---

CG CASE NUMBER: 661/91

IN THE SUPREME COURT OF SOUTH AFRICA (APPELLATE DIVISION)

In the matter between:
MOSES ZITHA Appellant

and

THE STATE Respondent

CORAM: JOUBERT ACJ, EKSTEEN JA et HOWIE AJA HEARD ON: 19 MARCH 1992

DELIVERED ON: 31 MARCH 1993

JUDGMENT HOWIE AJA

2

Appellant was convicted by a regional magistrate of attempted armed robbery and sentenced to six years' imprisonment. He appealed unsuccessfully against conviction and sentence to the Witwatersrand Local Division. With the leave of that Court he has appealed against the order dismissing his first appeal. In argument the present appeal has been confined to the conviction.

The prosecution evidence comprised that of the complainant, Amelia Maria Jordaan, and her husband, Louis Hendrik Jordaan. The defence was one of alibi and was advanced by way of the evidence of appellant, his wife and his mother.

The commission of the offence by three men at the Jordaan home on the late afternoon of 23 December 1989 in clear daylight was not in dispute. The fundamental issue in the case was whether appellant was one of the offenders. Mrs Jordaan was unable to identify him as such but her husband claimed in evidence
3 to have done so.
Briefly, the relevant incident was this. Mrs Jordaan returned to her home in Senderwood at about 6.15 pm on the Saturday concerned and parked her car in front of the garage. Having alighted, she was confronted by three men in the drive. Two ran up and pointed hand-guns at her and the third went to the other side of the car. She started screaming. One of the men tried to grab her car keys. To attract her husband's attention she went on screaming. A short time later Jordaan opened the garage door. With that, one of the men with her turned and pointed his gun at Jordaan. She screamed to her husband to fetch a firearm whereupon he closed the door and ran into the house. Before he could return the three men made their getaway in a car waiting for them in the street. Prior to making good their escape one of the group tried unsuccessfully to snatch the complainant's choker from her neck.

Jordaan's evidence was that he was watching

4 television when he heard screaming. Because it persisted and was coming from his own premises he went to investigate. He opened the garage door to see his wife being held at gunpoint by two men. The moment the door was opened the nearest gunman turned towards Jordaan and pointed his weapon at him. The distance between them Jordaan estimated as five paces. He said it was so close he could clearly see that the barrel of the firearm was rusted. Jordaan hastily retreated into the garage and closed the door. He went to fetch his own firearm but was unable to find it. He soon returned to the scene but the intruders had gone. According to Jordaan appellant was the person who pointed the gun at him.

It was common cause that appellant had for about five months been one of sixty-eight drivers employed by the business concern of which Jordaan was the owner and that appellant and several fellow employees were retrenched about two months before the

5

offence. It was also common cause that on an occasion
at the commencement of appellant's period of employment
he unwittingly entered Jordaan's office without
permission, looking for the transport manager, and
Jordaan angrily told him to get out.

Asked in evidence how well he knew appellant,

Jordaan said:

"Kyk, ek het horn nie persoonlik so goed geken nie, want ek meng nou nie so met die drywers nie. Ek sien hulle maar so met die verbygaan, u weet met die in- en uitkommery en alles

When the magistrate pointedly asked Jordaan if he saw appellant regularly in that period, he answered affirmatively.

Reverting to the day of the attempted robbery, Jordaan fairly conceded that he took fright when he saw what was happening. However, he said he was "honderd persent" able to see that appellant was the person nearest to him. He also described the clothing which appellant was wearing. He added that the other man

6

confronting his wife was tall, thin and of dark
complexion and that he would be able to recognize him if
he saw him again. Referring to the length of time for
which he had the scene under observation, Jordaan said:

"Nou praat ons van sekondes van tien, miskien 15 of 20 sekondes of minder."

Jordaan testified that until he checked his

firm's records the following Wednesday and learnt that

appellant had been retrenched, he had been under the

impression that appellant was still on the staff. He

said that he could not search the records sooner because

the offices were locked over the Christmas holidays.

His evidence in this connection reads as follows:

"So ons kantoor het eers weer die Woensdag oopgemaak en toe kon ek eers dokumentasie gaan kry, want ek se vir hulle: luister hierdie persoon werk vir ons, ek sê dit is een van ons drywers. Toe ons deur die dokumentasie gaan saam met my sekuriteitsman ... en ek wys vir horn die foto, toe se hy vir my: nee hierdie meneer het al twee maande terug ons diens verlaat."

Subsequently Jordaan gave appellant's details
7

to the police and later still he attended an

identification parade. It was common cause that at the
parade he immediately pointed out appellant. Although
his evidence as to this pointing out was not clarified I
presume that he identified appellant not merely as his

former employee but as one of the culprits.
The evidence in support of the defence case was that on the Saturday in question appellant was at home in Orlando East with various members of his family baking cakes in preparation for Christmas. The only time he was away from the house was when he and his wife went shopping from about noon until 5 pm.

The magistrate found Jordaan to be a reliable witness and he rejected the defence evidence as not reasonably possibly true. He therefore concluded that appellant had been correctly identified as one of the participants in the offence.

The rejection of the defence evidence appears to me to have been based on inadequate grounds but in
8 the view I take of the matter, it is unnecessary to make a finding to that effect. I shall assume that such rejection was justified.

The crucial question is whether Jordaan's

evidence identifying appellant was reliable to the

requisite degree. In this type of case the self-evident risk which the State must eliminate beyond reasonable doubt is that of mistaken identity. In the present context the risk was that Jordaan saw the offender in question as being appellant when there was a reasonable possibility that it was in fact someone else who looked similar.
I accept that when Jordaan testified he was sincere and sure in alleging that the person concerned was indeed appellant. It is also clear that the state of the light, his having known appellant before, the fact that he saw the intruder's face and the short distance between them were all factors conducive to a favourable opportunity for reliable identification.
9 However, it is axiomatic that the trier of fact must have regard to all the evidence and to all such considerations as reasonably invite clarification or investigation. He errs if he attaches undue weight to the conviction and assurance with which the identifying witness claims to be right. Obviously he also errs if he ignores or undervalues factors that militate against a reliable identification.
Correctly, the magistrate set no store by the pointing out at the identification parade. Jordaan was by that time clearly convinced that appellant was the person concerned and it was inevitable that if the latter was on the parade Jordaan would point him out forthwith. The enquiry, then, is whether the magistrate's treatment of the case constituted compliance with what the leading reported decisions enjoin not only in regard to single witness cases but more particularly as to identification matters.

In S V MTHETWA 1972 (3) SA 766 (A) at 768 A-C

10 it was said that because of human fallibility the reliability of the observation made by the identifying witness must be tested. The trial court must therefore consider i.a. the opportunity for observation, both as to time and situation; the extent of the witness's prior knowledge of the accused; and the accused's appearance. It goes without saying that where, as here, the accused has no legal representation, the judicial officer has no alternative but to seek, within the constraints of his function as impartial arbiter, to conduct the necessary testing himself. In cases where the identifying witness has known the accused previously, identification marks and facial characteristics are of mush less importance than where there has been no previous acquaintance but it is then necessary to focus upon the degree of prior knowledge and the opportunity for correct identification having regard to the circumstances in which it was made: R v DLADLA AND OTHERS 1962 (1) SA 307 (A) at 310 D-E.

11
It seems to me to follow from these cases that

the importance of physical appearance remains but that

it is in inverse proportion to the degree of prior
knowledge. On the question of prior knowledge the
magistrate said the following as to Jordaan's evidence:

"hy (het) die beskuldigde betreklik goed geken, want hy het 'n hele ruk by sy firma gewerk waar hy horn gereeld gesien het. Hy was derhalwe in staat om 'n paar dae later na sy firma terug te gaan en 'n foto van die beskuldigde te gaan uithaal."

In my view this assessment involves an

overstatement and also lacks critical analysis. Jordaan

claimed to have seen appellant at close quarters as an

employee on only one occasion and that was about seven

months before the offence. As regards the rest of his

time with the firm, Jordaan apparently saw him merely in

passing, as one of sixty-eight drivers. In leading

fashion the magistrate asked whether he saw appellant

"gereeld", with which Jordaan simply agreed. The word

"gereeld" is relative. The court's investigation fell

12

short of revealing how regularly, at what distance and
in what situations. Jordaan himself said he really did
not have contact with the drivers. It is true that in
his account of the encounter in Jordaan's office
appellant said that Jordaan "net (daarna) sleg vir my
gekyk" but that vague comment was left unexplained. In

the result, to say that Jordaan's evidence was that he
knew appellant "betreklik goed" put the position too
high. That conclusion could only properly have been

reached upon further enquiry. An illustration of the

nature and depth of the required investigation was given

by Williamson JA in S v MEHLAPE 1963 (2) SA 29 (A) at 33

B-D:

"(H)e said he had often seen the appellant before. The value of this alleged prior knowledge ... remained entirely uninvestigated. The court did not know how often he had seen this man, or when he had last seen him, or whether he had even seen him close by or had ever spoken to him or anything at all about the opportunities of accurate observation of the appellant's face afforded on the prior occasions; he said that he recognized him by his face. The magistrate may of course have seen that the appellant's

13

face was of a type which was easy to remember and later to recognize; but he made no findings in that regard."

In the present case some of these factors were touched upon but insufficiently to satisfy the court as to "the opportunities of accurate observation afforded on the prior occasions".
In my view Jordaan's evidence as to the extent of his prior knowledge of appellant was such that it remained of importance either to embark upon the further enquiry I have referred to or for the magistrate to canvass, among other things, the matter of appellant's physical appearance. That he failed to do. The reader of the record has no idea of appellant's recognisability.

Then, as regards the circumstances in which Jordaan's purported identification was made, there is nothing in the magistrate's judgment expressly or impliedly to demonstrate that he took into account the brevity of Jordaan's observation and the fact that it
14 was made in a situation which engendered tension and fear in the witness's mind. By reason of that factor alone Jordaan's judgment was apt to be hasty (cf MEHLAPE'S case, supra, at 33 E) quite apart from the matter of the time constraint which applied.

As to the matter of time, it is implicit in Jordaan's very description that his chance for observation was exceedingly short-lived. Moreover, to judge by his evidence that the screaming he heard went on for possibly as long as five minutes - which would clearly have been an extraordinarily long time in the circumstances - it appears that his ability to estimate time was questionable. Even his estimate of ten to twenty seconds seems long in the prevailing circumstances. He was not asked what degree of attention he gave to the appearance of the man nearest to him as opposed to the other matters he observed in the brief opportunity available to him. It was naturally inherent in the situation that concern for his
15 wife would probably have been uppermost in his mind rather than the identity of the intruders. And when the gun was turned on him there was every reason to think of his own safety as well. There were, consequently, factors present in this case which militated substantially against a reliable identification.
The manner of removing any reasonable possibility of error in any given case is a matter entirely to be governed by the circumstances of the case (MEHLAPE'S case, supra, at 32 H) but other features which in my opinion called for investigation here were Jordaan's reason for consulting his employees' records, the tenor of his first report to the police and the stage when he first formed the view that the offender in question was one of his drivers. On the evidence it is not apparent whether that view was formed at the moment he saw the man concerned or later and, if later, how long afterwards. His mere assertion that it was appellant that he saw cannot by itself provide the
16 answer. The self-evident possibility in the prevailing circumstances was that although he could immediately have acquired a mental picture of a familiar-looking face, he needed time and reflection to conclude where he had seen it before and who the person was. There would have been nothing unnatural or improbable in that thought process. Therefore it was an appropriate enquiry whether he consulted the photographs in his records in order to try to transform a provisional impression into a positive identification or whether he merely sought the name and address of the man he had already, in his mind, definitely identified. In the same context the content of Jordaan's first report to the police would have been relevant to establish the consistency and confidence of his identification, more especially as to whether he positively knew the man as one of his employees or whether he merely thought he was. The content of such a report was considered in R vs T 1958 (2) SA 676 (A) at 681 E-F.
17 In the absence of any exploration of these considerations I do not consider that the magistrate was justified in concluding that it enhanced the reliability of Jordaan's evidence that he went to consult his records some days later. Once he as much as suspected that the person in question might be appellant he would naturally have gone to the records to look for confirmation.
In the result I think that the onus on the prosecution was not discharged. And it is appropriate to. emphasise that this was so not because of inherent inadequacies in the vital witness's powers of observation or articulation but because the treatment of his evidence by the State and the trial Court failed to demonstrate that the reasonable possibility of error had been excluded.

The appeal succeeds and the conviction and sentence are set aside.

18

C T HOWIE AJA ACTING JUDGE OF APPEAL

JOUBERT ACJ)
) CONCUR EKSTEEN JA)

---