The release of DSM-5 and the preparations for the launch of the ICD-11 provoked a series of critiques of psychiatric classification, which continues to depend largely on clinical description. Among the immediate problems are those of arbitrary diagnostic thresholds, tendency to reification, rigid category boundaries, comorbidity, diagnostic 'epidemics' and differential diagnostic dilemmas. We argue that many of those problems stem from the polythetic-operational definitions of psychiatric categories, which thereby come to lack an organizing prototype-directed or gestaltic intelligibility principle. We illustrate these issues by briefly examining the current operational diagnosis of schizophrenia, its demarcation from affective illness and the status of the spectrum concept and the prodrome of schizophrenia. We point out that European research on schizophrenia always allocated an important diagnostic weight to a certain prototypical trait core of the illness, phenomenologically indispensable for its demarcation from other, nonschizophrenic psychotic conditions. We believe that the notion of self-disorder (reflective of the structural alterations of subjectivity), itemized into its various aspects in the Examination of Anomalous Self-Experience scale, is an important step forward in a more precise psychopathological articulation of that core, strengthening its clinical and research utility.