Essay · Trauma & Neuroscience
The Brain (Not the Body) Keeps the Score
For a decade, the dominant metaphor of trauma told us the body bears the scars of what the mind cannot face. New neuroscience suggests the metaphor, while compelling, is wrong in ways that matter.
In 2014, Bessel van der Kolk published what would become one of the most influential books in the history of trauma psychology. The Body Keeps the Score offered a powerful, compassionate account of how traumatic experience reshapes the person who survives it — not just psychologically, but physically. The body, van der Kolk argued, literally holds the memory of what the mind cannot process. Trauma lives in the flesh, in the fascia, in the patterns of breath and bracing that persist long after the danger has passed (van der Kolk, 2014).
It is a beautiful metaphor. It is also, according to a growing body of computational neuroscience, not quite right — and the inaccuracy is not merely academic. How we conceptualize where trauma lives shapes everything about how we think it can be healed.
The emerging view, articulated with particular force in a 2026 paper by Steven Kotler, Michael Mannino, Glenn Fox, and Karl Friston published in Frontiers in Systems Neuroscience, is this: the body does not keep the score. The brain keeps predicting it (Kotler et al., 2026).
What van der Kolk got right — and where the metaphor leads astray
To be clear: van der Kolk’s contribution to trauma science is enormous. He drew serious clinical and scientific attention to the embodied dimension of traumatic experience at a time when psychiatry remained stubbornly cognitive and verbal. His observations — that trauma survivors show altered patterns of breathing, posture, arousal, and interoception — are real and important. His invocation of Antonio Damasio’s somatic marker framework, the role of the amygdala, and prefrontal-limbic interactions reflects genuine engagement with the neuroscience of his era (van der Kolk, 2014).
But the book’s most memorable language — that trauma “lives in the body,” that it is stored in tissue outside the mind’s reach — has been taken fully and literally by many readers and many clinicians. That reading implies that the body is somehow an independent archive, that traumatic memory is inscribed in muscle and fascia rather than in neural circuitry. And that, Kotler and colleagues argue, is where the metaphor goes wrong (Kotler et al., 2026).
“The body participates in trauma, but as messenger, not archive. What endures after trauma is not a memory lodged in non-innervated tissue but a collapse of flexibility — a loss of the brain’s ability to fluidly switch among semi-stable network states.” (Kotler et al., 2026)
The paradox, as Kotler and colleagues note, is that the book unintentionally reinforces the very mind-body duality it claims to dissolve (Kotler et al., 2026). Every somatic symptom a trauma survivor experiences originates in neural processes. As Damasio himself argued in Descartes’ Error, we should not treat rational and emotional processes as separate systems — they are fully integrated and complementary (Damasio, 1994). There is no territory of the body not subject to the brain’s renderings.
Predictive processing: how the brain makes a world
To understand the alternative account, it helps to understand how contemporary neuroscience thinks about perception itself.
The brain is not a passive receiver of sensory information from the outside world. It is, at every moment, an active prediction engine. Rather than waiting for signals to arrive and then interpreting them, the brain continuously generates predictions about what it expects to sense — what the next heartbeat will feel like, what the stranger across the room is likely to do, whether the sound around the corner is threatening or benign. Incoming sensory signals are processed not raw, but as prediction errors: the difference between what was expected and what arrived.
This framework — known as predictive processing, or active inference — has become one of the most influential theories in cognitive and systems neuroscience. In its most rigorous formulation, articulated by Karl Friston through the free energy principle, the brain’s fundamental task is to minimize surprise: to maintain a model of the world that contains the least unexplained sensory data (Friston, 2010). Every perception, every emotion, every bodily sensation is the product of this ongoing generative process.
Critically, the brain does not weight all predictions equally. It assigns precision — something like confidence — to different predictions. High-precision predictions carry more weight in determining what we perceive and how we respond. And precision itself can be modulated: the brain can choose, in effect, to trust some signals more than others. A process known as sensory attenuation allows the brain to down-weight signals that are likely to be imprecise noise; in healthy functioning, this keeps perception flexible and context-responsive (Kotler et al., 2026).
Trauma as a disorder of prediction, not storage
Here is where the predictive processing framework transforms our understanding of PTSD.
Trauma does not store a memory in the body. What trauma does — computationally — is over-weight the precision of threat predictions. After extreme danger, the brain assigns excessive confidence to the expectation that danger persists. The prior belief that “I am unsafe” becomes deeply, steeply engrained in the brain’s generative model. Mathematically, this is like being trapped in a narrow valley in a landscape of possible beliefs: the sides are steep (high precision), and escape requires enormous energy (Kotler et al., 2026).
The consequences of this are exactly the symptoms we recognize in PTSD. Hypervigilance arises because the brain is searching the environment for evidence of danger with outsized attention. Flashbacks occur because threat-encoded memories are constantly primed to activate. Avoidance follows from the brain’s attempt to prevent prediction error. The “score” the body appears to keep is not stored in tissue. It is an artifact of circular inference: the brain predicts danger, senses arousal (racing heart, tight chest), and takes that arousal as confirmation that danger persists (Kotler et al., 2026). Convergence-divergence zones — neural hubs that integrate the parameters of a memory and its associated emotion — may become overly sensitive, triggering the cascade of conscious trauma experience in response to even distantly related cues (Damasio, 1989; Meyer & Damasio, 2009).
|
Healthy brain Metastable inference Neural networks fluidly shift between semi-stable states. The brain updates its threat model in response to new evidence. Arousal is recognized as imprecise noise. |
Traumatized brain Frozen prediction Danger priors are over-weighted with excessive precision. The brain is trapped in narrow threat-confirming loops. Arousal is misread as proof that danger persists. |
The neuroimaging literature is consistent with this picture. In PTSD, researchers reliably find a characteristic dissociation: the amygdala — central to threat detection — is hyperactivated, while the medial prefrontal cortex, which plays a critical role in top-down regulation and dampening threat responses, is hypoactivated. The more severe the symptoms, the weaker the prefrontal response (Shin et al., 2006). The brain’s regulatory hierarchy loses its ability to correct its own predictions.
Research on working memory and cognitive control reinforces this account. Reduced working memory capacity — a core marker of executive function — has been shown to predict the severity of intrusive memories even after accounting for anxiety and depression (Bomyea & Amir, 2011). Trauma’s persistence reflects a breakdown of top-down regulation rather than a bottom-up somatic residue: when prefrontal systems can no longer suppress or update threat representations, sensory and interoceptive signals dominate belief updating, creating the illusion that the body keeps score.
Metastability: what trauma destroys
The concept that unifies this picture is metastability — the brain’s capacity to continually shift between semi-stable network configurations in response to context. A healthy brain is not rigidly stable, nor chaotically disordered. It lives on a knife’s edge: organized enough to function, flexible enough to adapt. Neural coalitions assemble transiently, disassemble, and reform in new configurations. This is what allows us to fluidly shift between states of vigilance and rest, threat-detection and creative play, contraction and expansion. Hellyer and colleagues demonstrated that reduced metastability arising from structural connectome damage was associated with diminished cognitive flexibility and impaired information processing (Hellyer et al., 2015).
Trauma erodes this fluidity. The brain gets stuck — not in its body, but in a narrow basin of its own probability landscape. The dynamic repertoire shrinks. Network switching slows. The brain cannot easily move from threat-state to safety-state because the threat predictions are too precisely weighted, the sides of the valley too steep (Kotler et al., 2026).
This reframing matters enormously for treatment. If trauma is stored in the body, the therapeutic metaphor is excavation: find what was buried and release it. If trauma is a failure of metastable inference, the therapeutic metaphor is something more like training or calibration: the goal is to restore the brain’s capacity for flexible prediction-updating, to give the generative model new evidence that safety is possible, to widen the valleys so movement becomes possible again. Research confirms that when individuals with PTSD train cognitive control — through mindfulness, working memory practice, or executive-function exercises — symptoms decrease and regulation improves (Kotler et al., 2026).
Why so many different treatments can work
One of the most compelling features of the metastability framework is that it dissolves a long-standing puzzle: why do such radically different therapeutic approaches — exposure therapy, EMDR, mindfulness, somatic practices, psychedelics, vigorous exercise — all show meaningful effects in PTSD?
From a storage model, this diversity is hard to explain. From a metastability model, the answer is coherent: each of these approaches, in its own way, restores flexible coupling between large-scale brain networks. Each introduces variability into a system that has become too rigid. Each provides the nervous system with new precision-weighted evidence — through novel sensorimotor experience, through top-down regulatory practice, through chemically-induced increases in neural entropy — that the threat model needs updating. The mechanism is not specific content but dynamic reorganization (Kotler et al., 2026).
The hopeful implication
Resilience research offers a final, important piece of the picture. The majority of people exposed to traumatic events do not develop chronic PTSD. Bonanno and colleagues found that across large longitudinal cohorts, approximately 65.7% of trauma-exposed individuals follow a resilient trajectory (Bonanno et al., 2015). Most recover, often without formal therapeutic intervention (Bonanno, 2004).
This fact is more naturally explained by an inference model than a storage model: under typical conditions, the brain rebalances its predictive architecture, restoring metastability and flexibility on its own. Pathology arises not from the permanence of what was stored, but from the failure of that natural recalibration (Kotler et al., 2026).
This is, if anything, a more hopeful picture than the storage model. It frames the traumatized brain not as indelibly marked, but as temporarily stuck — as a system that has lost the capacity for movement and can, under the right conditions, find it again. Our task, as clinicians and as communities, is to provide the conditions — safety, challenge, novelty, co-regulation — under which the brain can update its model of the world.
The body does not keep the score. The brain keeps predicting it. And predictions, unlike scars, can change.
References
Bomyea, J., & Amir, N. (2011). The effect of an executive functioning training program on working memory capacity and intrusive thoughts. Cognitive Therapy and Research, 35(6), 529–535. https://doi.org/10.1007/s10608-011-9369-8
Bonanno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the human capacity to thrive after extremely aversive events? American Psychologist, 59(1), 20–28. https://doi.org/10.1037/0003-066X.59.1.20
Bonanno, G. A., Romero, S. A., & Klein, S. I. (2015). The temporal elements of psychological resilience: An integrative framework for the study of individuals, families, and communities. Psychological Inquiry, 26(2), 139–169. https://doi.org/10.1080/1047840X.2015.992677
Damasio, A. R. (1989). Time-locked multiregional retroactivation: A systems-level proposal for the neural substrates of recall and recognition. Cognition, 33(1–2), 25–62. https://doi.org/10.1016/0010-0277(89)90005-X
Damasio, A. R. (1994). Descartes’ error: Emotion, reason, and the human brain. Putnam.
Friston, K. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11(2), 127–138. https://doi.org/10.1038/nrn2787
Hellyer, P. J., Scott, G., Shanahan, M., Sharp, D. J., & Leech, R. (2015). Cognitive flexibility through metastable neural dynamics is disrupted by damage to the structural connectome. Journal of Neuroscience, 35(24), 9050–9063. https://doi.org/10.1523/JNEUROSCI.4648-14.2015
Kotler, S., Mannino, M., Fox, G., & Friston, K. (2026). The body does not keep the score: Trauma, predictive coding, and the restoration of metastability. Frontiers in Systems Neuroscience, 20, 1812957. https://doi.org/10.3389/fnsys.2026.1812957
Meyer, K., & Damasio, A. (2009). Convergence and divergence in a neural architecture for recognition and memory. Trends in Neurosciences, 32(7), 376–382. https://doi.org/10.1016/j.tins.2009.04.002
Shin, L. M., Rauch, S. L., & Pitman, R. K. (2006). Amygdala, medial prefrontal cortex, and hippocampal function in PTSD. Annals of the New York Academy of Sciences, 1071, 67–79. https://doi.org/10.1196/annals.1364.007
van der Kolk, B. A. (2014). The body keeps the score: Brain, mind, and body in the healing of trauma. Viking Press.
