It can be circumvented only if criminologists use official data to compute crime rates and to identify criminals or delinquents. That proposed strategy does not necessarily entail acceptance of the reactive conception, especially in view of an argument about etiological theories that purport to answer two of the four major criminological questions, those having to do with variation in the crime rate and with individual differences as regards criminality. Any such theory will be empirically indefensible unless it encompasses (1) some etiological condition as the independent variable, (2) the frequency of some type of behavior as an intervening variable, (3) an official criminality variable (for example, an official crime rate), and (4) a reactive variable, one which pertains to the behavior of legal officials and supposedly determines the connection between the intervening variable and the dependent variable. The argument bears on Marxist and conflict criminology only insofar as advocates of those perspectives genuinely pursue etiological theories about crime. Finally, apart from any substantive consideration, criminological theories will remain defective until criminologist adopt formal theory construction. (Publisher abstract)