A philosophical question divides the field of addiction research. Can a psychological disorder count as an addiction absent a common underlying physical basis (neurological or genetic) for every case of the disorder in the category? Or is it appropriate to categorize a disorder as an addiction if the symptoms of and diagnostic criteria for it are sufficiently similar to those of other disorders also classified as addictions—regardless of whether there is some underlying physical basis common to each case of the disorder? The question concerns the scope and validity of the scientific concept of addiction and, more broadly, what is required for a psychological concept to count as scientific. The case of pathological gambling (PG) raises this question nicely. ‘‘Should pathological gambling be considered an addiction?’’ asks Howard Shaffer (2003, 176). He specifies the question further (2003, 177–178): ‘‘When clinicians and scientists identify a behavior pattern as an addiction, even if they can identify it reliably according to DSM criteria, how do they know that it is indeed an addiction?’’ He warns that, as it stands, ‘‘the concept of addiction represents a troublesome tautology’’ (2003, 178): a subject S is addicted if and only if S engages in repetitive behavior with negative consequences against S’s better judgment. The problem with this concept is that it provides no way to distinguish behavior that cannot be controlled from behavior that is merely in fact not controlled. This ‘‘lay’’ concept is of little scientific value, Shaffer (2003, 179) argues, saying that ‘‘for addiction to emerge as a viable scientific construct . . . investigators need to establish a ‘gold standard’ against which the presence or absence of the disorder can be judged.’’ PG has no such gold standard, says Shaffer. What would such a gold standard be? Shaffer suggests ‘‘neurogenetic or biobehavioral attributes’’ (ibid.). Such attributes will have to be identified because ‘‘if pathological gambling represents a primary disorder orthogonal both to its consequences and the laws of probability, then clinicians and scientists should be able to identify the disorder without knowing the winning or losing status of the gambler’’ (ibid.)...