David Herbert has been the primary exception to the geographical mainstream by suggesting that the social processes underlying crime pat- terns (including the process ofcontrol) are of key explanatory significance. But he also points out that this significance remains largely untested in the geographic literature (I977. ZI6-l9: I979. ll8-I9: I982. 21-29). Although Herbert believes that the themes of spatial distribution and environment fall most clearly into the domain ofa geographical approach. he suggests that these cannot form the exclusive terms of reference for the geographer: as with other types of urban phenomena. spatial pattems of crime are the outcome of social processes. Thus he contends