Diagnosis in psychiatry is not well based in science, mainly because we do not know the causes of most mental disorders, and are forced to identify them by their signs and symptoms, Diagnosis could be made more scientific by using external sources of validity, such as biomarkers for endophenoypes, but there is not enough knowledge do so. In general, attempts to apply reductionist models to the mind are limited in principle because neuroscience studies mental phenomena on a different level, and tends to exclude psychosocial factors. In practice, diagnoses tend to be used largely as heuristics leading to treatments believed to be effective. Moreover, there is no well-defined boundary between psychopathology and normality.