Paper 07
Extinction versus reconsolidation: two routes to changing a fear
A whitepaper on the two distinct ways a fear memory can be changed, why extinction leaves the original trace intact while reconsolidation rewrites it, and why that difference decides whether fear stays gone.
The problem: fear that comes back
Suppose you learn that a tone predicts a shock. After a few pairings the tone alone makes you flinch, sweat, and freeze. Now suppose we want to undo that. There are two fundamentally different things we could do to the fear, and they are not the same operation even though they can look identical from the outside. One builds a new memory on top of the old one. The other reaches into the old memory and changes it. The clinical stakes of telling them apart are large, because the first route has a built-in failure mode: the fear tends to come back.
This paper is the companion to the foundation paper on the labile window. That paper introduced reconsolidation and named reactivation-extinction as its flagship behavioral application. Here we slow down on the contrast itself: what exactly extinction is, why extinguished fear returns, what it would mean to rewrite rather than compete, and how firm the evidence is on each side.
Extinction: new learning, not unlearning
Extinction is what happens when you present the tone over and over without the shock. The flinch fades. It is tempting to read this as erasure, as if the tone-shock association had been deleted. It is not. The classic and well-established view, running from Pavlov through modern work, is that extinction is new learning. The animal does not forget that the tone once predicted shock. It learns a second, competing thing: that the tone now predicts safety. Both memories coexist. Which one drives behavior at any moment depends on context and cues.
The evidence that the original trace survives extinction is not subtle, and it comes in three named phenomena. These are the fingerprints of a memory that was covered over rather than rewritten.
| Phenomenon | What you do | What happens | What it proves |
|---|---|---|---|
| Spontaneous recovery | Extinguish the fear, then simply wait | Fear returns on its own with the passage of time | The safety memory faded faster than the fear memory underneath it |
| Renewal | Extinguish in one context, test in another | Fear returns outside the extinction context | Extinction was context-bound; the fear memory was not |
| Reinstatement | Extinguish, then deliver an unsignaled shock | Fear to the tone comes back | A reminder of the danger revived the intact original |
Each of these is a demonstration that the fear memory was still there the whole time. Extinction did not touch it. It laid down a safety memory beside it, and under the right conditions the safety memory loses the competition and the fear reappears. This is established, in animals and in humans, and it is the central clinical problem with exposure therapy: exposure is extinction, and extinction relapses.
EXTINCTION
original fear trace stays intact
a new safety trace is laid down alongside it
behavior at any moment = whichever trace currently wins
RECONSOLIDATION
recalling the fear trace makes it briefly labile
it then re-stores in a changed form
behavior = that one trace, now updated
Reconsolidation: changing the original trace
Reconsolidation is the other route. As the foundation paper describes, Nader, Schafe and LeDoux (2000) showed that recalling a consolidated fear memory throws it back into a fragile, protein-synthesis-dependent state. During that brief labile window the original trace itself is open to change, and what re-stabilizes can differ from what you started with. This is not a new competing memory. It is the same memory, re-stored in an altered form.
The distinction is mechanistic, not just descriptive. If you weaken a fear by blocking reconsolidation, you are subtracting strength from the one and only fear trace. There is no second memory waiting to win back control, because you did not create one. The prediction that follows is the crucial one: a fear reduced through genuine reconsolidation should not show spontaneous recovery, renewal, or reinstatement, precisely because the original trace, the thing those phenomena revive, is no longer intact. In animal work this prediction holds up well. The absence of return of fear is the signature people look for as evidence that they touched the original trace rather than layering a competitor on top.
That gives us a clean conceptual test, summarized here.
| Question | Extinction | Reconsolidation update |
|---|---|---|
| What is created | A new safety memory | No new memory; the old one is edited |
| Original fear trace | Left intact | Destabilized and re-stored changed |
| Return of fear (recovery, renewal, reinstatement) | Expected | Predicted absent |
| Depends on a labile window | No | Yes |
| Requires prediction error to work | No | Yes, to open the window |
Reactivation-extinction: making extinction rewrite
Now the idea that connects the two. Extinction is easy, safe, and behavioral. It needs no drug and it is already what exposure therapy does. Its only flaw is that it competes instead of rewriting. Reconsolidation rewrites but is harder to trigger deliberately and, in humans, harder to trust. So the obvious question is whether you can run extinction inside the reconsolidation window and get extinction to rewrite the original trace rather than compete with it.
That is the reactivation-extinction paradigm. The recipe is to present a single reminder cue that reactivates the fear memory and opens the labile window, wait through the window (on the order of ten minutes to an hour), then run ordinary extinction while the trace is still destabilized. The hope is that the safety learning gets written into the reconsolidating original rather than stored as a separate competitor. If it works, you get extinction's safety and convenience with reconsolidation's permanence, and the return of fear does not happen. The rodent version of this result was reported by Monfils, Cowansage, Klann and LeDoux (2009).
Schiller et al. (2010) reported the analogous effect in humans. Participants underwent fear conditioning, then a reminder followed by extinction inside the presumed reconsolidation window. Compared with standard extinction, the reactivation-first group showed no return of fear when tested later, and the authors reported that the effect persisted at a follow-up roughly a year afterward. This is the single most cited demonstration that a behavioral procedure might reach the original trace in people.
The honest evidence picture
Here the paper has to hold two evidence levels apart, and the house rule of this corpus is not to blur them.
Extinction and its return-of-fear phenomena are established. Spontaneous recovery, renewal, and reinstatement are robust, replicated across species and laboratories, and they are the reason relapse after exposure therapy is a real clinical problem rather than a theoretical worry. Nothing in the contested column below weakens any of this.
The animal reconsolidation mechanism is established. Reactivation returns a fear trace to a labile, synthesis-dependent state, and interfering with re-stabilization weakens the original memory in a way that resists recovery, renewal, and reinstatement. This is a strong, mechanistic, replicated body of work.
Reconsolidation-based updating in humans, and reactivation-extinction specifically, is contested. The Schiller result is striking, but the replication record is genuinely mixed. Several careful attempts have failed to reproduce the effect, and the paradigm appears sensitive to conditions that are hard to control: whether the reminder actually induced prediction error and opened the window at all, the timing of the interval, the age and strength of the memory, and individual differences. A protocol that fails to open the window will look exactly like a failed reconsolidation effect while never having engaged the mechanism, which makes null results hard to interpret. The animal-to-human translation gap is wide here, and stating that plainly is a design commitment of this work, not a hedge.
Why the distinction matters clinically
The whole point of separating these two routes is that they predict different clinical futures for the same patient.
Exposure therapy is extinction. It works, often dramatically, but because it creates a competing safety memory rather than rewriting the fear, it carries extinction's relapse liabilities. A patient treated in the clinic, which is one context, can experience renewal when they meet the feared thing in the world, which is another context. A patient who has recovered can be reinstated by a fresh trauma. Spontaneous recovery means gains can erode with time alone. None of this is a failure of effort. It is the structural signature of a treatment that competes instead of rewrites.
If reconsolidation-based updating could be made reliable in humans, it would attack the disorder at its root: the original trace itself, edited so there is nothing intact left to recover, renew, or reinstate. That is the prize, and it is why the field keeps pushing on reactivation-extinction and on pharmacological reconsolidation blockade despite the messy human evidence. Lane, Nadel, Greenberg and Ryan (2015) pressed the same logic into psychotherapy broadly, arguing that durable therapeutic change in general depends on reactivating an emotional memory and letting a corrective experience re-store with it, rather than on building a competing memory beside it. That is a reconsolidation claim about why some therapy sticks and some relapses, and it is worth taking seriously at the right evidence level.
One circuit-level note ties this back to the rest of the corpus. Roozendaal and McGaugh established that noradrenaline in the basolateral amygdala gates how strongly an arousing event is stored. That same gate is what pharmacological reconsolidation approaches reach for from the other end, using beta-adrenergic blockade to damp the noradrenergic signaling during the labile window so that a weaker emotional charge re-stabilizes. Arousal is the dial on trace strength both when the fear is written and when it is rewritten.
What to take away
- Extinction does not erase fear. It creates a new, competing safety memory and leaves the original fear trace intact. This is established.
- The proof that the original survives is spontaneous recovery, renewal, and reinstatement: three ways extinguished fear comes back. These are robust and are the reason exposure therapy relapses.
- Reconsolidation is the other route. Recall makes the original trace labile, and an update installed in that window changes the one trace itself rather than adding a competitor.
- A fear reduced through genuine reconsolidation should not show return of fear, because the trace that recovery revives is no longer intact. In animals this signature holds.
- Reactivation-extinction is the attempt to make extinction rewrite rather than compete, by running extinction inside the reconsolidation window. Monfils et al. (2009) reported it in rodents and Schiller et al. (2010) reported it prevents the return of fear in humans.
- Keep the levels straight. Extinction and its return-of-fear phenomena are established; the animal reconsolidation mechanism is established; reconsolidation-based updating in humans, reactivation-extinction included, is contested, with a mixed replication record and a wide translation gap.
Sources
- Nader, Schafe and LeDoux (2000). Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature 406, 722 to 726. doi:10.1038/35021052
- Schiller, Monfils, Raio, Johnson, LeDoux and Phelps (2010). Preventing the return of fear in humans using reconsolidation update mechanisms. Nature 463, 49 to 53. doi:10.1038/nature08637
- Monfils, Cowansage, Klann and LeDoux (2009). Extinction-reconsolidation boundaries: key to persistent attenuation of fear memories. Science. (Original rodent reactivation-extinction result.)
- Bouton (2004). Context and behavioral processes in extinction. Learning and Memory. (Framework for extinction as new learning, and for spontaneous recovery, renewal, and reinstatement.)
- Lane, Nadel, Greenberg and Ryan (2015). Memory reconsolidation, emotional arousal, and the process of change in psychotherapy. Behavioral and Brain Sciences 38, e1. doi:10.1017/S0140525X14000041
- Roozendaal and McGaugh, on basolateral-amygdala noradrenergic gating of memory consolidation (review literature; corpus meta-analysis: noradrenergic-signaling-in-the-basolateral-amygdala).
Evidence tags used above (established, emerging, contested) follow the corpus convention and keep the animal-to-human translation gap visible.