The Memory You Lost and Found
Series 05: Who You Are When You Forget
Miriam Torres is 85 and has advanced Alzheimer’s. Her daughter Lorena has not been recognized in two years. Lorena visits every Sunday anyway. She sits beside her mother’s bed at the memory care facility, holds her hand, talks about the week, and receives no indication that her mother knows who she is or hears what she is saying. Two years of Sundays. Two years of visiting a woman who looks at her the way she looks at the aide or the wall or the window.
This Sunday, Lorena has brought nothing special. No recordings, no photographs, no particular plan. She is sitting beside her mother’s bed, holding her hand, humming a lullaby. She is not humming it as a strategy. She is humming it because it is what her body does when she sits beside her mother. Miriam’s lullaby. The one Lorena has known since infancy.
Miriam’s eyes open. She looks at Lorena’s face. She says her name. “Lorena.” Once. Clearly. With the specific tone that mothers have for their firstborn. Then she is gone again. The eyes close. The recognition retreats. Lorena will live on that moment for the rest of her life.
Memories Are Not Files#
The metaphor most people carry about memory is a filing cabinet: memories stored in folders, retrieved by pulling the right drawer. Dementia, in this metaphor, is the files being deleted. The person forgets because the memory no longer exists. The forgetting is permanent. The file is gone.
The neuroscience says something different. Memories are not files. They are patterns of neural activation distributed across multiple brain regions. A single memory of a Tuesday afternoon in 1973 is encoded simultaneously in the hippocampus (the context), the amygdala (the emotion), the visual cortex (what things looked like), the auditory cortex (what was playing on the radio), and the motor cortex (what the body was doing). The memory is not stored in one place. It is stored in the pattern of connections between many places.
What cognitive decline disrupts is often the retrieval pathway, the ability to activate the full pattern on demand, not the pattern itself. The neurons that encoded the memory may still hold their piece of it. The connections between them may be weakened but not severed. The memory may still be there. The door to it may be locked. And sometimes, under specific conditions that neuroscience is beginning to understand, the door can open.
What Recovery Actually Looks Like#
Not Hollywood. Not a sudden restoration of full memory and personality. Not the scene where the person with dementia looks at their child and says “I remember everything.” That scene does not happen.
What recovery actually looks like: a fragment. A name said once. A gesture. A look of recognition that lasts five seconds. Sometimes twenty minutes of lucidity and conversation. Sometimes one word that holds everything. The door opens a crack, a shaft of light falls through, and the door closes again.
The range matters because expectations shape experience. The family that expects full restoration will experience a fragment as a failure. The family that understands the neuroscience will experience a fragment as what it is: evidence that the pattern is still there, that the door can open, and that the conditions under which it opens are worth understanding.
Lorena’s moment was one word. “Lorena.” Spoken once. The word held two years of absence and a lifetime of presence. It was not a return to normal. It was a retrieval, momentary and complete, of the specific neural pattern that connects a mother to her firstborn daughter. The pattern activated. The name came through. Then the activation faded and the silence returned.
Music as a Retrieval Key#
Music is the most powerful and most documented trigger for episodic retrieval in dementia. BML-05.10 covers the neuroscience and the clinical applications in depth. Here, the mechanism matters for understanding why Lorena’s humming opened the door.
Musical memory is stored in brain regions, the cerebellum, supplementary motor area, and basal ganglia, that Alzheimer’s pathology affects later than the hippocampus and prefrontal cortex. A song that was emotionally significant during a formative period is encoded across motor, emotional, and contextual memory networks simultaneously. When the song plays, it activates all of these networks at once. The activation can cascade into associated episodic memories, pulling them into accessibility through the musical channel even when the direct retrieval channel is closed.
Lorena was not using a strategy. She was humming her mother’s lullaby because it was the song her body knew. The lullaby activated the musical memory, which activated the emotional memory, which activated the association between the melody and the person who always sang it, which activated the recognition of the face above the hand holding hers. The cascade took seconds. It produced one word. Then the cascade subsided and the retrieval pathway closed.
Scent as a Retrieval Key#
Olfactory processing is the only sensory pathway that bypasses the thalamus and connects directly to the hippocampus and amygdala. Smell is the most direct neurological route from stimulus to memory. This is why a childhood smell can produce an emotional response in a person who cannot remember where they are.
The research on scent-based memory retrieval in dementia is promising but methodologically challenging. Studies are small. Outcomes are measured inconsistently. Personal significance matters enormously: the generic lavender that produces nothing from one person produces a twenty-minute conversation from another whose grandmother grew lavender. The practical application is specific: identify the personally significant scents from the biographical profile in BML-05.07. The spice that was always in the kitchen. The soap the person used for forty years. The perfume the spouse wore. Each of these is a potential retrieval key calibrated to one specific person’s neural landscape.
BML-05.11 covers the broader sensory approach. Here, the principle is the same as for music: the scent is not a treatment. It is a key. Sometimes the key fits. Sometimes the door opens. Sometimes it does not. The family that has the key is better positioned than the family that does not.
Motor Cues and Touch#
The body performing a familiar movement can reopen retrieval pathways that verbal and visual prompts cannot reach. Harold’s hands on the tools in BML-05.08 activated motor memories that cascaded into procedural knowledge, which cascaded into teaching behavior, which is a form of identity expression. The movement was the key. The identity was behind the door.
Touch is the most direct pathway available when verbal and visual retrieval have closed. Holding a hand. Stroking hair. The physical contact that the body has known for decades activates somatosensory memory, which connects to emotional memory, which can produce recognition that does not arrive through any other channel. Miriam may not have recognized Lorena’s face. She may have recognized Lorena’s hand, the specific feel of the hand that has held hers every Sunday for two years, and the recognition may have traveled from the somatosensory cortex to the language centers that produced her daughter’s name.
This is speculative. The neuroscience of touch-based retrieval in advanced dementia is not well studied. What is documented is that physical contact produces measurable physiological calming in people with advanced dementia, reduces agitation, and increases positive affect. Whether it also produces memory retrieval moments is a question the research has not yet answered definitively. The families who maintain physical contact are not waiting for the research. They are holding the hand because it is the right thing to do, and if the door opens, they are there.
What the Family Should Know#
The retrieval window is not predictable and not controllable. Creating the conditions that make it more likely is not the same as guaranteeing it will happen.
The conditions that increase the probability of a retrieval moment: the right time of day (cognitive function fluctuates; many people with dementia have a best window, often in the morning or early afternoon). A calm environment (overstimulation closes doors; quiet opens them). Personal sensory materials (the specific music, scent, or object with documented biographical significance). Physical presence and touch. Emotional safety (the person who is afraid does not retrieve; the person who feels safe may).
None of these guarantee anything. The family that brings the music every Sunday and plays it at the optimal time in a calm room while holding the person’s hand and using the correct biographical materials may experience nothing for months. The family that does nothing in particular may have a moment arrive unbidden.
The family that tries every approach and experiences no retrieval moments has not done something wrong. The biology is the biology. Some doors are more firmly closed than others. Some doors will not open again. Creating the conditions is the work the family can do. The outcome belongs to the biology, and the biology does not owe anyone a moment.
Lorena, After#
Lorena does not tell her siblings what happened. She does not want to explain it. The moment is hers, the lullaby, the name, the tone. She does not want it discussed, analyzed, or turned into evidence about their mother’s cognitive status. It was not a data point. It was her mother saying her name.
She comes back the next Sunday. She hums the lullaby. She holds her mother’s hand. Miriam does not say her name. Lorena stays the full hour. She comes back the Sunday after that. Miriam does not say her name.
Lorena keeps coming. Not because it worked. Because it could. Because the neuroscience says the pattern is still there, that the door can open, and that the conditions she creates by sitting beside her mother and humming the lullaby and holding the hand are the conditions that give the door its best chance. She has stopped measuring Sundays by whether the moment happens. She measures them by whether she showed up for the possibility.
The door can open. She wants to be there when it does.
How this article connects to others in Blue Mirror.
Sources cited in this article.
- Cuddy, Lola L., et al. "Music, Memory, and Alzheimer's Disease: Is Music Recognition Spared in Dementia, and How Can It Be Assessed?" Medical Hypotheses, vol. 64, no. 2, 2005, pp. 229-235.
- El Haj, Mohamad, et al. "Music Enhances Autobiographical Memory in Mild Alzheimer's Disease." Educational Gerontology, vol. 38, no. 1, 2012, pp. 30-41.
- Herz, Rachel S., and Jonathan W. Schooler. "A Naturalistic Study of Autobiographical Memories Evoked by Olfactory and Visual Cues." American Journal of Psychology, vol. 115, no. 1, 2002, pp. 21-32.
- Glachet, Ophélie, and Mohamad El Haj. "Emotional and Phenomenological Properties of Odor-Evoked Autobiographical Memories in Alzheimer's Disease." Brain Sciences, vol. 9, no. 12, 2019, p. 339.
- Harrison, Briony E., et al. "Touch in Dementia: A Review of the Literature." Western Journal of Nursing Research, vol. 32, no. 6, 2010, pp. 779-797.
