Abudalla Nabulere & 2 Ors v Uganda [1978] UGSC 5 (5 October 1978)

Evidence Law|Evaluation of Evidence
Case summary
The court considered whether the trial court erroneously relied on identification of a single witness to convict the appellants. The court held that, where the quality of identification was good, a court could safely convict on uncorroborated evidence of a single identifying witness provided that it adequately warned itself of the special need for caution.



(Coram: Saied C.J., Nyamuchoncho, J.A. Ssekandi, J.A)



2. KULUSENI MUBALA………………………………………………………. APPELLANTS

(Appeal from a judgment of the High
court at Kampala (Allen, J. dated
23rd February, 1978.)

Criminal Session No. 8 of 1978



This appeal arises from conviction of the three appellants by the High Court sitting at Tororo for the murder of Maimuna Kiiza on 17th September, 1975. The deceased lived in a one—roomed hut with a friend Mary and a young boy Magidu (A2’s son). She was at one time married to A3 and A2 is their son. Al is also the son of A3 but by another marriage.

The facts as accepted by the learned trial the three appellants entered the deceased’s hut at night and on the orders of A3, A2 and A1 cut her on the head and shoulder wits pangas. She died instantly. Mary got off the bed during the attack and ran out to the verandah while raising an alarm. The three appellants followed her and A1 cut her on the left upper arm and A2 on the left side of the scalp her left arm was so badly cut that it was later amputated at the hospital. The

three appellants then ran way. Mary’s alarm was answered first by Kazimbye (PW3) and Kamuma (PW4), neighbours of the deceased. When they arrived, Mary recounted to them what had taken place. At the trial, Kazimbye and Kamuma gave different versions of what had taken them that night.

While Kazimbye testified that Mary named all three appellants as the assailants, Kamuma said that he heard her mention Al only. The next person to arrive at the scene was a chief called Peta. At first Peta said that he did not bother to ask Mary who the assailants were but in cross examination he changed his story and said that when he asked her, she said she did not know them. The learned trial judge rejected the evidence of Peta and Kamuma on this issue and believed that of Kazimbye that Mary named all three appellants.

The learned trial judge also accepted the evidence that prior to the day of the murder, the second appellant had uttered threats against the life of the deceased. The second appellant told his sister Nasimu (pw6) that the deceased had caused the appellant to suffer from a painful disease and if he had the opportunity he would cut her to pieces. Nasimu, in turn, informed the deceased who reported the matter to the local chief, Peta. (PW5). The chief did not do anything about the report; instead, he advised the deceased to return the following morning as it was already dark. As it turned out, she did not live to see the chief again.

Mr. Mpungu, for the three appellants, submitted before us that the learned trial judge erred in believing the testimony of Mary in view of the evidence of at least one witness, who answered the alarm, to the effect that she recognised only the first appellant among the assailants. He also submitted that Mary’s account of what happened that night could not he believed because her behaviour soon after the deceased was attacked was credible. He submitted that no one could believe her testimony that when Al and A2 were cutting the deceased he ran out to the verandah and stayed there until they came out and cut her. Ha also draw our attention to the fact that Mary testified that the hut in which they lived was dark. In his submission, the witness could not have been able to see what each of the assailants did in the house as she claimed.

Mr. Mpungu also submitted that the learned trial judge misdirected himself on the burden of proof of alibi and on identification. On identification, he submitted that the appellants were wrongly convicted on the uncorroborated testimony of a single witness.

We are greatly indebted to the forceful submissions made by Mr. Mpungu on behalf of the appellants. This is a difficult case and has caused us some anxiety. The appellants were convicted essentially on the identification evidence of Mary who is the only person that survived the attack, apart from Magidu who did not testify due to tender age. However, in a careful and well reasoned judgment, the learned trial judge evaluated all the evidence in this case. He was satisfied that Mary was a witness of truth. This was also the opinion of both assessors. The judge considered the effect of the evidence of Kamuma and Peta on her testimony. In him judgment, Kazimbye’s evidence that Mary named all the three appellants that night was preferable to that of the other two witnesses which he rejected for reasons given in the judgment. With regard to Kamuma he held that it may very well be that he had forgotten what Mary told her since the events took place two and a half years ago. Kamuma had told the police that Mary and the little boy Magidu named the three appellants as tile assailants but in court he said that Mary mentioned the name of the first appellant only. The trial judge also rejected the testimony of Peta. He found that Peta was a shifty and obviously an unreliable witness. Consequently, the judge found as a fact that Mary had been consistent in her story and because she was an honest, straight forward sod truthful witness he believed that her identification of the three appellants was correct.

While we appreciate yr. Mpungu’s efforts to persuade us to differ from the findings of fact mode by the trial court we think that we cannot do so in this case. The judge and the assessors had the advantage of seeing and hearing the witnesses. We have not. An appellate court has indeed jurisdiction to review the evidence in order to determine whether the conclusion originally reached upon that evidence should stand. But this is a jurisdiction which is always exercised
with caution. It is not enough that the appellate court might itself have come to a different conclusion. The appellate court will only interfere with the findings of fact of a trial court if there is no evidence to support a particular conclusion. But, if the evidence as a whole can reason be regarded as justifying the conclusion reached at the trial, the view of the trial judge as to where the credibility lies is entitled to great weight especially where there is conflict of testimony. In our view, the conclusion of the trial judge is amply supported by the evidence and we see no reason to interfere.

We are equally not persuaded by Mr. Mpungu’s other submission with regard to the behaviour of Mary at the time of the attack. In the cool and detached atmosphere of the court-room, it is so much easier to criticise the reactions of ether people in times of crisis. We think that, as far as possible, allowance must be made for individual characteristics when judging the behaviour of others in such circumstances. The true test in such cases is not what a reasonable man sitting in a court—room could do but whether the witness, giving allowance for his back-ground and the situation he was in, would have reacted in the manner he did. In this case Mary said that when her friend was attacked in the house she ran to the verandah while raising an alarm. Mr. Mpungu questioned the probability of her running out but staying on in the verandah. We are unable to agree, in the circumstances that this was improbable. Judging from the record, it would seem that the appellants followed her immediately after she went on the verandah. There is no evidence as to how long she stayed there before the assailants cut her outside. Even if she was on the veranda for some minutes, her behaviour cannot justifiably be impugned, for she may have thought no ham would come to her subsequently, or she may have feared to ran far out without knowing what had become of her friend who was being attacked inside the hut, or she may have stopped where she did lest there be other companions of the assailants outside. All these are possibilities which cannot be excluded especially as she was never cross—examined on the matter.

On the issue of alibi, wile, we agree with Mr. Mpungu that it is not for the accused to prove it, we do not think that where none is set up, the trial judge is required to speculate as to whether or not an alibi is available to the accused. The appellants did not set up any alibi worth considering at the trial. They contented themselves with rehearsing what they did the morning following the murder. We do not think on the evidence it was open to the trial court to assume from that evidence that the appellants set up an alibi. The learned trial judge correctly thought they did not.

The remaining question is the reliance placed by the trial court upon the sole identification of Mary to convict the appellants. A conviction based solely on visual identification evidence invariably causes a degree of uneasiness because such evidence can give rise to miscarriages of justice. There is always the possibility that a witness though honest may be mistaken. For this reason, the courts have over the years evolved rules of practice to minimise the danger that innocent people may be wrongly convicted. The leading case in East Africa is the decision of the former Court of Appeal in Abdalla Bin Wendo and Another v. R. (1953), 20 EACA 166 cited with approval in Roria v. R. (1967) EA 583. The paragraph which has often been quoted from Wendo (supra) is at page 168. The ratio decidendi discernible from that case is that:—

(a) The testimony of a single witness regarding identification must be tested with the greatest care.

(b) The need for caution is even greater when it is known that the conditions favouring a correct identification were difficult.

(c) Where the conditions were difficult, what is needed before convicting is ‘other evidence’ pointing to guilt.

(d) Otherwise, subject to certain well known exceptions, it is lawful to convict on the identification of a single witness so long as the judge adverts to the danger of basing a conviction on such evidence alone.

The safe—guards laid down in “enc1o are in our view adequate, if properly applied, to reduce the possibility of a miscarriage of justice occurring. It will be observed that there is no requirement in law or practice for corroboration. In applying Wendo there have sometimes been references to the need for corroboration where the only evidence connecting the accused with the offence is the identification of a single witness. We think that this is not correct. First, there is clear statutory provision that for the proof of any fact, a plurality of witnesses is not necessary: see s. 132 of The Evidence Act (cap.43). Secondly, there is no particular magic in having two or more witnesses testifying to the identity of the accused in similar circumstances. What is important is the quality of the identification. If the quality of the identification in not good, a number of witnesses will not cure the danger of mistaken identity, hence the requirement to look for ‘other evidence’.

Where the case against an accused depends wholly or substantially on the correctness of one or more identifications of the accused, which the defence disputes, the judge should warn himself and the assessors of the special need for caution before convicting the accused in reliance on the correctness of the identification or identifications. The reason for the special caution is that there is a possibility that a mistaken witness can be a convincing one and that even a number of such witnesses can all be mistaken. The judge should then examine closely the circumstances in which the identification came be made, particularly, the length of time the accused was under observation, the distance, the light, the familiarity of the witness with the accused. All these factors go to the quality of the identification evidence. If the quality is good, the danger of a mistaken identity is reduced but the poorer the quality, the greater the danger.

In our judgment, when the quality of identification is good, as for example, when the identification is made after a long period of observation or in satisfactory conditions by a person who knew the accused well before, a court can safely convict even though there is no ‘other evidence to support to identification evidence; provided the court adequately warns itself of the special need for caution. If a more stringent rule were to be imposed by the courts, for example if corroboration were required in every case of identification, affronts to justice would frequently occur and the maintenance of law and order greatly hampered.

When, however, in the judgment of the trial court, the quality of identification is poor, as for example, when it depends solely on a fleeting glance or on a long observation wade in difficult conditions; if for instance the witness did not know the second accused before and saw him for the first time n the dark or badly lit room, the situation is very different. In such a case the court should look for ‘other evidence’ which goes to support the correctness of identification before convicting on that evidence alone. The ‘other evidence’ required may be corroboration in the legal sense; but it need not be so if the effect of the other evidence available is to make the trial court sure that there is no mistaken identification. A good example is the case of Wasajja v. Uganda (1975) EA 181. The coincidence of a person previously identified behaving strangely by putting up a fabricated alibi of his movements at the time the offence was committed or telling lies in some material aspect of his evidence can, in a proper case, amount to ‘other evidence’ sufficient to support a conviction.

In the instant case the learned trial judge properly warned himself of the dander of convicting on the evidence of identification alone and cautioned himself on the possibility of a mistaken identity. He was satisfied that the quality of identification was good in this case. The judge found Mary an honest witness and accepted the correctness of her identification. Mary saw the appellants first cutting the deceased albeit in a badly lit hut but she saw then later outside on the verandah under bright moon light.

The appellants carried a torch in the hut and we do not agree with Mr. Mpungu that Mary could not see what the appellants, whom she knew well before, were doing. On the verandah she had ample opportunity to see them clearly when they began cutting her. In the judgment of the trial court there was no chance of any mistake in identification. We agree. We can see no reason to think that the learned trial judge was wrong in concluding that Mary positively identified the three appellants as the assailants. We think that the circumstances in this case were such as to make Mary’s identification safe to support a conviction.

In any event there was evidence accepted by the learned trial judge. Mary knew all three appellants well before and in her case it was more recognition than mere identification. There was evidence at least from Kazimbye that soon after the attack Mary named all the three appellants to the first people to answer the alarm. The trial judge accepted the evidence Nasimu (PW6) that A2 had uttered threats against the life of the deceased the day before the murder. It seems that A2 had contracted gonorrhea and he foolishly believed that the deceased had bewitched him. All this was in our judgment clear evidence which went to support the correctness of Mary’s identification of the three appellants and we are satisfied that they were rightly convicted of murder.

Before leaving this judgment, we would like to comment on one other point raised by Mr. Mpungu regarding the reference in the judgment of the court below to what Magidu told those who answered the alarm. It is indeed unfortunate that any mention was made of what Magidu, who did not testify, may have said. We are however, satisfied that the judge did not in any way rely upon anything that Magidu may have told the witness which is hearsay but, that he decided the case on the admissible evidence which have reviewed earlier in this judgment, and that no miscarriage of justice has occurred.

This appeal is accordingly dismissed.

DATED AT KAMPLA this 5th day of October 1978.

Sgd: (M. Saied)

Sgd: (P. Nyamuchoncho)

Sgd: (F. M. Ssekandi)

Mr. P.S. Mpugu of Mpugu & Balikuddembe Advocates for the Appellant

Mr. Emesu, Senior State Attorney, for the Director of Public Prosecutions.

I certify that this is a

true copy of the original

(M. Ssendegeya)


▲ To the top