Are traumatic memories indelible?

Created with the van der Kolk (1994) ingest, which put the wiki’s two trauma sources into direct disagreement on a single, consequential question.

What both sides agree on

This is not a dispute about where trauma lives. Both sources agree on a surprising amount:

• Trauma is held sub-cortically and in the body, not as ordinary narrative memory. See traumatic-memory, core-response-network.
• It is re-lived rather than recalled — expressed as somatic sensations, conditioned responses, and states rather than as a dated autobiographical event.
• It is triggered by arousal / bodily state, and persists for decades unmodified by ordinary experience.
• The therapeutic target is the subcortical machinery, reached bottom-up through the body, not top-down through talk.

Given that much agreement, the disagreement is sharp and specific.

The disagreement: permanent, or reversible?

Van der Kolk: indelible-but-inhibitable. He adopts LeDoux, Romanski & Xagoraris (1991): cortical lesions prevent extinction of a conditioned fear response, so “once formed, the subcortical traces of the conditioned fear response are indelible, and… ‘emotional memory may be forever.‘” On this reading (with Kolb 1987), PTSD is impaired cortical control over an intact subcortical trace. The trace cannot be deleted; therapy strengthens the cortical/hippocampal brake and relocates the trauma in time, and symptoms re-emerge whenever inhibition weakens (drugs, alcohol, sleep, aging, reminders). Figure 2’s caption is the precise statement: the emotional memories “are thought to be relatively indelible, but their expression can be modified by feedback from the prefrontal cortex.”

Payne/Levine: reversible. For SE, trauma is a dysfunctional dynamical state of the core-response-network, “in principle fully reversible” by shifting the network to a different attractor. Completing the thwarted defensive movement discharges the held activation and lets the memory integrate as an ordinary past event — the state is resolved, not merely inhibited. SE is emphatic that this is not extinction: extinction (per McNally 2007) only overlays a competing association on an intact fear memory and is easily reinstated — which is essentially van der Kolk’s indelibility picture — whereas SE claims a discontinuous change robust to re-evocation.

So the two positions are almost mirror images of the same clinical fact (persistent, arousal-keyed, body-held trauma), and they disagree on whether the persistence reflects a permanent trace under a brake or a stuck state that can be un-stuck.

The irony at the centre: both cite LeDoux

The debate is unusually clean because both sides lean on the same animal-research tradition and reach opposite morals:

• Van der Kolk cites LeDoux’s indelibility result (cortical lesions block extinction → the trace is forever).
• SE cites LeDoux’s escape-completion result (Amorapanth, Nader & LeDoux 2000: letting a rat complete a blocked escape made the conditioning “immediately disappear”) and the threat+restraint PTSD model (Shors et al. 1989) → the blocked action is the pathogen, so unblocking it is the cure.

Whether these are genuinely in conflict or describe different phenomena (extinction of a cue-fear association vs. completion of a defensive motor act) is exactly what is unresolved. See joseph-ledoux, survival-circuits. Note also that the strong “emotional memory is forever” reading van der Kolk quotes has been complicated by the memory-reconsolidation literature (Nader and others, post-2000, not in raw/), which shows reactivated memories can become labile and be altered — a body of work that, if anything, moves LeDoux’s own tradition toward the reversibility pole after 1994.

Why it matters, and what would move it

This is not a philosophers’ quarrel — the whole difference between “manage a permanent condition” and “resolve it” rides on it, and so does the choice between inhibition-strengthening treatments (van der Kolk’s pharmacology; arguably interoceptive-exposure as extinction) and transformation-claiming ones (SE). What would discriminate:

• A test of SE’s differential prediction (movement-specific completion beats generic exercise and beats non-completing re-exposure — see biological-completion), run by non-affiliated investigators. SE staking reversibility on a mechanism nobody has tested is the debate’s weakest link on that side.
• Long-term relapse data: indelibility predicts return-of-fear under stress/reinstatement; genuine reversibility predicts robustness to re-evocation.
• Reconsolidation-based erasure results (propranolol-during-reactivation and similar) bear directly on whether the trace is truly permanent.

The wiki holds it open, with two first-hand sources on opposite poles and the shared source (LeDoux) cited by both — a textbook case of the kind of contradiction the wiki exists to record rather than resolve.