On 30 December 2009, two heavily pregnant women (the first complainant being 9 months pregnant) were walking through a bushy area in Mathe village, Tsholotsho, with their 9-year-old cousin. They encountered a man carrying a stick who initially asked if they had seen his cattle. He then grabbed the boy, ordered the complainants off the road into the bush, and assaulted them when they resisted. The man proceeded to rape the first complainant three times and forced the second complainant to perform oral sex on him twice, all in the full view of the 9-year-old boy. After the assault, he allowed them to return to the road. The complainants reported the incident to their aunt, described the assailant and his clothing, leading to the arrest of the appellant. The complainants did not ordinarily stay in that area, coming from Gwayi 2 area in Nyamandlovu district, and did not know the locals. However, the 9-year-old boy lived in the neighbourhood and knew the appellant.
The appeal against both conviction and sentence was dismissed. The appellant's conviction on 3 counts of rape and 2 counts of aggravated indecent assault was upheld, as was the sentence of 18 years imprisonment for the rape counts and 15 years imprisonment for the aggravated indecent assault counts (total 33 years), with 8 years suspended on the usual conditions, resulting in an effective sentence of 25 years imprisonment.
The binding legal principles established are: (1) Identification evidence from a child witness can be reliable where the child had prior knowledge of and familiarity with the accused over a substantial period; (2) Where a young witness provides specific, verifiable details about the accused that are not disputed or challenged at trial, such details can constitute independent corroboration of the identification evidence; (3) Good identification based on prior knowledge does not require corroboration, but poor identification does (applying S v Nkomo 1989 (3) ZLR 117 (S)); (4) An appellate court will not interfere with a trial court's findings on credibility of witnesses absent a misdirection; (5) An appellate court will not interfere with a sentencing court's discretion unless there is a misdirection, and it is not sufficient for an appellant to merely argue that the sentence is too severe (applying S v Chiweshe 1996 (1) ZLR 425 (H) and Mkombo v the State HB 140/10).
The court made observations about the particularly callous and sadistic nature of the crimes, noting the appellant "exhibited callousness of the highest order and appeared to derive sadistic pleasure in abusing heavily pregnant women in the full view of a 9 year old child" and that "He got his just deserts." While not strictly necessary for the legal determination, the court emphasized the extreme vulnerability of the victims (being heavily pregnant, with one being 9 months pregnant) and the psychological trauma inflicted by forcing these acts to occur in front of a young child. The court also observed that the trial magistrate "stretched backwards to find corroboration" even though the identification was already good identification that did not require corroboration.
This case is significant in Zimbabwean criminal law for its treatment of identification evidence from child witnesses. It demonstrates the courts' approach to assessing the reliability of identification evidence from young witnesses, particularly where the witness had prior familiarity with the accused. The judgment clarifies that where a child witness provides identification evidence containing specific, verifiable details that are not disputed or challenged, such details can constitute corroboration of the identification. The case also illustrates the court's willingness to uphold substantial sentences for particularly callous sexual offences involving vulnerable victims, and reaffirms the principle that appellate courts will not interfere with sentencing discretion absent a misdirection. The judgment emphasizes the distinction between "good identification" (based on prior knowledge) and "poor identification" in determining whether corroboration is necessary.