This presentation aims to provide an integrative account of relational trauma through the lens of the Predictive Processing framework (Clark, 2016; Friston, 2010). Predictive Processing conceptualizes the brain as a generative model that anticipates incoming sensory data, reduces prediction error, and minimizes expected free energy. Central to this process is precision weighting, namely, the confidence attributed to priors, sensory evidence, or policies for action. When early relational trauma is present, precision allocation becomes distorted, producing maladaptive predictive dynamics.


Low precision assigned to priors results in hypervigilance, biased likelihood models, and heightened sensitivity to potential threat. At the same time, when rigid and global negative priors dominate ("I am in danger"), prediction errors accumulate at higher levels of the hierarchy, preventing adaptive learning and reinforcing avoidance or dissociation as pragmatic policies. These strategies reduce short-term distress but consolidate cycles of maladaptive self-protection, a process described as "bad bootstraps" and loss of metastable attunement. As a consequence, individuals remain locked into rigid, global patterns, unable to flexibly update priors or use emotions as adaptive signals.


Clinically, this framework sheds light on the paradoxical presentations of traumatized patients: simultaneous hypervigilance and avoidance, unstable yet rigid predictive models, and diminished capacity to incorporate corrective experiences. Psychotherapy can thus be conceptualized as a relational space for recalibrating precision, fostering epistemic trust, and restoring generative flexibility. By providing a stable environment and co-constructing new predictive models, therapy enables patients to reengage with novelty, update maladaptive priors, and develop more adaptive strategies for minimizing prediction error and free energy over time.


This predictive processing perspective not only integrates developmental, neurocognitive, and clinical findings but also offers testable hypotheses for advancing therapeutic interventions in attachment-related trauma.