The author suggests Charles' analysis should be interpreted with reference solely to Freudian metapsychology. From this perspective, the indisputable interest of the model of melancholic identification in understanding the patient's depressive manifestations gives way to that of primary identification. The intimate, albeit conflicting, dialogue between the two models of identification highlights the (superego) demand for internal coherence inherent in the great metapsychologies of the past, beginning with Freud's.