In April 1997, when Zephany was taken from Groote Schuur Hospital, DNA analysis was available but not as fast, cheap, or widely deployed as it became by 2015. A DNA test in 1997 required a known suspect or a known comparative sample. It required laboratory infrastructure. It required a candidate to test.

The case did not fail to resolve in 1997 because the technology was inadequate. It failed to resolve because there was no suspect to test, no candidate to compare, and no mechanism available to identify either. The crime scene left no DNA evidence. The perpetrator left no trace that the technology of the time could act on without a known suspect.

By 2015, nothing about the technology’s fundamental logic had changed. DNA parentage testing was still: obtain a sample from the candidate, obtain samples from the alleged biological parents, compare, produce a probability result. What had changed was the speed — twenty-four hours versus weeks — and the accessibility — R3,000 rather than a larger cost requiring more institutional resource.

What had not changed at all was the logical dependency: the technology needed a candidate. The candidate was provided by a school hallway encounter, a selfie, and a father’s recognition of his daughter’s face.

The case sits at the intersection of human pattern recognition and laboratory science. Neither one, alone, was sufficient. The recognition produced the candidate. The science confirmed the connection. Together, they resolved a case that had been open for seventeen years.

It is a small and specific story about how investigations work, embedded in a much larger and more complicated story about what it means to be found.

ADVERTISEMENT
Advertisement
Share.