We are currently facing the challenge of improving treatments for psychiatric disorders such as major depression. Notably, antidepressants have an incomplete efficacy, mostly due to our limited knowledge of their action. Here we present a theoretical framework that considers the distinction between instructive and permissive causality, which allows formalizing and disentangling the effects exerted by different therapeutic strategies commonly used in psychiatry. Instructive causality implies that an action determines a specific effect while permissive causality allows an action to take effect or not. We posit that therapeutic strategies able to improve the quality of the living environment or the ability to face it, including changes in lifestyle and psychotherapeutic interventions, rely mainly on instructive causality and thus shape the individual's ability to face the psychopathology and build resilience. By contrast, pharmacological treatments, such as selective serotonin reuptake inhibitors, act primarily through a permissive causality: they boost neural plasticity, i.e. the ability of the brain to change itself, and therefore allow for instructive interventions to produce beneficial effects or not. The combination of an instructive and a permissive action represents the most promising approach since the quality of the living environment can shape the path leading to mental health while drug treatment can increase the likelihood of achieving such a goal.