Despite many authors in psychiatry endorsing a naturalist view of the mind, many still consider that mental dysfunctions cannot be reduced to brain dysfunctions. This paper investigates the main reasons for this view. Some arguments rely on the analogy that the mind is like software while the brain is like hardware. The analogy suggests that just as software can malfunction independently of hardware malfunctions, similarly the mind can malfunction independently of any brain malfunction. This view has been critically examined in recent literature. However, other less discussed reasons suggest that mental dysfunctions cannot be reduced to brain dysfunctions because mental dysfunctions are appropriately ascribed at the level of intentional mental states, while brain dysfunctions are solely related to abnormalities in anatomy and physiological processes. This paper questions why such a view would be upheld. The discussion is framed within the interface problem in the philosophy of cognitive science, which concerns the relationship between personal and subpersonal levels of explanation. The paper examines the view that an autonomist perspective on the personal/subpersonal distinction could justify the separation of mental dysfunctions, described in intentional terms, from brain dysfunctions, described in anatomical or physiological terms. Ultimately, the paper argues that the autonomist view cannot be upheld in psychiatry and, consequently, does not provide a principled justification for rejecting the reduction of mental dysfunctions to brain dysfunctions.