Memory is not a recording. It is a reconstruction—every time you remember something, your brain rebuilds that memory from fragments, fills gaps with inference, and colors the whole thing with current emotional state and belief.1 The thing you remember happening is not what happened; it's what your brain constructs now in response to a cue, shaped by what you've been told, what you've read, what fits your current narrative about yourself.
This means something radical: your past is not fixed. It's being rewritten continuously, and it can be deliberately reshaped by someone who understands how memory reconstruction works.1
Think of memory not as a filing cabinet but as a collaborative storytelling session your brain runs with itself every time you try to recall something. And if someone else is in the room during that storytelling, they can shape the narrative.
Encoding — When something happens, your brain doesn't record it comprehensively. It selects which features to encode based on attention and emotional intensity.1 A traumatic event gets encoded in vivid sensory detail. A boring meeting gets encoded as sparse generalities. This means the intensity of your attention at the moment determines what gets stored.
Storage — Once encoded, memories are stored not as unified files but as distributed patterns across multiple brain regions. The emotional tone lives in the amygdala. The visual details in visual cortex. The spatial context in hippocampus. This distribution means memories are inherently fragmented—no single unified storage location.
Retrieval and Reconstruction — When you retrieve a memory, your brain doesn't pull out a file. It reconstructs the memory by re-activating the stored patterns and filling in the gaps.1 This is where the vulnerability lives: gap-filling is inference, and inference can be shaped.
The Loftus Effect demonstrates this brutally: in her classic experiments, researchers showed participants video of a car accident, then asked questions about what they saw. Participants who were asked "How fast were the cars going when they smashed into each other?" later "remembered" broken glass that wasn't in the video. The verb smashed shaped not just how they described the memory but what they reconstructed as having happened.1
Memories are most vulnerable to modification during and immediately after retrieval. The act of remembering reactivates the memory, temporarily destabilizing it. During this window, new information can be integrated into the reconstruction—and the person won't notice it's been added.1
This creates operational opportunity: if you can get someone to retrieve a memory (tell you the story), then feed them new information during the retrieval window, they will reconstruct the memory incorporating the new information. They won't experience it as being told something false; they will experience it as remembering something they'd forgotten.
The Misinformation Effect:1 People exposed to false post-event information integrate that false information into their memories and become confident in the false memory. The confidence is not manufactured; it's real. They genuinely remember it that way.
Perception appears to be direct—you see what's in front of you. But perception is actually predictive construction. Your brain generates predictions about what should be there given context, then checks incoming sensory data against those predictions.1 What you perceive is the match between prediction and data.
This means perception is vulnerable to prior belief and expectation. Show someone ambiguous visual information and tell them it's a duck, they'll see a duck. Tell them it's a rabbit, they'll see a rabbit. The sensory data is identical; the perception differs because the predictive model differs.
The Implication: You don't perceive reality directly. You perceive your brain's prediction about reality, corrected by sensory data. But if the prediction is strong enough, the brain will interpret the sensory data to match the prediction.1
Magicians exploit this: they guide your attention to where they want you looking (shaping your prediction about where the action is), and your visual system cooperates by seeing what the prediction expects. You're not deceived about what's in front of you; you're constructing perception consistent with where you expect the action.
Lung's framework treats memory malleability and perception construction as exploitable vulnerabilities: understanding how memory gets reconstructed means you can introduce false information into someone's memory and they will integrate it as genuine recollection. Understanding perception construction means you can shape what someone sees by shaping their expectations.
Cognitive psychology (Loftus, Schacter) agrees that memory is reconstructive and vulnerable to misinformation—but frames this as a feature of normal cognition, not a design flaw. Memory flexibility allows us to update beliefs, correct misunderstandings, and integrate new information. The same mechanism that makes us vulnerable to misinformation also makes us capable of learning.
The tension: Is memory malleability a vulnerability to be exploited or a capacity to be understood and integrated? A person who knows their memories are reconstructions can develop meta-awareness—noticing when new information is being integrated, questioning whether a "memory" is actually a reconstruction from suggested information. This shifts the vulnerability from automatic to navigable.
Where psychology describes how memory malleability works (the mechanism of reconstruction and gap-filling), behavioral-mechanics describes how to deliberately exploit that mechanism. The tactical deployment: get someone to retrieve a memory (activate the reconstruction process), introduce false information during the vulnerable window, and they integrate it as genuine recollection.
The psychological mechanism and the tactical deployment are identical in form—only the consciousness differs. A therapist might use memory reconstruction therapeutically (helping someone integrate fragmented trauma memories into coherent narrative). An operative uses the same mechanism tactically (implanting false information into memory). The reconstruction process doesn't know the difference between therapeutic and tactical application.
The insight neither domain produces alone: This reveals that the vulnerability isn't in memory itself but in the asymmetry of knowledge. Memory is malleable—that's how it works. But if you have knowledge of how that malleability works and the other person doesn't, you have leverage. A person who understands their own memory reconstruction can develop epistemic humility (questioning whether a retrieved memory is actual recollection or reconstruction + inference). This doesn't make memory less malleable; it changes what they can do with that knowledge.
Perception is constructive—shaped by expectations and predictions. Behavioral-mechanics describes how to deliberately shape someone's expectations so their perception constructs the reality you want them to see. This is the foundation of magic, misdirection, and propaganda: guide the prediction, and perception follows.
Together, memory malleability and perception construction describe the full cycle of reality-construction: what someone perceives shapes what they encode into memory; what gets encoded shapes what they can retrieve; what they retrieve (as reconstructed) becomes the "evidence" for their beliefs about what's real.
The memory malleability framework assumes that because memories are reconstructions, they're infinitely modifiable—that you can implant any false memory and it will stick. But there's a limit: emotional coherence and identity stability create resistance. You can implant a false detail into a memory. But implanting a false memory that contradicts someone's core identity or emotional narrative is much harder. The person will reject it or reinterpret it.
This suggests that the most protected memories are those most integrated into identity. The vulnerabilities are peripheral memories—the details no one cares much about. This means genuine deep change (shifting someone's core beliefs about themselves and reality) requires more than memory manipulation; it requires destabilizing identity first, then implanting new narratives, then allowing them to integrate into a new identity.
What is the relationship between memory accuracy and identity stability? If someone's core identity rests on a false memory, does revealing the falsity destabilize their sense of self? And does that destabilization create either openness to new narratives or desperate clinging to the false memory as identity-anchor?
Can someone develop the capacity to notice when a memory is being modified in real-time? Meta-awareness of memory reconstruction (knowing that retrieving a memory destabilizes it) might allow someone to "hold lightly" to retrieved memories, noticing when new information is being integrated. What would that capacity look like?
How do repeated false narratives become experienced as memories? If someone hears the same false story multiple times, when does it stop being "information I was told" and start being "something I remember"? Is there a threshold, or is it a gradual shift?