South African criminology's structural aetiology is in crisis. This dissertation offers a novel account of the nature, origin, severity, implications, and possibilities of that crisis. It suggests that, rather than a normative problem, it should be understood as an empirical one, related to the challenge of crime prevalence measurement. The question of crime prevalence patterns and trends has mistakenly been treated as trivial. This dissertation conducts meta-theoretical and historical analyses to reveal a fundamental criminological quandary: making defensible and testable claims about aggregate crime prevalence patterns and trends is at once both indispensable and impossible. This dilemma is in some respects inherent to the task of primary criminology, but its origin and manifestation are also uniquely crippling and revealing in the South African context. The aetiological crisis is more severe, more fundamental, and more complex than previously thought. In demonstration of this, this dissertation seeks to establish, as defensibly as possible, just one observation about long-term South African crime prevalence trends that would seem to require explanatory effort. It collects official South African police murder statistics over the longest-possible time frame and at the lowest-possible level of aggregation and combines them with census data using Geographic Information System technology. The result is by far the most extensive and defensible possible description of South African long-term crime prevalence patterns and trends. It shows a large, unprecedented, widespread murder rate decrease from 1994 to 2011. This poses problems for existing theory and reveals the discipline's failure to even identify that which is relatively unequivocal and requires explanation. This dissertation concludes that there is an unidentified void at what should be the empirical heart of South African criminology. There is much to gain in engaging head-on the question of how to go about systematic empirical observation in the context of profound ambiguity about the meaning and measure of crime.