Psychedelics may work by restoring the brain's readiness for change, not by driving change directly
Neuroscientist Gül Dölen argues that psychedelics do not induce plasticity so much as restore the brain's capacity to undergo it. If she is right, the conditions inside the therapeutic window matter as much as the drug itself.
The therapeutic promise of psychedelics has long been framed around neuroplasticity: the substances promote new neural connections, and those connections enable change. Gül Dölen, a neuroscientist whose work centers on psychedelic mechanisms, argues that framing is imprecise in a way that matters clinically. Psychedelics, she contends, do not induce much plasticity on their own. What they appear to do is restore the brain’s capacity to undergo plasticity, a property she describes as metaplasticity.
The distinction is more than semantic. Dölen has found that despite binding to different receptors, various psychedelics converge downstream on this same effect. The implication is that the receptor-level differences between substances, the feature that has long anchored pharmacological classification, may matter less than the shared downstream function they produce. What the substances share, on her account, is not a binding target but a destination: the restoration of conditions under which the brain can again be shaped.
Where Dölen’s framing becomes most concrete is in its account of how psychedelics interact with memory and learned behavior. Her argument is that these substances help treat conditions like post-traumatic stress disorder by reopening the window of time during which patterns were originally formed. Once that window is open, memories and behaviors can be reconfigured to fit present circumstances. The therapeutic work, on this account, happens inside the reopened window, not because the substance is driving it.
The way that these psychedelics are helping for diseases like PTSD is that they are enabling a sort of reopening of that window of time when you learned, so that you can reconfigure the memories and patterns of behavior in a way that is appropriate for your current circumstances.Gül Dölen
That account offers a plausible explanation for something practitioners and observers have noted but struggled to systematize. Joe Hudson, whose work focuses on emotional and nervous-system change, describes high-dose psilocybin as capable of changing nervous systems but doing so very undependably. Under Dölen’s model, the undependability is not a failure of the drug’s mechanism. It is the expected outcome of a mechanism that restores conditions for change without determining what change occurs. What happens inside the reopened window depends on the environment, the therapeutic structure, or the experiences that accompany it. The substance is permissive, not directive.
This framing carries real consequences for how clinical trials and therapeutic protocols should be designed. If psychedelics are primarily creating a window rather than driving outcomes within it, then the conditions inside that window, the quality of therapeutic support, the setting, the preparatory work, become the active ingredient in a meaningful sense. A study measuring outcomes from drug administration alone would be measuring the window, not what is done with it.
The mechanistic picture is still in motion, and Dölen’s framework should be read as an organizing hypothesis rather than a closed account. Multiple pathways are likely in play, and the field is working out their relative contributions. But the metaplasticity framing is already doing useful conceptual work.
What Dölen’s framework offers, even at this early stage, is a tool for thinking about variability, the hard-to-predict spread of outcomes that has made psychedelic-assisted therapy difficult to evaluate and harder to scale. If the substance restores readiness and the environment determines direction, then outcome variability is structurally built in, not a sign that the underlying mechanism is absent. That reframe does not make the clinical path easier. It may, however, point toward where the design problems actually sit.