Abstract
Currently, mental disorders are usually conceptualized as a hidden causal factor, manifested by its symptoms. This notion rests upon the reflective latent model, which is implicitly at work every time complex symptomatology gets summarized by a single number or a categorical state.
The present paper reflects on the quantitative, testable implications of this psychometric model and shows how its restraints are untenable for most mental disorders. The observed data are instead consistent with mental disorders be- ing complex dynamic systems.
Instead of being treated as interchangeable measures of the same latent factor, symptoms likely act as independ- ent causal entities, directly affecting each other. In recent years, this shift in ontological stance toward psychopathology has laid a basis for adapting the network theory.
Under this theory, a mental disorder is a relatively stable emergent state, which arises due to a pronounced and recurrent interaction of causally linked symptoms. It is discussed how models embedded within the network theory can help provide insight into the etiopathogenesis of mental disorders and address clinical intervention.
In conclusion, limits and future challenges to the network theory are discussed.