Early memories don't hide because they're shameful. They hide because they fragmented before language could hold them. A child experiences something—pain, terror, violation, overwhelming sensation—but the words to describe it do not yet exist. The narrative mind is not yet developed. The event is recorded in the body, in sensation, in affect, in the nervous system's response—but not in language.
This creates a particular problem: the fragmentation that is universal in memory storage becomes extreme in pre-verbal trauma. The event cannot be integrated into narrative at the moment it occurs because the child has no narrative mind yet. There is only sensation, fear, pain, and the body's response. The nervous system records this. But without language and narrative integration, the fragments scatter across somatic channels, affective patterns, and sensory pathways—with no narrative thread connecting them.
Kaufman identifies this as "distinctive discontinuity" applied to the pre-verbal realm.1 The memory is stored with discontinuity so extreme that conventional memory retrieval (asking "what happened?") will not work. Language-based approaches cannot access what was never encoded in language.
The biological feed is the developing nervous system. In the first two to three years of life, the nervous system encodes experiences in multiple non-verbal channels simultaneously. Sensory impressions register in the cortex. Threat activates the amygdala. The body maps the experience in proprioception and somatosensation. Affect patterns emerge. But the hippocampus, which coordinates these separate systems into coherent episodic memory, is not yet fully developed. And the prefrontal cortex, which would create narrative integration, is still emerging.
When a significant experience occurs in this window, it is encoded in the raw sensory and affective systems but not integrated into narrative consciousness. The child cannot say "this happened to me because..." The child has only the direct experience: sensation, fear, pain, the response of their body, the affect state that was triggered.
This encoding is not stored wrong or incompletely—it is stored in the only way the developing nervous system can store it. It is adaptive. But it creates a problem for later retrieval: how do you access something that was never encoded as "a thing that happened" but only as "a state I was in"?
The systemic feed is the availability of somatic and affective attunement after the event. Can the child's body be held and soothed? Can their nervous system be regulated by proximity to a calm adult? Can their affect be named and mirrored even before the child has language? In these conditions, the pre-verbal fragments begin to organize around the relational experience of being tended to. The trauma is still encoded somatically, but it is held in a relational context.
But if the child is alone after the event, if the affect is not mirrored or named, if the body is not soothed—the fragments remain scattered and uncontextualized. They stay activated. They continue to organize the nervous system's response to threat or intimacy or touch.
Tomkins's concept of distinctive discontinuity becomes crucial here. The discontinuity is the gap between the fragments. But in pre-verbal memory, the discontinuity is more extreme because there is no narrative thread connecting the pieces. A smell might trigger fear. A touch might trigger pain. A voice might trigger rage. But these activations happen separately, without any narrative that says "these are all parts of one event from when I was two years old."
The recovery task is not to reconstruct "what happened"—this may never be possible. The recovery task is to find the discontinuity point: the one accessible fragment that is already activated and can be used as an entry.
This fragment might be:
Any of these can be the distinctive discontinuity point. It is the one piece that is already accessible, already activated. Starting from it, other fragments can gradually be contacted and brought into relationship.
What early memory recovery reveals is that the developing child encodes experiences differently than the adult does. There is no "true memory" from early childhood that is being recovered—there is only the fragmented somatic and affective record of what occurred and what the nervous system learned from it.
Recovery of pre-verbal memory is not archaeological (excavating a buried treasure) but reconstructive (assembling fragments into a new coherence). The child's body learned something from the early experience. The nervous system organized itself around it. But the memory of what happened—if it exists at all—is inaccessible because it was never encoded as memory. It was encoded as nervous system organization.
The deepest implication: healing early trauma does not require remembering what happened. It requires understanding what the nervous system learned—what patterns of response, what expectations about safety or danger, what responses to touch or closeness or authority became organized into the body. Once you can name these patterns and their sources, you can work with them. You can update them. The original event remains largely inaccessible. But the nervous system's response to it can be addressed.
Consider a person who experiences intense panic when being held or touched. The panic seems to emerge from nowhere—there is no narrative memory of abuse, no clear explanation. But in therapy, working with the somatic fragment itself, the person begins to contact something in the body.
There is a constriction in the throat. There is a sense of being unable to breathe. There is a feeling of being trapped. These activate together, as a unit, whenever certain types of touch occur.
The therapist works with the fragment as it is, without demanding narrative. "Stay with the sensation. Don't try to remember anything. Just be with what's happening in your body." The person remains in contact with the panic, the constriction, the trapped feeling—but in a safe container, with a present, regulated adult.
Over time, other fragments begin to emerge. A sense of being held down. A smell—of sweat, of alcohol, of cologne. A sound—of breathing, of a voice. A feeling of being small, powerless, unable to escape. But importantly: no clear narrative emerges. The person may never remember a specific event.
Yet the healing is real. By staying in contact with the fragmented somatic memory—not forcing it into narrative, but allowing the nervous system to process it as a present experience in a safe context—the activation begins to discharge. The body begins to learn that being held can be safe. The panic response begins to attenuate. The fragment, worked with somatically, begins to integrate.
If you are working to recover or understand early memories:
Step 1 — Find the accessible fragment: What is the one somatic, sensory, or affective activation that is already present and accessible? A feeling in your body? A sound that triggers something? A sensation that emerges in certain situations? Do not force it. Find what is already activated.
Step 2 — Activate the fragment safely: In a safe container—therapy, somatic practice, trusted relationship—deliberately activate that fragment. Stay with the sensation or feeling without trying to make it into a story. Notice what is there without forcing narrative.
Step 3 — Allow adjacent fragments to emerge: As one fragment activates, what other sensations, sounds, or feelings emerge naturally? Do not suggest or imagine. Let the nervous system produce what is there.
Step 4 — Resist the urge to narrativize prematurely: Your conscious mind will want to construct a story to hold the fragments together. This is natural. But resist the urge to cement the story too quickly. Let the fragments speak first. The narrative may come later—or it may never come, and that is okay.
Step 5 — Work with the nervous system response, not the historical fact: The healing is not about knowing what happened. It is about the nervous system learning that the activation can be tolerated, that presence is safe, that the response can change. Work with what the body is learning, not with what the mind is trying to remember.
Step 6 — Integration without forced coherence: Over time, the fragments may naturally integrate. Or they may remain somewhat separate. Either way is fine. The point is that they are being processed in safety, and the nervous system is recalibrating. That is healing.
Early memory recovery fails when:
The person is not in a safe enough container: Pre-verbal activation is intense and overwhelming. Without real presence and real safety, the nervous system will not allow the fragments to activate. The person will freeze, dissociate, or shut down rather than feel the full affect.
Narrative is forced too early: The person or therapist forces a story onto the fragments before the nervous system has had time to process them somatically. The story may be plausible, may feel true, but it is constructed rather than recovered. It can prevent actual nervous system healing.
The fragments are so extensively dissociated that no entry point is available: Some very early or very severe trauma fragments so completely that the person has no accessible discontinuity point. Even with safety and support, the entry cannot be found. Different approaches may be needed—longer safety building, group work, other somatic practices.
Additional trauma occurs before integration is complete: The early fragments are beginning to activate and process when new trauma occurs, re-fragmenting what was beginning to cohere. The nervous system returns to protection. The early work must wait.
Evidence: Distinctive discontinuity in pre-verbal memory is documented in neuroscience. The developing brain encodes experiences differently than the adult brain. Children under three have limited narrative memory but robust somatic and affective memory. Somatic therapies that work with pre-verbal activation show measurable changes in nervous system regulation. Clinical observation confirms that working with fragments somatically can produce healing without narrative recovery.
Tensions: Yet tension remains. Can fragments that remain largely inaccessible to consciousness really be understood or healed? Some therapists argue that without narrative, there is no real "memory recovery"—only nervous system calming. Others argue that requiring narrative is itself a harmful demand, one that replicates the original trauma (the original event was pre-verbal; forcing it into language re-traumatizes). The tension between narrative understanding and somatic knowing is real and unresolved.
Open Questions:
Kaufman's distinctive discontinuity principle, rooted in Tomkins's affect theory, reveals something crucial about early memory: it is not that the events are forgotten or repressed in the Freudian sense. It is that they were never encoded in the form that conventional memory retrieval can access. They exist as fragments in the nervous system. This convergence with somatic psychology (which has long worked with pre-verbal activation) challenges the purely cognitive approach to memory recovery.
The tension with cognitive psychology is productive. Cognitive approaches privilege language and narrative as the mechanism of healing. But early memory is pre-narrative. Insisting on narrative may actually prevent access to the deeper somatic and affective organization that holds the early experience.
Where psychology describes memory as a cognitive function (encoding, storage, retrieval), somatic practice reveals that the body remembers before language, and that the body's memory is primary. The nervous system encoded the early experience somatically and affectively before any cognitive integration occurred. Language came later. If you treat memory recovery as a language task, you bypass the actual location where the memory exists.
The tension reveals that consciousness is not where memory lives—not in early trauma. Memory lives in the nervous system, in the body's patterns and responses. Talking about early trauma without engaging the body's activation may feel therapeutic but can also retraumatize—it keeps the person in narrative mind, trying to think their way through something that exists in the body.
The insight neither domain generates alone: early trauma requires somatic recovery work. Memory integration cannot happen purely through narrative. The body must be involved. The person must contact the activation somatically, in safety, and allow the nervous system to process what the mind cannot access. Language becomes integrated later—or not at all—but somatic processing is the primary work.
Where developmental psychology describes the emergence of memory capacity over time—pre-verbal to sensorimotor to narrative—it reveals that early experiences are encoded in the body before language develops. The child learns about safety, threat, intimacy, and power through embodied experience. Only later do these become narratives, stories, conceptual understanding.
The tension reveals that early trauma is stored in the developmental substrate before language exists. To access it, you must work at that substrate level. But most therapy is conducted in language. Language therapy can address post-childhood trauma, but for early trauma, the entry point is somatic and affective—the levels at which the original encoding occurred.
The implication: healing early trauma may require working in different modalities than healing adult trauma. The person may need somatic experiencing, dance, touch (carefully), body-based attunement before narrative work becomes possible. Forcing narrative too early can prevent this essential somatic processing.
Your earliest experiences are stored in your body, not your mind. If you experienced trauma before language developed, you will not find it by trying to remember or free-associate. You will find it through your body's responses—the panic that emerges when held, the rage that activates without clear trigger, the numbness that follows intimacy. These are not false memories or imagination. They are the nervous system's record of early experience. Healing them does not require remembering what happened. It requires staying with what the body is experiencing, in safety, and allowing the nervous system to complete its processing. Your body knows. Your mind may never know. That is enough.
Question 1: If healing early trauma does not require narrative memory, how do you know healing has occurred? What is the evidence that the nervous system has actually integrated something when you have no story to show for it?
Question 2: Some people never recover narrative memory of early trauma but achieve full nervous system healing. Others recover narrative but cannot shift the nervous system's response. Which is more important—the story or the somatic change?
Question 3: Is there a point at which continuing to search for early memories becomes a form of re-traumatization? How do you know when to stop excavating and start accepting that some things will remain inaccessible?