Identification parades: an empirical survey of legal
recommendations and police practice in South Africa.
Colin Tredoux *
Department of Psychology
University of Cape Town
* To whom correspondence should be addressed
In court a witness will swear under oath that the accused is the person whom they saw
committing the crime. This can lead to the conviction of a guilty criminal, but can also
lead to the conviction of an innocent suspect. Identifications are often tested by means
of identification parades, which have become an accepted part of police practice in most
countries. A substantial international literature exists on the topic, but little South
African research has been conducted. In this study the practice of identification parades
in the greater Cape Town area is investigated. We report interviews with attorneys,
prosecutors, and police officers, as well as data collected during participant observation
of police training. We conclude that there is much room for improvement in police and
legal practice in South Africa, and show how international research on identification
parades could be usefully drawn on.
’n Getuie in die hof sal onder eed sweer dat die aangeklaagde die persoon is wat hulle
gesien het die betrokke misdaad pleeg. Dit kan lei tot die skuldigbevinding van die
skuldige, maar kan ook tot die skuldigbevinding van’n onskuldige lei. Uitkennings
word dikwels getoets deur middel van uitkenningsparades, wat in meeste lande deel
geword het van aanvaarde polisie praktyk. Uitvoerige internasionale literatuur bestaan
oor die onderwerp, maar in Suid Afrika is daar nog min navorsing oor hierdie
onderwerp gedoen. In hierdie studie is die gebruik van uitkenningsparades in die groter-
Kaapstad-gebied ondersoek. Onderhoude met prokureurs, aanklaers en die polisie word
rapporteer, sowel as data wat tydens observasie van polisie instruksie ingesamel is.
Daar is tot die gevolgtrekking gekom dat daar baie ruimte vir verbetering in polisie- en
regspraktyk in Suid Afrika is; ook wys die studie how internasionale navorsing wat
reeds gedoen is behulpsaam kan wees.
Witnesses to crimes often have crucial information about the physical identity or appearance of the
perpetrator, and are frequently asked to make a personal identification. Unfortunately, this process
is error-prone, and criminal history is replete with examples where this has led to tragic miscarriages
of justice. In the South African Law Journal, several articles have acknowledged the inherent
dangers of identification evidence. The ‘Justice of the peace’ writes, in 1926:1
... mistaken identity is the most likely and common cause of miscarriages of justice, and such
miscarriages not only shock the public conscience but give rise to doubt and uneasiness as to the
administration of justice. (At 287).
South African courts recognise the danger posed by a bald statement of identification,2 and
approach identification evidence with caution. One of the safeguards against the vagaries of
identification evidence is the police ‘identification parade’, or ‘lineup’. The identification parade is
widely regarded as an important safeguard to an accused who is implicated by identification
evidence, both in South Africa, and abroad.
Identification parades are, in our opinion, attributed too much protective power: less protection is
accorded to an accused by an identification parade than is commonly supposed. We will present
this argument in three stages. In the first stage, we will briefly review regulations which have come
to govern the conduct of identification parades. In the second stage, we will provide evidence from
interviews with prosecutors and attorneys which suggests that there is much disagreement about the
evidentiary value of identification parades within the legal profession. Thirdly, we will show that
there are significant problems with identification parades at the level of police practice. Finally, we
will review some international psycho-legal research on identification parades, and show that there
is much we can learn from this research.
The nature of the problem
From a legal point of view the question that an identification parade attempts to answer is whether
the suspect (the person that is thought to have committed the crime) is in fact the culprit (the person
who actually did commit the crime). Several methods have been used to answer this question. In a
dock identification, the witness is asked during the court case if the person in the dock is the culprit.
A confrontation occurs when the witness is shown a single suspect and asked whether this is the
culprit. These methods have been dismissed as largely unacceptable due to the leading nature of the
question involved, where the suspect is the only real answer available to the witness. This was
recognised in England as early as 1860,3 and the identification parade was invented at this time with
the explicit purpose of providing a better method of identification. In an identification parade, the
suspect is placed amongst known innocents and the witness is asked if he/she can recognize the
culprit. This is a good way to use the ability of the witness to recognize someone rather than recall
their features, but has also been found to be problematic. Internationally, there is a recognition that
identification parades need to be treated with caution and that they need to be conducted according
to specific principles for them to be fair. Even then, they are not infallible procedures4.
South African courts have also recognised that identification parades are not infallible. Thus a
cautionary rule exists for identification parades, as mistakes have been found to occur even when
the witness is being honest and the procedure of the parades was correct5. Further it is also
recommended that the evidence should not only be admitted but the circumstances surrounding and
the procedures of the actual parade should be examined by the court for factors that may influence
the outcome of the parades6.
In most countries, including South Africa, there is thus a recognition in the legal system that
identification parades do not yield perfect information, as they rely on human memories (which are
not perfect recorders)7, and meticulous police procedure. It is clear that human memories are more
difficult to control than police procedure, and much effort has been expended in developing
regulations and recommendations that control the way in which police conduct identification
parades. It is worth dwelling briefly on these.
Law and regulations governing identification parades in South Africa
South African criminal law is based in substantial part on English criminal law, and we thus look to
English history for the origin of identification parades. Identification parades can be traced as far
back as March 1860, when they were instituted by a Metropolitan Police order. The order stated that
the police should place a suspect amongst his/her peers and then ask the witness to select the person
seen performing the crime8.
The earliest reported South African cases involving identification parades that we have been able to
find were in 1932 and 19359. Since then, however, courts have addressed themselves on a frequent
and regular basis to the question of identification parades, and there is now a substantial body of
regulations and recommendations. We will point to a few of these only.
It is recommended that the investigating officer of the case should not be in charge of the parade as
this will heighten the suspicion of unfair conduct in the courts10. The accused is given the right to
have legal counsel present at the parade11. This however does not place the onus of ensuring a fair
parade on the defence, but the onus for ensuring the parade is fair rests with the prosecution12, who
is here represented by the police. The witness should not be allowed to see the suspect before the
parade and the foils on the parade should be strangers to the witness13. The witnesses should also be
shown the parade separately and should not discuss the parade amongst themselves14.
The number of foils on the parade should be 8 (or 10 in the case of 2 suspects)15. All people on the
parade should be of similar build, height, age and appearance, and they should all be of about the
same occupations and be similarly dressed and of the same sex and race16. Furthermore, the witness
should be told that the culprit may or may not be on the parade and that they should indicate if they
can make an identification or not17.
Previously witnesses were required to touch the shoulders of people they pointed out on a parade,
but it has been recommended that this should not be a requirement, and that especially in sensitive
cases, the witness be allowed to only point at the person or view the parade from behind a one way
mirror18. The identification parades should also be recorded in an acceptable way - which is usually
by completing Form SAP 329 - and it is suggested further that a photograph be taken of the
composition of the parade as well of each identification that is made19.
The only statutory law that exists in relation to identification parades is the Criminal Procedure Act
Any police official may . . . of any person arrested upon any charge . . . of any such person released on
bail or on warning under section 72 . . . make a person . . . available or cause such person to be made
available for identification in such a condition, position or apparel as the police official may
This law places the authority and responsibility on the police to conduct identification parades, and
it is thus important to examine police practices in respect of identification parades, which we do
later in the article.
However, the question of practice is also relevant in the case of other legal professionals in the
criminal justice system, namely prosecutors, attorneys-general, and attorneys. Even though there
are many recommendations to guide parade construction and procedure, these will be interpreted
and applied to particular cases, and there is room for considerable variation in this respect. There is
in addition the question of whether the extant recommendations are sufficient, and whether there is
consensus among criminal justice professionals in this regard.
Interviews with prosecutors and attorneys
We interviewed several legal professionals, namely three who use identifications at parades in the
business of criminal prosecution, and two who regularly defend accused people who have been
identified at identification parades. Specifically, we interviewed the Attorney General of the
Western Cape, Mr. Frank Kahn, two senior state prosecutors, and two practising attorneys.
The prosecutors we interviewed should not be considered representative of all prosecutors,
especially as they both work only at high court level, and because it is clearly not possible to base
generalisations on two cases. One of the prosecutors was approached directly, and the Attorney
General referred us to the other. The attorneys approached for comment should also not be
considered representative, as we approached them on the basis of their reputed experience and
knowledge of identification parades. While these professionals are not representative, their
opinions do give us some idea of how legal professionals view identification parades.
The interviewees were all posed questions concerning their understanding of i) identification
parades and their use in the criminal justice system, and ii) their familiarity with international
research in this regard. The interviewees were also asked about their experiences in specific cases,
and whether they believed that alternative methods of conducting identification parades might
improve some of the identified problems.
The Attorney General decides what cases are to be prosecuted and thus has much influence on the
types of evidence that are seen as acceptable, or problematic. In the interview, he claimed that
identification parades were rarely problematic, and that it was not worth investigating them in any
detail. He did not seem to be familiar with any of the literature or research conducted about the
problems associated with identification parades. He further pointed out the good use that
identification parades can be put to, for example in the case of Avzal Simons, better known as the
“Station Strangler”, where the identification parade played a large part in securing his conviction.
State prosecutors deal with the cases that the Attorney General has decided to prosecute and
identification parade evidence may be an important part of these cases. They are thus the ones who
get the identification parade evidence from the police and need to decide whether it helps or hinders
their case, since they will have to present it in court. Thus the opinions that they have of
identification parades is integral in determining which parades will be presented in court and those
that will not.
The two senior state prosecutors who were consulted for their opinion did not in principle see a
serious problem with identification parades, as presently practiced in South Africa. One saw them as
completely unproblematic, and suggested that they are to be preferred over and above the use of
photospreads or other alternative forms of parade, as they contain more information and clues about
the suspect’s behaviour. This prosecutor went on to defend the parade in the “Station Strangler”
case as completely unproblematic and the way it was treated in the court as fair (see later
discussion). She did not seem to be familiar with any of the research or literature regarding
The other prosecutor felt that there were some problems associated with the conduct of parades.
Inexperienced police officers often conduct parades and do not check all the requirements, and this
leads to mistakes. Often identification parade evidence is not led in the court as the prosecutor feels
the case is strong enough without it, or can see that it will be dismissed by the courts due to
problems. He felt that one-way mirrors should be standard use for identification parades and was
interested in the option of alternative parades, although he acknowledged that the courts seemed to
prefer corporeal parades21. He was not familiar with any of the research about identification
parades, but was interested that there were many (possibly better) alternatives to the parades used in
Defence attorneys often have to argue against identification parade evidence. The attorneys can be
present at the identification parades22 and have access to the Form SAP 329 after the parade23, but
while at the parade, are only allowed to observe. This does not place any obligation on the accused
or the defence attorney to ensure that the parade is conducted properly. They can, however, use
their observations in court to attempt to show that the parade was not conducted regularly.
The attorneys we interviewed felt that police make serious mistakes in the conduct of most
identification parades, and it is accordingly not difficult to get an identification parade dismissed as
evidence. Only reasonable doubt needs to be created for the court to dismiss the evidence. The
attorneys suggested that many problems are associated with identification parades, and they were
fairly familiar with the international literature in this regard.
In general, legal professionals involved in the use of identification parades at the level of the courts
seem to have quite different views on identification parades, ranging from complete acceptance, to
skeptical use, and expedient rejection.
While the law has certain requirements and certain expectations of what identification parades
should look like, there is still an enormous amount of flexibility and space that is left for the
interpretation of certain recommendations. Thus, in the practical application of the law and the
precedents that have been set, the police play a very important role in translating the law into
practice. The police are in fact the only people explicitly granted the power to conduct identification
parades, and thus decisions about the finer working of the parades are left to them. This means that
when looking at the state of identification parades in South Africa, the South African Police
Services are an integral part of those who need to be consulted to get a comprehensive picture of
what is actually happening.
It has been noted on more than one occasion by the courts that the conduct of identification parades
is a matter largely in the hands of the police, and that this therefore gives them a wide discretion.24
For this reason, it is important to examine the in-house procedures the South African Police Service
follows in respect of identification parades.
Identification parades in the South African Police Service (SAPS)
We approached the SAPS in the Western Cape for their co-operation in respect of research into
police practices with respect to identification parades. This co-operation was given from the highest
down to the lowest level, which can probably be attributed to the new policies of open and
Various people in the police who are involved in identification parades were interviewed. First of
all, two senior police officers involved in the training of detectives were interviewed, and this was
complemented by observer participation in training in identification parades, given at a detectives’
training course. Finally, videotaped recordings of four identification parades were obtained and
The South African Police Service (SAPS) is a large organization, and one should naturally expect
police practices to vary considerably from region to region, and we do not claim that our review and
analysis can be generalised to all centres in which the SAPS conducts identification parades.
The training given for identification parades consisted of a 2 hour lecture and a 2 hour practical
session. This training is part of a 4 week course which is offered to police officers training to
become detectives. The training programme is arranged by the police officers stationed at the police
training college. There are several training centres around the country, and they have a common
syllabus, but the way this is presented varies from place to place.
The lecture details the procedure of the identification parade according to the Form SAP 329. The
emphasis is largely on the right of the police officer to be in charge and to take control of the
situation and the importance of him knowing what the rules and procedures are. The instruction
focuses on what the police officer in charge of the parade needs to organise (the venue, time, place,
foils, escorts and guards for the witnesses, interpreter, video unit and lawyer), who needs to be at the
parade, and who may and may not be at the parade. Much time is spent in outlining the practical
details of how a parade is organized, and the procedures that need to be followed. Some advice is
given about how to handle the selection of the foils, but this is largely anecdotal, and little formal
advice is given as to how to select foils. The police booklet is referred to as the source for detailed
information and the Form SAP 329 is vaunted as a good guide on how to conduct and record
parades. No detailed explanations are given as to the reasons for the procedures or the way things
are done, but emphasis is placed on the fact that the parade is part of the evidence that should be
supplied if the identity of the culprit is uncertain.
In the practical training component, trainees constructed parades, and conducted them with the
assistance of their fellow trainees. The trainers watched this and made comments on the procedure
and criticisms as to things that could be improved. They also added possible scenarios that could
happen and how the “officer in charge” would react. This was a fairly artificial situation as there
were only about 20 people present to choose foils from, and the only people who practised to “be in
charge” were people who had previously conducted parades. The SAPS video unit was present at
the training session and they videotaped “parades” for demonstration purposes. The video unit
record all identification parades that are held behind one way mirrors, and also record other parades
when specifically requested to do so. They have therefore observed many identification parades and
were able to give advice to the trainees. This advice consisted mainly of advising the officer in
charge to be familiar with the procedures and the Form SAP 329 and to be confident about what
they are doing.
Conduct of parades
The training that police officers get with regard to identification parades is not standard and is not
necessary for them to hold parades, thus is was felt that the best way to get an idea of what really
happens at the parades would be to view several parades and then see how these compare to the
principles set out by the law and by the police recommendations. These can then also be analysed to
get an idea of what useful recommendations would be as to ways of improving identification
parades. Access to identification parades is limited to certain people and thus watching the live
parades was not an option, but once the videos of identification parades are admitted as evidence
they are public property and the police video unit also made videos of identification parades
available for use in this case. Thus one parade was obtained from the Supreme Court archives and is
the identification parade from the Avzal Simons case, better known as the “Station Strangler”. The
other three identification parades were from the archives of the SAPS video unit and thus no further
information about the cases or whether they were eventually used in the cases is available.
Review of the 4 parades
As it is not feasible to distribute video tapes of the parades with this article, each parade will be
described here, so as to outline the events on the video tapes. We will also provide a critical
discussion of each of the parades.
The “Station Strangler” parade
The parade took place on 14/4/1994 at the Belville-South police station. The accused in this case,
Avzal Simons, better known as the “Station Strangler”, appears on parade behind a one way mirror,
with 8 other men. Two witnesses view the parade, one points out the accused while the other points
out no one.
The videotape begins with the officer in charge of the parade, the video crew and assistant being
introduced by the commentator. The officer in charge is in the room with the accused and the foils
and the video crew is behind the one way mirror. This set up explains the poor footage on the video,
due to difficult lighting conditions. The officer in charge then tells the accused about the parade
procedure; shows him behind the one-way mirror and confirms that he declined to have legal
representation present. The officer in charge further asks the accused to hand out the numbers to all
on the parade and asks whether he is satisfied with the parade. The accused initially says that he is,
but later does raise an objection - he points out the he is the only one with a facial scar on the parade
and the culprit is alleged to have a scar too. The officer in charge responds by saying that he is
independent and does not know what the case is about.
This part of the parade occurs mostly according to police protocol. The suspect is informed of the
procedure and his rights. What is of concern from a psychological point of view is that the foils
know who the suspect is and might through their behaviour indicate this to the witness.25 It is also
mentioned that the accused picked his own foils. This might give the accused the feeling of being in
control, but the responsibility for ensuring a fair parade still rests with the police and not the
accused. The accused can not be expected to know how to construct a fair parade and thus allowing
this and asking the accused if he/she is satisfied with the parade is without purpose. The accused
can contest the fairness of a parade in court, but at the actual parade the responsibility is solely with
the police to ensure a proper parade. In this case the accused did raise the objection of a facial scar
which was a very important one and the officer in charge simply brushed it aside26.
The officer in charge of the parade then takes down the names and addresses of the people on the
parade, this has the purpose of making these people traceable in case this is needed when the case
goes to court. The officer in charge explains to the accused and foils that he will be on the other
side of the mirror and will call the witnesses to view the parade, but between each witness will let
them know what has occurred.
Looking at the construction of the parade, we see 9 men of varying height standing against the back
wall of the room facing the one way mirror. The quality of the video footage does not allow us to
see the men’s facial features clearly and thus we cannot comment on the similarity of these. All
those on the parade are of the same race and sex though. Looking at their clothes, it is apparent that
seven of the men are dressed in fairly plain coloured clothes, mainly in shades of white, light blue
and beige. The accused, in contrast, is wearing a multi-coloured T-shirt and bright pants. One other
person is wearing a red T-shirt and blue jeans. This makes the suspect and the other man in the
bright clothing stand out amongst the other men on parade and thus the viewer’s attention is drawn
to them. This does not concur with the recommendation that the clothing of parade members should
be similar. This problem is compounded by the fact that the accused is also the one with the
heaviest build on parade and this adds to his noticeability amongst the foils. Thus, on the level of
construction, the parade unfairly biases the suspect, as he is different from all the others in two very
obvious respects, and the witness’s eye may be be drawn to him for this reason only.
After constructing the parade, the first witness is called. He is a boy of approximately 8 years of age.
He is afraid to enter the darkened viewing room on his own and thus his mother is allowed to
accompany him. The boy is made to stand in the corner of the viewing room so that he cannot see
the parade while the officer in charge explains the procedure to him. The instructions are very stiff
and formal and not aimed at someone his age. The concept of the one-way mirror is not clearly
explained to him. He is asked to say the number of the person he picks out, but no check is made
that he is able to read. The whole procedure seems very daunting to him and he is obviously
uncomfortable as he then faces the parade. He stand and looks at the parade for about five minutes
without saying a word and then simply says that he does not remember. He is then dismissed from
the room and the officer in charge informs the accused and the foils, via the intercom system, that
- 10 -
the witness did not point anyone out. The accused is then given the chance to change number,
position and make further requests; he swaps number and position.
The next witness, a woman, is then called and escorted to the parade by a police officer. The officer
in charge instructs the woman to stand in the same corner as the boy did and he then informs her
that she is here to point out the person she saw on the particular day. She is asked to say the number
of the person she points out, but is told not to worry as she is behind a one-way mirror and the
people on the parade cannot see her. This middle aged woman looks at the parade and tentatively
indicates that she thinks it is number 27. This is in fact the accused, but she then asks that he come
forward and walk back to his position so that she can see him walking from behind. She then also
requests to see him while facing towards the left. This she also asks one of the foils to do (the one
wearing the red T-shirt). Finally she requests that the accused say the words “loop vinnig” so that
she can see his lips. She then ends by saying that the accused does look like that person, but that his
hair style was different then. She is then thanked and dismissed from the parade.
This part of the parade occurs according to the principles set out by the police and the law, but what
is of note is that there are also many other cues in the process that bias the accused. The clothing
makes him distinctive and this is substantiated by the fact that the witness here only wanted to take
a closer look at the two people who were wearing bright clothing. Then the officer in charge also
allows the accused to stand with his left side toward the witness for a long time while she looks at
him and he writes notes - when the other man is standing forward with his left side toward the
mirror he is allowed to step back as soon as the witness makes her first comment and he does not
write any notes about this request. In an unconscious and probably unintentional way this says to
the witness that the other man is not worth looking at but the accused is worth looking at. Also, the
body language of the foils on parade shows that they are quite relaxed, leaning against the walls and
not standing up straight, while the accused is standing quite straight and responds to the requests in
a very eager manner. The instructions that the witness is given also indicate to her that she should
pick someone out and there is no warning that the culprit might not be on the parade at all. (The
exact words used were: “Jy is vandag hierso om ‘n verdagte of verdagtes uit te wys op hierdie
parade. As jy nou na die parade kyk, indien jy iemand uitwys, moet jy net die nommer van die
persoon hard op vir ons noem”).
The officer in charge then thanks her and she leaves the parade viewing room. The officer in charge
then again informs the accused and the foils that the witness did point out the accused, but said that
his hairstyle was different. The accused then wants to know what the witness said the hairstyle was.
- 11 -
The officer in charge informs the accused that the witness will later say what the hair style was, and
he ends the parade.
This parade can be argued to contain a bias against the accused by the way in which he is dressed,
by his build in comparison with other parade members, and by other more subtle cues from the foils,
himself and the officer in charge.
The two hat parade
This parade opens with a view of the entire parade in front of the one-way mirror and we see that
there are 8 people on parade. This means that the requirement of 8 foils plus the suspect was not
adhered to. The 8 people are all of the same sex and race. On the parade there are two people who
are wearing woollen caps (the accused is one of these), the rest are without hats. The suspect on this
parade also has a moustache, which is a feature that is not shared by all the people on the parade.
The clothes worn by parade members are fairly similar. The height of all on parade is about equal,
except that one man is quite a lot taller than the rest. The video tape does not show the officer in
charge talking to the foils and the accused, but starts with the first witness entering the viewing
room. This witness is also asked to stand in the corner while the officer in charge instructs the
witness that the he should point someone out if they are on the parade, but also warns him that the
culprit might not be on the parade at all. The one-way mirror is explained to him and he is told that
he can make requests regarding the parade e.g. asking a parade member to stand closer to the glass.
He is then asked if he understands this and if he has any further questions.
These instructions are clear, to the point, clearly warn the witness that the culprit may not be there
and allow the witness to raise any problems or questions that he/she may have.
The first witness looks at the parade and very quickly points out no. 3, who is the accused. What
becomes very clear then is that the accused has his woollen hat pulled very low into his face and is
holding his number up to his chin, as if he were hiding, while all the others are holding their
numbers in front of their chests. This behaviour on the part of the accused makes him look very
suspicious and draws attention to himself.
When the witness has left the viewing room the accused is told that he was pointed out and is then
given the chance to change position, number, clothing or make any other requests. He does not want
to change anything and thus the next witness is called. The second witness is absent and thus the
third and fourth witnesses are called one after the other and the same procedure is followed, with
good clear instructions and each witness pointing out the accused.
- 12 -
This parade can be commended for the instructions that are given to the witness, but the
construction that allowed two people with hats is very problematic: if the officer in charge were
aware of more subtle cues that bias the accused - like holding his number to his chin - he could warn
the accused of this and make the parade even fairer.
The one scarf parade
This parade consisted of 9 people, and 2 witnesses viewed it. For this parade, a legal representative
was also present. The videotape starts with the officer in charge and the camera man being
introduced through verbal commentary. The officer in charge informs the accused and the foils of
the parade procedure. Regarding the construction of the parade, the height of those on the parade
varies by about a head and a half. The hair styles also vary from practically clean shaven, to a small
afro. The accused has a full beard, but some on the parade only have a moustache while others are
clean shaven. These features lead to considerable variation and cannot satisfy the requirement that
foils be of similar height, hair style and length. One of the men is also wearing a pair of shorts and
another is wearing a scarf around his neck, and holding a cap in his hand. Thus, the quality of
construction in this parade is also questionable, in terms of the similarity of the clothing of the foils
and the accused.
The officer in charge then proceeds to call the first witness. He gives the witness instructions that he
is here to point out a person who robbed him on a particular day. He is told that he is behind a one-
way mirror and that parade members cannot see him. He is not warned that the culprit might not be
on the parade.
The witness then faces the parade. He says quite quickly that the person who robbed him had a scarf
around his mouth like the one man is wearing around his neck, but further than that he can not say
who it was. Here it becomes clear that the presence of an item of clothing that was important in the
crime is not advisable on the parade and should be avoided as it can bias the accused if he happens
to be wearing it, or alternatively draw attention away from him if another is wearing it. The witness
is then dismissed. The legal representative, who till now had only been observing silently follows
the witness out of the room and looks where he is being led to.
The second witness is then called and given instructions. This time the officer in charge does warn
the witness that the culprit may not be present. This witness points out the accused when she views
the parade. She is then dismissed and the parade ended.
The prison parade
- 13 -
The fourth parade takes place at Pollsmoor Prison with awaiting trial prisoners. When identification
parades take place here, the police need to co-operate with the prison authorities in order to obtain
foils, and to make other arrangements for the parade. Parades are usually held in prison if the
suspects are prisoners or awaiting trial there. The foils that are used in these cases are usually other
prisoners and the prison authorities usually pick these out. The parade that we have on video tape
takes place in a large room, possibly a hall in the prison. There are 18 men on the parade and they
are standing against one wall. Amongst these there are 4 suspects. According to the principle that
there must be 8 foils plus an additional 2 for each additional suspect, the parade meets this criterion,
but on the grounds of similarity, there does not seem to be any real resemblance between the people
on parade. They look more like a gathering of young men, with not even their clothes, heights or
builds being similar.
This type of parade, i.e. where there are multiple accused, apparently occur fairly frequently. They
presumably pose less work for police, and it is therefore understandable that they use this method,
particularly since the law allows them to do so. The problem is that the accused are seldom very
similar to each other and are thus not effective foils to each other. This makes the selection of
further foils problematic, as it is not clear whom they should be similar to, and this results in the
situation where there appear to be no underlying shared features between those on the parade.
The first witness in the parade in question is then called. He stands just outside the doorway to the
room and can thus not see the parade. The officer in charge instructs him to touch the shoulders of
those people that he identifies, but the rest of the instructions are not very clear on the tape and we
cannot comment on them.
The witness walks into the room and touches the shoulders of two people on the parade in fast
succession. He is asked to repeat this and pose for a cameraman who is also present. At this the
witness looks clearly uncomfortable. He proceeds to point out 4 people on the parade. He is then
led away by a prison official through a different door. The next witness is called. The same
procedure is followed with him and he points out three people on the parade. He is then led away
through the same door.
This parade clearly shows the difficulties with parades that contain several suspects. The suspects
will usually not be similar to each other and thus it is difficult to find foils that are similar to the
suspects. In practice this means that it will be very difficult for these parades to meet the criteria of
similarity and thus will usually be unfair for the suspects. The witness in this parade was also
brought face to face with the suspects and had to touch them on the shoulder. This used to be
- 14 -
requirement, but is not longer neccessary. The witnesses behaviour here shows that coming face to
face with the suspects can be very anxiety provoking and is thus not in the interest of the witness.
This is also especially true for cases where the witness might be threatened by the suspcet or in
sensitive cases such as rape. Thus while this parade fulfills most of the legal requirements, there are
problems associated with these requirements as this parade shows.
Conclusion and recommendations
The review of these 4 parades suggests that there are problems associated with the implementation
of identification parades in the South African criminal justice system. In all the parades that were
reviewed, the accused was biased by the procedure of the parade and thus not afforded the
protection that an identification parade should afford a suspect. While these identification parades
are not representative of all identification parades that are conducted in South Africa, they do
indicate that serious problems are occurring. The parades show that the recommendations set out by
the law and by the police to guide the conduct of parades are not sufficient to ensure fair parades.
This problem is further exacerbated by the fact that the people involved in conducting and
evaluating identification parades do not recognise that parades are frequently problematic.
In this respect, psychology has an important role to play. Much research has been done around the
construction of identification parades and how different practices affect the correct and incorrect
identification rates of witnesses. Several things are clear from this research. Firstly, the way
parades are conducted in South Africa is inoptimal. Secondly, there are methods of conducting
identification parades which show better identification results - that is, they afford more protection
to suspects, while not affecting the number of identifications of guilty suspects. Thirdly, these
alternative methods make it easier for the police to construct identification parades. Finally, these
alternate parade procedures are less stressful for witnesses.
One of these alternatives is the use of photographs or videotapes in place of corporeal parades.
Photographs are placed on a board or a table and arranged with the same precautions used for a live
parade. A video parade is set up combining snippets of filmed foils into a parade format. Although
it is a popular legal belief that videos and photographs contain less information than the live
person27, research has shown that there no real differences exist in the rate of identifications across
live parades, video parades, black and white or colour photograph parades28.
Practically, photo or video arrays are also much more convenient, and effective. They can be set up
faster, as people do not need to be found to serve as live foils - there are methods of arranging a
- 15 -
large number of photographs so that they can be accessed swiftly. Indeed, a very large data base can
be set up, making more foils available, and therefore making suspect-foil similarity easier to
establish. The behaviour of people on the parade is by definition not a problem, eliminating possible
bias against the suspect. The anxiety of witnesses is greatly reduced in that they do not need to see
the suspect face to face29
In addition, research has examined different procedures and structures for identification parades.
One of these options is known as the ‘sequential’ parade. In the case of ‘sequential’ parades, parade
members are presented to the witness individually, in a randomised sequence. The witness needs to
decide for each person shown, if he/she is the culprit or not. If the witness decides that the person
shown them is not the culprit, then the next parade member is shown. The witness in this way is
forced to make an absolute choice about each person seen, rather than comparing the people on
parade to each other30.
This method has been fairly thoroughly tested, and shown to be highly robust and substantially
superior to the tradition simultaneous parade, as used in South Africa. Studies have found that the
correct identification rate is about the same for sequential and simultaneous parades, but that the
false identification rate is very significantly lower in the case of sequential parades. This means that
witnesses are still able to point out the culprit, but that they are much less likely to point out an
innocent person if the culprit is not on the parade31. This method works best if the witness does not
know how many people there will be on the parade32. Further research has found that sequential
parades are also robust against other biases. Suggestive instructions and lower suspect-foil similarity
do not affect sequential parades as much as simultaneous parades33.
There is a great deal more psychological research on identification parades, but to review it here is
beyond the scope of this paper. We have simply chosen two well researched alternatives to
traditional parade methods, and shown how they may improve police practice. Readers interested
more generally in psychological research on identification parades are referred to the overviews
provided by Gary Wells34, Elizabeth Loftus35, and Ross and colleagues36.
The identification parade was originally devised in order to improve police practice in securing
evidence of identification. There is little doubt that the identification parade improved police
practice in the mid 19th century, but there are also clear indications that the identification parade
- 16 -
does not offer satisfactory protection against mistaken identifications, especially in its practical
execution in South Africa. There are many ways in which current legal and police practice in South
Africa can be improved, particularly if cognisance is taken of modern psycho-legal research.
- 17 -
‘Identification’. South African Law Journal (1926), 287. (No author is given).
R. v. Shekelele and another 1953 (1) (SA) 636 (T), S. v. Mehlape 1963 (2) (SA) 29 (A), and many
other cases express the need for caution. Indeed, Didcott, J. in S. v. Ngcobo 1986 (1) (SA) 905 (N),
at 906 says:
The danger of mistaken identification is one to which all judicial officers are or should certainly be alive. Enough has been said
about it over the years, and in various parts of the world to see to that.
Lord P Devlin Report to the Secretary of State for the Home Department of the Departmental
Committee on Evidence of Identification in Criminal Cases. (1976).
GL Wells ‘What do we know about eyewitness identification?’ (1993) 48(5) American
S v Mthetwa 1972 (3) SA 766 (A) and S vs Mehlape 1963 (2) SA 29 (A) and S v Khumalo 1991
(4) 310 (AD)
S v Mlati 1984 (4) SA 629 (A)
M Lezak Neuropsychological Assessment 3. Ed (1995)
JW Shepherd, HD Ellis & GM Davies Identification Evidence: A Psychological Evaluation (1982)
Mkize v. R. 1932 (1) (PH) H17 (N), R. v. Olia 1935 (SA) 213 (T)
R v Dickman 1910 5 C A R 270
S v Jija and others 1991 (2) SA 52 (E)
S v Mlati 1984 (4) SA 629 (A)
Kola v Rex 1949 (1) P H, H 100 (A)
R v Nara Sammy 1956 (4) SA 629 (T)
R v Oglia 1935 T P D 213
This is a combined rule becomes evident from several cases where individual recommendations
were made in this regard. For Example: R v Masemang 1950 (2) SA 488 (A) , S v Mlati 1984 (4)
SA 629 (A) and S v Pelwan 1963 (2) PH H 237 (T)
- 18 -
R v Nara Sammy 1956 (4) SA 629 (T)
S v Leepile 1986 (4) 3 661 (W)
S v Mlati 1984 (4) SA 629 (A)
The Criminal Procedure Act No 51 of 1977, Section 37(1)
Live corporeal parades are identification parades where the suspect and foils are present in person
and the witness views the people in person. This is as opposed to photospreads where the witness
will view photographs of the suspect and foils. Other alternatives are video parades and voice
Par. 11 of SAP Form 329
Erasmus J in S v Jija & others 1991 (2) SA 52 (E)
S. v. Sibanda and others 1969 (2 ) (SA) 345 (T).
RS Malpass and PG Devine, (1984). Research on suggestion in lineups and photospreads. In GL
Wells & EF Loftus (Eds). Eyewitness testimony: Psychological perspectives. Pp 12-37. (1984).
The issue of the facial scar was raised in the court, but the judge did rule that this could not have
been what the witness used to identify the accused as she could not have seen the scar at the
distance she was standing from him at the time.
S. v. Shandu 1990 (1) (SACR) 80 (N).
BL Cutler, GL Berman, S Penrod & RP Fisher ‘Conceptual, Practical and Empirical Issues
Associated with Eyewitness Identification Test Media’. In D Ross, JD Read & MP Toglia (eds.)
Adult Eyewitness Testimony: Current Trends and Developments (1994) 163
Cutler, Berman, Penrod & Fisher op cit. (n 26)
EF Loftus & JM Doyle Eyewitness Testimony: Civil and Criminal (1987)
G Koehnken, RS Malpass & MS Wogalter ‘Forensic Application of the Line-up Research’ In S
Sporer, RS Malpass & G Koehnken (eds.) Psychological Issues in Eyewitness Identification (1996)
GL Wells op cit (n 4)
- 19 -
RCL Lindsay, JA Lea, GJ Nosworthy, JA Fulford, J Hector, V LeVan & C Seabrook (1991) in
Wells op cit (n 4)
GL Wells Eyewitness identification: A system handbook. (1988), and GL Wells, 1993 op cit (n 4)
Loftus and Doyle, op cit
D Ross, JD Read and M Toglia. Adult Eyewitness testimony: Current trends and developments.