 of 1997
;
[3] Robbery with
aggravating circumstances – in respect of the violent taking of
the firearm from Mr Dlamini, as defined in
section 1
of the CPA -
read with the provisions of
section 51(2)
and schedule 2 of the
Criminal Law Amendment Act 105 of 1997
and further read with the
provisions of
sections 92(2)
,
256
and
260
of the CPA;
[4] Possession of an
unlicensed firearm, in contravention of
section 90
, read with the
provisions of
section 1
,
103
,
117
,
120
(1),
121
and schedule 4., of the
Firearms Control Act 60 of 2000
, and further read with the provisions
of
section 250
of the CPA.
[5]
The unlawful
possession of ammunition, in contravention of
section 90
, read with
the provisions of
section 1
,
103
,
117
,
120
(1),
121
and schedule 4, of
the
Firearms Control Act 60 of 2000
, and further read with the
provisions of
section 250
of the CPA.
__________________________________________
GRANT
AJ
JUDGE
OF THE HIGH COURT OF SOUTH AFRICA
GAUTENG
LOCAL DIVISION, JOHANNESBURG
COUNSEL
FOR THE STATE: ADV RUBIN
INSTRUCTED
BY: THE NATIONAL PROSECUTING AUTHORITY
COUNSEL
FOR ACCUSED: ADV MOSOANG
INSTRUCTED
BY: LEGAL AID SOUTH AFIRCA
DATE
OF HEARING: 29 MAY 2019
DATE
OF JUDGMENT:
18 JUNE 2019
[1]
See page 37ffDT Zeffertt & AP Paizes
The
South African Law of Evidence
3rd ed (2017).
[2]
See chapter 2: ‘The South African evidentiary system and
theories of proof’, in ibid, at p 31ff.
[3]
See, in particular, pages 34ff; ibid. One such example which reveals
the conflicting decisions that may arise from a
deductive/mathematical
mode of reasoning relative to an intuitive
mode is included in the quote below.
[4]
Ibid page 37-8.
[5]
See
John Dewey,
Logical
Method and Law
,
10 Cornell L. Rev. 17 (1924).
[6]
I acknowledge here that in some respects our mental processes are
beyond our comprehension, either as unconscious or unfathomable.
[7]
Sidney N. Lederman, Alan W. Bryant & Michelle K. Fuerst
Sopinka,
Lederman & Bryant: The Law of Evidence in Canada
5 ed (2018) page 72.
[8]
See the discussion under ‘Circumstantial Evidence’ of
the reliance by our courts on the analogy of a mosaic –
at
paragraph [86]ff.
[9]
See CIRCUMSTANCIAL EVIDENCE
on
page 27ff.
[10]
One
may notice that, in attempting to extract and isolate the logical
form of an argument, the premises as stated above the line,
and the
conclusion (which is supposed to follow), is stated below the line.
[11]
Most closely affiliated with Pascal - Zeffertt & Paizes
The
South African Law of Evidence
3rd ed (2017) at 34ff.
[12]
Most closely affiliated with Bacon - ibid at 34ff.
[13]
This is, I acknowledge, to attach a label to what we do – or
ought to do. I also acknowledge the danger of unthinkingly
attaching
a label expecting that in doing so a problem is thereby solved. I
mean however to attach the label deliberately to
attract to what it
is we do or ought to do, the attributes which have come to be
associated with abductive reasoning –
especially in its
attributes discussed below. If, however, we need, by the dictates of
law or practice, to deviate from what
ordinarily follows on
abduction, then we cannot allow a label to prevent us from doing
what is necessary.
[14]
Widely credited with formalising this form of logic (C.S. Pierce
Writings
of C.S. Peirce: a Chronological Edition,
Vol 1 (1865)). See, for instance, Giovanni Tuzet 'Legal Abductions'
In
Legal
Knowledge and Information Systems: Jurix 2003
edited by D. Bourcier 41-9 (2003); Lipton, P. (1991),
Inference
to the Best Explanation
,
Routledge, London. Pierce wrote prolifically and both introduced and
developed the concept of abduction and what it must be
understood to
mean over nearly half a century – from about 1863 virtually
until his death in 1914. Despite the prominence
given to Pierce, I
acknowledge that the adoption of abduction is not to tie us
slavishly to whatever he said. It is to rely on
a rich source of
thought on how we reason. If reason dictates that we depart from
what Pierce has said, then so it must be.
[15]
C.S. Pierce
Writings
of C.S. Peirce: a Chronological Edition,
Vol 3 (1878) 323-38; Pierce
Writings
of C.S. Peirce: a Chronological Edition,
Vol 1 (1865) 180, 266; C.S. Pierce
Writings
of C.S. Peirce: a Chronological Edition,
Vol 1 (1866) 362, 430.
[16]
There is, in this a semblance of the what has come to be known as
the second of the ‘cardinal rules of logic’ from
the
case of
R
v Blom
1939 AD 188.
This is perhaps no accident – as submitted (see
[20] and [21]), the mode of reasoning proposed is to treat all
evidence
as circumstantial evidence.
[17]
C.S. Pierce
Writings
of C.S. Peirce: a Chronological Edition,
Vol 2 (1982) 22. Peirce, C. S. ‘On the Logic of drawing
History from Ancient Documents especially from Testimonies’

(1901),
Collected
Papers
v.
7, paragraph 219. C. S. Pierce, A Letter to F. A. Woods (1913),
Collected
Papers
v.
8, paragraphs 385–388.
[18]
Tuzet 'Legal Abductions' In
Legal
Knowledge and Information Systems: Jurix 2003
edited by Bourcier 41-9 (2003) 45.
[19]
Peirce, Charles S. (1901). The Proper Treatment of Hypotheses: a
Preliminary Chapter, toward an Examination of Hume's Argument

against Miracles, in its Logic and in its History.
[20]
Pierce stated – in a criticism of Descartes (René
Descartes 'Sixth Meditation' In
Meditations
on First Philosophy
(1641/1996)) method for confirming existence (which came to be known
as
Cartesian
Dualism
):
‘
We
cannot begin with complete doubt. We must begin with all the
prejudices which we actually have when we enter upon the study
of
philosophy. These prejudices are not to be dispelled by a maxim, for
they are things which it does not occur to us
can
be questioned. Hence this initial skepticism will be a mere
self-deception, and not real doubt; and no one who follows the

Cartesian method will ever be satisfied until he has formally
recovered all those beliefs which in form he has given up. It is,

therefore, as useless a preliminary as going to the North Pole would
be in order to get to Constantinople by coming down regularly
upon a
meridian. A person may, it is true, in the course of his studies,
find reason to doubt what he began by believing; but
in that case he
doubts because he has a positive reason for it, and not on account
of the Cartesian maxim. Let us not pretend
to doubt in philosophy
what we do not doubt in our hearts.’ (Some Consequences of
Four Incapacities,
Journal
of Speculative Philosophy
(1868) 2, 140-157, at 140.)
[21]
Pierce explained: ‘… it is no longer the reasoning
which determines what the conclusion shall be, but it is the

conclusion which determines what the reasoning shall be. This is
sham reasoning.’ C.S. Pierce
Collected
Papers of C.S. Peirce
Vol 1 (1931-1958) 57.
[22]
See note 5 and associated text.
[23]
A. Jøsang
Subjective
Logic: A Formalism for Reasoning Under Uncertainty
,
Springer 2016.
[24]
See
note 20.
[25]
See note 20.
[26]
People
v. Collins
,
68
Cal.2d 319
,
438 P.2d 33
(1968).
[27]
See paragraph [94]ff.
[28]
See paragraph [26]
above.
[29]
Zeffertt and Paizes state: ‘
An
election, in short, must be made between these two schools of
thought. It is unlikely that either of the approaches is universal,

which means that different problems may have to be solved in
different ways, here with a little Bacon, and there with a touch
of
Pascal.’ (
Zeffertt
& Paizes
The
South African Law of Evidence
3rd ed (2017) 38)
[30]
Contained in the quote at paragraph [12]
.
[31]
In its most basic form, it would run as follows: from blood
collected at the scene of a crime (leaving aside issues of chain

evidence and other ancillary issues), it is established that the
perpetrator has a rare blood characteristic: only, let’s
say,
1% of people share this characteristic. The accused has this blood
characteristic. Let me call these bits of information
“the
initial premises”. The (fallacious) argument may be made that,
therefore, we can be 99% certain that the accused
is the
perpetrator. The flaw in the logic lies in looking at the statistic
in isolation – instead of considering, as one
must, who else
shares this blood characteristic? In a population of 10 million, it
is 1% of 10 million: 100 000 people. Thus,
the flaw is revealed that
in truth, the probability of the accused being the perpetrator is
1/100 000. Thus, whereas the probability
that the accused is the
perpetrator, is proposed, in error, to be 99%, it is, instead,
0.00001%. The “trick” to avoiding
the prosecutor’s
fallacy is to take account of the “big picture”. More
information is required before one can
make anything of the data
contained in the initial premises.
[32]
In
S
v Stevens
1961
(3) SA 518
(CPD) at pages 518H to 519C;
S
v Papiyana
1986
2 PH H115 (A) at 206;
S
v Mthembu and Others
1988
(1) SA 145
(A) at 155G-H;
S
v Maputle
2002
(1) SALR 550
(WLD) at 552;
S
v Phiri
(2033/05)
[2005] ZAGPHC 38
, particularly at paragraph 15ff.
[33]
See note 20. Even social psychology and in particular in the
critical schools of thought this method is known as ‘bracketing’.
[34]
It
appears that our Courts have adopted the approach pronounced by
Denning J in
Miller
v Minister of Pensions
([1947)
2 All ER 372
(KB) at 374ff): ‘“we think it more probable
than not”, the burden is discharged, but if the probabilities
are
equal it is not.’
[35]
In
Ex
parte Slabbert
&
Prinsloo:
Re R v Slabbert
&
Prinsloo
1944
TPD 327
at 331.
[36]
1944
AD 314.
[37]
At
page 320ff.
[38]
At
page 320ff.
[39]
On page 45ff.
[40]
R v
Blom
1939 AD 188.
[41]
Ibid,
p
ages
202-203.
[42]
Zeffertt & Paizes
The
South African Law of Evidence
3rd ed (2017) at 103.
[43]
R v
Blom
1939 AD 188
, p
age
201.
[44]
Watermeyer AJ, ibid, p
age
201.
[45]
At page 205.
[46]
This circularity was argued by Counsel (Snitcher) for the accused
(see page 198) – but ignored.
[47]
Zeffertt & Paizes
The
South African Law of Evidence
3rd ed (2017) 103ff.
[48]
Chamberlain
and Another v The Queen [No 2]
1984 153 CLR 521 536.
[49]
See, in particular
Shepherd
v The Queen
[1990] HCA 56
;
(1990) 170 CLR 573
270ff.
[50]
R v De
Villiers
1944 AD 493
508-9;
S
v Cwele and another
2013 (1) SACR 478
(SCA) paragraph [19].
[51]
S v
Hadebe
1997 (2) SACR 641
(SCA). The Constitutional Court also adopted this
mode of reasoning in
S
v Molimi
[2008] ZACC 2
;
2008 (3) SA 608
(CC) at paragraph 51 – there holding that the
mosaic
was not sufficiently complete.
[52]
Moshephi
and Others v R
(1980-1984) LAC 57.
[53]
Moshephi
and Others, a
t
59F-H
[54]
Where Q may be anything that requires proof for any conclusion to
follow - even if that conclusion is a secondary fact, upon
which a
further inference must be drawn.
[55]
Proved at whatever the appropriate standard is for each item. Again,
here, we must observe that nothing inconsistent with the
inference
sought to be drawn may form part of the ‘givens’ - the
basis for drawing the inference.
[56]
See paragraph [86]ff.
[57]
Where common purpose by prior agreement is not implicated.
[58]
Common
purpose may be formed by prior agreement or active association (
S
v Thebus
[2003] ZACC 12
;
2003 (6) SA 505
(CC);
S
v Mgedezi and Others
1989 (1) SA 687
(A);
S
v Safatsa and Others
1988
(1) SA 868
(A)
)
- each of which is subject to specific rules as to whether
individuals are to be regarded as being part of a “common
purpose group”.
[59]
CR Snyman
Criminal
Law
6th ed (2014) 255-64; Jonathan Burchell
Principles
of Criminal Law
5th ed (2016) 467-93; Andrew Paizes 'Common Purpose by Active
Association: Some Questions and Some Difficult Choices' (1995)
112
SALJ; Andrew Paizes ''Mistake as to the Causal Sequence' and
'Mistake as to the Causal Act': Exploring the relation between
Mens
Rea and the Causal Element of the Actus Reus.' (1993) 110 SALJ.
[60]
S v
Thebus
[2003] ZACC 12
;
2003 (6) SA 505
(CC);
S
v Mgedezi and Others
1989 (1) SA 687
(A).
S
v Makhubela and Another
2017 (2) SACR 665
(CC);
S
v Jacobs And Others
2019
(1) SACR 623 (CC).
[61]
See
S v
Nhlapo and Another
1981
(2) SA 744
(A).
[62]
S v
Nkombani and Another
1963
(4) SA 877 (A).
[63]
Rumpff JA (Holmes JA concurring, Steyn CJ dissenting) stated:
According to the facts, in sofar as it relates to the robbery of

garages on that particular evening, [X] had
dolus
indeterminatus
in
the sense that he foresaw the possibility that anyone who was, or
could have been, involved in the robbery, could have been
killed by
the revolvers of the robbers, and he was indifferent as to who could
die (own translation, original emphasis; ibid
p 892).
[64]
S v
Nhlapo and Another
1981
(2) SA 744 (A).
[65]
Bold emphasis added, ibid p 750-1.
[66]
James Grant 'Common Purpose:
Thebus
,
Marikana and Unnecessary Evil' (2014) 30 SAJHR.
[67]
Recognised in
S
v De Blom
1977 (3) SA 513
(A).
[68]
Snyman
Criminal
Law
6th ed (2014) 63.
[69]
S
v Nkosi
1998
(1) SACR 284 (W).
[70]
A
t
page 286H – I.
[71]
S v
Kwanda
2013 1 SACR 137 (SCA).
[72]
2017 (2) SACR 665 (CC).
[73]
Mhlantla J (Mogoeng CJ, Nkabinde ADCJ, Cameron J, Froneman J, Jafta
J, Khampepe J, Madlanga J, Mojapelo AJ, Pretorius AJ and
Zondo J
concurring).
[74]
Emphasis added; paragraph 45-47
[75]
2018
(1) SACR 522 (WCC).
[76]
At paragraph 132.
[77]
Which provides: ‘
If
a person would commit an offence if he-
(a)
carried on any occupation or business;
(b)
performed any act;
(c)
owned or had in his possession or custody or used any article; or
(d)
was present at or entered any place,
without
being the holder of a licence, permit, permission or other authority
or qualification (in this section referred to as
the
'necessary
authority'
), an accused shall, at criminal proceedings upon a
charge that he committed such an offence, be deemed not to have been
the holder
of the necessary authority, unless the contrary is
proved.’
[78]
S
v Fransman
2000
(1) SACR 99
(W).
[79]
As per the ultimate test accepted in
S
v Coetzee and others
1997 (1) SACR 379 (CC).
[80]
2003 (6) SA 505 (CC).
[81]
At paragraph. 56.
[82]
S v
Thebus
2003 (6) SA 505 (CC).
[83]
Goldstone, O'Regan, Ackermann and Mokgoro JJ.
[84]
Moseneke J, Chaskalson CJ and Madala J.
[85]
As
discussed further in paragraph, under the heading, “
THE
SUBSTITUTION”
on
page 68ff.
[86]
Under section s167 of the CPA, which provides: ‘
The
court may at any stage of criminal proceedings examine any person,
other than an accused, who has been subpoenaed to attend
such
proceedings or who is in attendance at such proceedings, and may
recall and re-examine any person, including an accused,
already
examined at the proceedings, and the court shall examine, or recall
and re-examine, the person concerned if his evidence
appears to the
court essential to the just decision of the case.’
[87]
Discussed below under the heading ‘CREDIBILITY AND RELIABILITY
OF THE WITNESSES’/FOR THE STATE on page 60ff.
[88]
Discussed below in paragraph [212]ff.
[89]
Under the heading ‘THE SUBSTITUTION’ on page 68ff.
[90]
Section 208 of the CPA provides: ‘
An
accused may be convicted of any offence on the single evidence of
any competent witness.’
[91]
See
R v
Mokoena
1932
OPD 79
at page 80.
[92]
1971 (3) SA 754
(A) at page 758
[93]
Discussed in more detail below, under the heading ‘THE
SUBSTITUTION‘ on page 68ff.
[94]
On one such occasion, A1 denied being anywhere near the front of the
cashier’s counter – in a protracted argument
–
whereas he had just testified to that fact. He had, he explained,
gone into the store to buy airtime and stood in front
of the counter
waiting in the queue.
[95]
Contrary to the dicta in the minority judgment of Malan JA in
R
v Mlambo
1957
(4) SA 727
(A). See instead the position in
S
v Mtsweni
1985 (1) SA 590
(A), in particular at 593ff,
S
v Campos
2002 (1) SACR 233
(SCA).
S
v Burger and Others
2010 (2) SACR 1
(SCA).
[96]
S v
Deppe
[2013] ZASCA 4
,
S
v Mtsweni
1985 (1) SA 590
(A),
S
v Campos
2002 (1) SACR 233
(SCA),
S
v Burger and Others
2010
(2) SACR 1
(SCA). See also Zeffertt & Paizes
The
South African Law of Evidence
3rd ed (2017) 127ff.
[97]
Under the heading ‘FAILURE TO PROVIDE A VERSION‘ on on
page 46ff.
[98]
Discussed above at paragraphs [218]-[220].
[99]
Discussed above at paragraphs [218]-[220].
[100]
See
S v
Mthetwa
1972
(3) SA 766
(A) at 768; and
S
v Carolus
[2008] ZASCA 14
;
2008
(2) SACR 207
(SCA) at para. 17.
[101]
Under the heading ‘OBLIGATION TO PRODUCE LICENCE’ on
page on page 45ff.
[102]
Paragraph 53;
Director
of Public Prosecutions, Gauteng v Pistorius
(96/2015)
[2015] ASCA 204 (3 December 2015); see also
S
v De Oliveira
1993
(2) SACR 59
(A).