The Illusion of Permanence

Memory, Influence and Identity

Memories feel stable. It is their most important quality – and their most consequential flaw.

What we remember, we experience as fact rather than reconstruction. The past appears fixed, authentic, our own. This impression is not merely useful; without it, identity could not exist. It is, nevertheless, a fiction.

Since the late 1990s, neuroscience has documented a mechanism whose implications remain, even now, significantly underestimated. Every memory a person calls into conscious awareness is at that moment permanently deleted from long-term storage. What remains exists only as volatile content held in a separate, temporary state – nowhere else, and only briefly. In that state, it is open. Susceptible to whatever is present during these hours: a reaction, a word chosen with care or without it, the framing of a question. When the memory is eventually written back into permanent storage, it is not the same memory. The process is known as reconsolidation.

Nothing in what returns to storage signals that anything has changed. It does not present itself as a revised version. It carries the full weight of the original – the sense of always having been exactly so – with no awareness of alteration, no trace of what preceded it. What was changed and what was there before are, by every measure available to the mind, identical.

Elizabeth Loftus and John Palmer established as early as 1974 how little is required to produce this effect. Eyewitnesses to the same accident were questioned using different words: one group was asked how fast the cars were going when they "contacted" each other; the other, when they "smashed" into each other. A single substitution, nothing more. The second group subsequently recalled higher speeds – and broken glass that had not been there. Nader, Schafe and LeDoux (2000) and Schiller and colleagues (2010) confirmed the same mechanism at the neurobiological level. Memory is rewritable at every activation. Not in principle. In fact.

That memory should be constructed this way is not a failure of design. It reflects a long evolutionary preference for adaptability over fidelity – a system that integrates experience, adjusts to new contexts, and revises its own record accordingly proved more viable than one that did not. Identity, built on this foundation, operates by the same logic: it absorbs change without registering it as such. The chameleon does not know it has changed colour.

What served adaptation well becomes, in a different setting, a structural vulnerability – wherever someone brings knowledge and method to bear on a memory while it remains open.

Any activation opens the window. Not only deliberate questioning, but every form of contact: a response, a term introduced into the conversation, an interpretation offered before one was sought. Memory makes no distinction between the intentional and the incidental. It registers what is present.

In legal practice, this has been understood for decades. The right to question a witness is, in effect, write access to that witness's memory. Skilled lawyers work this window with precision – through the construction of questions, the selection of language, the architecture of the examination itself. The memory that leaves the room is not the one that entered it. This is not manipulation in any legal sense. It is craft.

Conversation always alters memory. What differs is not the mechanism but what is brought to the table – and by whom. Someone speaking casually about what occupies them opens the window over content that stays within the register of the exchange. Someone who opens up to a psychologist brings the most sensitive material of their inner life into a room where interpretation is both expected and professionally authorised – and where the consequences of that interpretation are often not fully understood even by the psychologist. This is surgery on an open heart, conducted on the assumption that the person holding the instrument knows precisely what they are doing.

Reconsolidation is not specialist knowledge. It belongs to the foundations of scientific psychology. To know it and interpret early regardless is not an oversight.

Examined from outside, established psychological conversation formats reveal a structural consequence that is barely avoidable within them. Every school provides an interpretive frame – prescribing which questions are asked, which patterns carry meaning, which explanations are offered. Interpretations come early because the method requires it, delivered in the school's own vocabulary, into a system that at that moment has no defence.

The schools contradict one another on fundamental questions. What one identifies as cause, another treats as symptom. What one dissolves, another reinforces. The same starting point can yield entirely different anchored memories depending on the format applied. Whether any school arrives at the correct interpretation cannot be determined – yet all have rewritten the memory in their own terms. None of these versions can be distinguished from the original. The original no longer exists.

Anyone entering such a conversation anticipates stimulation – ideas to return to later, judgements to form in one's own time. What is not known: a single question is sufficient to alter a memory, as Loftus demonstrated. An interpretation, once spoken, alters it always – at the moment of hearing, before any response to it is possible. Even disagreement engages an already altered memory. The original is not recoverable.

This is not a detail to be initialised away in terms and conditions.

Protection from this mechanism requires a form of conversation that understands it and accepts its implications. I call this reconsolidation-sensitive conversation architecture. The original is not touched while it remains open – no interpretation while the memory is active, no question that frames what has been said before it has been fully expressed. First accounts are preserved. Silence is held. Meaning is not offered before the person has arrived at it themselves.

When interpretation becomes necessary – and it does, for experience without integration remains inert – it is deferred to a moment when the memory has restabilised. And it does not engage the original directly. The work is done on new material, developed together within the conversation, structurally analogous to the original experience but without access to it. The original remains untouched. The interpretation is placed alongside it – offered, not inscribed. Whatever transfer occurs happens internally, within the person, and what emerges does so in a condition of stability, under their own authority.

Slowness, in this framework, is not temperament. It is the structural condition under which memory and identity can be protected.

Memory is the substance of what we hold to be true – about ourselves, about others, about what has taken place. Access to memory is access to the person. Not as metaphor. In the neurobiological sense. What is altered there alters who a person understands themselves to be – without indication, without remedy, and without their knowledge, because what now resides there carries the full conviction of having always been so.

This occurs daily. In millions of conversations. In settings designed to generate trust – and in which the mechanism described here operates without being named, understood, or constrained. To know that it exists is where protection begins.

References

• Schacter, D. L. (1999). The Seven Sins of Memory: How the Mind Forgets and Remembers.
• Tempel, T., & Pastöter, B. (2014). Abrufeffekte im Gedächtnis: Ein Überblick zur aktuellen Grundlagenforschung. Hogrefe Publishing.
• Schiller, D., Monfils, M.-H., Raio, C. M., Johnson, D. C., LeDoux, J. E., & Phelps, E. A. (2010). Preventing the return of fear in humans using reconsolidation update mechanisms. Nature, 463(7277), 49–53.
• Nader, K., Schafe, G. E., & LeDoux, J. E. (2000). Fear Memories Require Protein Synthesis in the Amygdala for Reconsolidation After Retrieval. Nature, 406, 722–726.
• Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13(5), 585–589.

---

Publications The Conversation Contact Main Page

---

Copy link Share by email

---