Illusions of Normality: A Methodological Critique of Category-Specific Naming (published with peer commentaries and reply)

Category-specific disorders are perhaps the archetypal example of domain-specificity — being typically defined by the presence of dissociations between living and nonliving naming ability in people following neurological damage. The methods adopted to quantify naming across categories are therefore pivotal since they provide the criterion for defining whether patients have a category effect and necessarily influence the subsequent direction and the interpretation of testing. This paper highlights a series of methodological concerns relating to how we measure and define category (or any) dissociations. These include the common failure to include control data or the use of control data that is inappropriate e.g. at ceiling, unmatched. A review of past cases shows that the overwhelming majority suffers from these problems and therefore challenges conclusions about the purported empirical demonstrations of dissociations and double dissociations in the category specific literature. This is not a refutation of category deficits, but skepticism about the current existence of any convincing empirical demonstrations of category specific double dissociations. As a potential solution, certain minimal criteria are proposed that might aid with the attempt to document category effects that are more methodologically convincing.