Significant changes have occurred in the developmental role of Hox genes, even within groups of arthropods that already have complex body plans and many different segment types. This is hard to reconcile with the 'selector gene' model for Hox gene function. Selector genes act as stable binary switches that direct lineages of cells to adopt alternative developmental fates. This model suggests that the regulation of selector genes can only evolve through mutations that alter the identity of whole developmental compartments -in the case of Hox genes, whole segments. Once segments have evolved distinct morphology and function, such mutations will result in dramatic homeotic transformations that are unlikely to be tolerated by natural selection. Thus we would expect the developmental role of these "master control genes" to become frozen as body plans become more complex. I argue for a revised model for the role and regulation of the Hox genes. This provides alternative mechanisms for evolutionary change, that may lead to incremental changes in segment morphology. The summation of such changes over long periods of time would result in differences in Hox gene function between taxa comparable to the effects of gross homeotic mutations, without the need to invoke the selective advantage of hopeful monsters.