Traumatic memory

The organizing idea of van der Kolk (1994) — that trauma is a memory-system disorder, and specifically one in which the body holds what the narrative cannot. See bessel-van-der-kolk.

The two-system split (Figure 1)

Van der Kolk’s Figure 1 divides memory into:

• Declarative / explicit — facts and events; conscious, verbal, constructive, integrated into existing schemata, and therefore distorted by later experience and current emotional state. Hippocampally mediated.
• Nondeclarative / implicit — skills and habits, emotional associations, and conditioned sensorimotor responses; unconscious, procedural, not symbolically encoded.

His load-bearing claim: trauma interferes with declarative memory but not with the implicit system. So the conditioned emotional response, the somatic sensation, the flashback survive intact and are re-enacted, while the conscious narrative is fragmented, absent, or unreliable. This is why, on his account, the traumatized person “relives” rather than “remembers.”

Why the split happens: arousal breaks the hippocampus

The mechanism is the amygdala/hippocampus dual-memory model (see amygdala for the neural detail). Under moderate arousal the amygdala’s significance-tagging enhances hippocampal consolidation — hence hypermnesia for emotionally important events. Under overwhelming arousal, excessive amygdala activation inhibits hippocampal function, and (per Piaget’s “failure of semantic memory”) the experience is organized “on a somatosensory or iconic level (such as somatic sensations, behavioral enactments, nightmares, and flashbacks).” Norepinephrine’s inverted-U on consolidation lets the same event produce coexisting hypermnesias and amnesias — the paradox at the heart of PTSD.

The clinical name for the extreme case is “speechless terror”: the arousal is so high the event never reaches words or symbols at all.

State-dependency

A corollary van der Kolk stresses: trauma memories are state-dependent. “High states of arousal seem selectively to promote retrieval of traumatic memories, sensory information, or behaviors associated with previous trauma.” Information encoded in an aroused/altered state is retrieved best when that state recurs — which is why a physiological surge (or a drug that produces one, e.g. lactate or yohimbine) can precipitate a flashback, and why dissociative amnesia can lift when the state returns. The trauma memory is not freely accessible; it is keyed to the body’s condition. This is the mirror image of the interoceptive point below: arousal triggers the memory, and the memory triggers arousal — a loop that can escalate subclinical into clinical PTSD.

The interoceptive reading

The wiki files this concept as trauma-literature material, but its centre of gravity is interoceptive. Van der Kolk’s sharpest functional claim is that traumatized people show a “loss of the capacity to use affect states as signals” — feelings stop working as cues to appraise a situation, so the person goes “immediately from stimulus to response without psychologically assessing the meaning of an event.” Read in the wiki’s vocabulary, that is a failure at the top of the interoceptive hierarchy: the bodily signal is present (often over-present) but no longer informs a flexible model of what is happening. And “the misinterpretation of innocuous stimuli… as potential threats” is a prior of danger overriding benign interoceptive/exteroceptive input — interoceptive-inference gone wrong, described decades before the term. Compare schema-guided-symptom-perception and Barrett’s chronic-misprediction account.

What is durable and what is contested

• Durable: the declarative/implicit dissociation, the state-dependency of retrieval, and the clinical observation that trauma is re-lived somatically. These are the parts later work built on.
• Contested: two claims the 1994 paper makes confidently. First, that traumatic flashbacks are “exact replicas” less subject to distortion than ordinary memory — the reconsolidation and false-memory literatures (not in raw/) complicate the “indelible accurate replay” picture the paper leans on. Second, and the one that is a live wiki debate, whether the underlying trace is permanent. See are-traumatic-memories-indelible — van der Kolk says indelible-but-inhibitable; SE says reversible. The whole difference between “manage the trace” and “resolve the state” rides on it.