The two appellants were convicted of rape in the Regional Court. The complainant and her boyfriend, Khuduga, testified that at approximately 20h30, the first appellant struck Khuduga with a stone from behind. While Khuduga went to seek help, both appellants forcibly dragged the complainant to a nearby informal settlement where they took turns to have sexual intercourse with her without her consent. The complainant identified both appellants as her attackers, though visibility was poor due to darkness with only distant shop lighting. The appellants advanced a different version: the second appellant met the complainant in a tavern, they shared drinks, and later had consensual sexual intercourse at his parental home. Both appellants were arrested the following day after Khuduga pointed them out. Common cause facts included: the appellants were in each other's presence that evening, there was an altercation between the first appellant and Khuduga, the second appellant had sexual intercourse with the complainant, visibility was poor, and the first appellant lived in the same area as the complainant and Khuduga.
The first appellant's appeal against conviction and sentence was dismissed. The second appellant's appeal against conviction (insofar as it might properly be before the court) and his appeal against sentence were dismissed.
The binding legal principles established are: (1) Prior knowledge or acquaintance between an identification witness and an accused provides a significant safeguard against mistaken identification, even in conditions of poor visibility (R v Dladla; S v Zitha applied). (2) Contradictions between a witness's oral evidence and their police statement should not be summarily rejected without careful evaluation of underlying factors such as language and cultural differences between the witness and the statement-taker, particularly where the witness is unsophisticated and of modest educational level (S v Mafaladiso applied). (3) Minor contradictions within a witness's evidence or between witnesses on non-material matters may suggest absence of fabrication rather than unreliability. (4) Where faced with two mutually destructive versions, a court is justified in assessing probabilities and may reject the defence version as false only if satisfied beyond reasonable doubt after considering the evidence as a whole (Stellenbosch Farmers' Winery v Martell; S v Trainor; S v Mbuli applied). (5) Appellate interference with sentence on the striking disparity criterion requires the appellate court to have formed a definite view as to the sentence it would have imposed and a striking degree of disparity between that sentence and the one imposed (S v Pieters; S v Matlala applied).
The court noted that the application by the second appellant for leave to appeal against conviction was irregular because no such appeal had been before the Full Court, and leave ought not to have been granted by the Supreme Court of Appeal. The majority judgment of the Full Court also inadvertently erred in purporting to dismiss an appeal by the second appellant against conviction when no such appeal was before it. Despite these irregularities, the court nonetheless considered whether the second appellant was correctly convicted and found the conviction unassailable. The court also remarked that the case was expressly noted as having "no precedential significance" at the outset of the judgment.
This case is significant in South African criminal law for reaffirming established principles regarding identification evidence, particularly emphasizing that prior knowledge of an accused by a witness provides a significant safeguard against mistaken identification. It demonstrates the proper approach to evaluating discrepancies between oral testimony and witness statements, particularly where language and cultural barriers exist in statement-taking. The judgment reinforces the cautionary approach to single witness testimony in sexual offence cases while showing that such evidence can be sufficient where the witness is found credible and reliable. It also illustrates the application of the test for competing versions and the assessment of probabilities in criminal trials, requiring the court to reject a defence version as false beyond reasonable doubt only after considering all evidence. On sentencing, it confirms the limited scope for appellate interference based on the striking disparity criterion.