Scientific Objectivity and Criminology
To begin, we argue that the criminal-law definition of crime is logical only when considered in reference to the system of law that produces it. Thus when that legal definition is used to study the causes of crime, it results in a tautology—meaning it is true by definition— and therefore suffers from significant shortcomings associated with tautological thinking (White 1958). How is this a tautology? Conceptually, there is no distinction between the concept of crime and its measurement—the criminal law. In the relevant literature on tautological arguments, we can say that this kind of statement about crime is simply a reference to itself (White 1958). It is similar to saying that “a crime is an act that violates criminal law because it is a crime.” The tautological nature of the criminal-law definition of crime makes it appear that there is no way to err when we measure crime as a violation of law as the data are only about violations of the criminal law.
The problem with the criminal-law definition of crime is one of validity and reliability. A concept is valid when it measures what it claims to measure. On the surface—and absent any critical analysis of the criminal law—the definition of crime appears to be valid. When a researcher says he or she has measured a violation of criminal law in the journal Criminology, for example, we typically assume that the measure is a valid indicator of crime. This may not, however, be the case, since not all crimes are reported and recorded. In fact, the validity of crime statistics was once seriously questioned within the discipline (for a review see Skogan 1974). As Skogan (1974, 27) points out, a segment of criminologists seriously questioned crime statistics, noting that they “are merely artifacts . . .” and he suggests that perhaps criminologists ought “not to use them for research purposes at all.” In fact, crime was often treated as an organizational outcome that was created by variations in the creation and application of criminal law (McCleary, Nienstedt, and Erven 1982). This view, however, has changed and crime has become reified by the discipline of criminology—it appears as a real reflection of individual behavior when it is constantly used to study individuals who are “criminal.” Criminologists no longer view crime as a simple artifact of society and organizations. Today, crime data are viewed as valid when studying crime because they are believed to represent the actual underlying criminal behavior—even if all violations of criminal law are not reported. As a result, the validity of crime is rarely given much thought and crime statistics are typically accepted as valid—as if the debate about the definition of crime is settled. We argue that crime statistics are dependent and represent what is unique about society and its organizations at a particular point in time as opposed to something about human behavior that is independently measured.
Can we restate the idea that a crime is a violation of criminal law in some way that establishes its independence? One way to do this would be to discover the nature of the behaviors that are defined as crime in the criminal law. If the criminal law is a conceptually valid concept, then when applied, all the behaviors that share the same set of characteristics that make them crimes ought to be defined as crime by the criminal law. If, for instance, we were able to say that all crimes have characteristics X1, X2, X3, and X4, then the criminal law should define all behaviors with those characteristics as crimes and exclude those behaviors that do not have those characteristics. Moreover, any measure of crime would need to count those individuals who possessed characteristics X1, X2, X3, and X4. If this is not the result, then there is a problem. The problem is that the criminal law concept used to define crime is really a measure of the behaviors that law identifies as crime and that are applied to people’s behaviors. Part of the issue is whether the criminal law, as both a concept and a measure, does what we think it does—it identifies the unique behaviors that are crimes, and identifies those behaviors using criteria that can be replicated independently from the law in action. In other words, the important question lurking here rather quietly behind the scenes is whether the legal definition of crime is objective and scientific. That question can only be addressed by asking whether we can identify the rules that shape the construction of law and to verify that the creation of the behavioral rules of law would produce the same set of behaviors the criminal law identifies as crimes. We argue that the criminal law is not representative of what might be derived as the nature of crime. We will return to this issue of the nature of crime later.
Crime is also problematic because it is not reliable. Reliability refers to the consistency of a measurement. Unfortunately, the definition of crime changes constantly as law changes so that measurement is not a reliable indicator of behavior over space and time. Thus any reliable measure of crime must be independent of space and time so that it produces consistent results wherever and whenever it is applied. This is not possible for the discipline of criminology when it relies on the state definition of crime. Thus to rectify the tautological nature of the concept of crime, it is necessary to establish some description of crime that establishes its independence from law—from the measurement of law breaking—so that it is more reliable across time and place.