Confusing operational predicaments and cognitive explorations: comments on Rossmo and Snook et al.

Snook, Taylor, and Bennell (Applied Cognitive Psychology, 2004 vol. 18, pp. 105-121) have carried out some elegant, but rather simple-minded experiments, which demonstrate, unsurprisingly, that people can employ limited geometrical principles in a sensible way. Having drawn these principles from a literature on the geographical patterns found in criminal behaviour, Snook et al. suggest that most people could geographically profile crimes without the need for much training or assistance from commercially available geographic profiling software. Rossmo (Applied Cognitive Psychology, 2005 vol. 19, pp. 651-654) challenges this suggestion by claiming that the experiments ignore the practical limits to the application of geometrical principles derived from criminals' spatial behaviour. These challenges are more appropriately construed as sources of error in the modelling of criminal behavioural geography than as operational constraints. Current weaknesses in the modelling of criminal actions show the need for a deeper understanding of offenders' decision processes. This debate also draws attention to the need for research on how investigators can actually make use of decision support systems derived from studies of criminal behaviour. Geographic profiling is not a distinct process but a set of hypotheses about patterns in criminal cognitions and actions. These hypotheses are worthy of further development and test rather than being left in the realm of detective hunch and conjecture. Copyright © 2005 John Wiley & Sons, Ltd.