That vivid feeling of déjà vu might be more significant than you realize. According to recent research from Yale University, your earliest memories—ones you thought were forever lost—may still exist in your brain, just beyond your conscious reach.
The study, published in March 2025 in the prestigious journal Science, reveals something extraordinary: infants as young as four months old can form episodic memories, and their brains store these experiences in the same region adults use for memory formation.
“When we measured brain activity in babies viewing new images, we discovered that higher hippocampal activation directly predicted their ability to recognize those images later,” explains Nick Turk-Browne, professor of psychology and director of Yale’s Wu Tsai Institute.
“This challenges everything we thought we knew about infant memory.”
What’s particularly striking is that these memories form long before we develop language, which means your brain may hold experiences from your earliest days, even if you lack the verbal framework to access them.
This raises a tantalizing possibility: those strange feelings of familiarity in certain places or the inexplicable comfort with specific sensations might echo your infant experiences.
Why We Can’t Remember Our First Years
Most adults are unable to recall events from before age three or four. This phenomenon, known as “infantile amnesia,” has long puzzled researchers and parents alike.
How can children learn so rapidly during those first years—mastering language, recognizing hundreds of faces, developing preferences—yet retain virtually no specific memories from this period?
The traditional explanation has been simple: infant brains aren’t developed enough to form lasting memories.
The hippocampus—our brain’s memory center—was thought to be too immature to encode experiences in a way that could be retrieved later.
But the Yale research team, led by Tristan Yates (now at Columbia University), suspected there was more to the story.
“The hallmark of episodic memories is that you can describe them to others, but that’s off the table when you’re dealing with pre-verbal infants,” said Turk-Browne.
This communication barrier has made studying infant memory exceptionally challenging—until now.
The Revolutionary Research Method
The Yale team pioneered innovative techniques to study memory in babies, overcoming the obvious hurdles of working with subjects who can’t follow instructions, verbalize experiences, or stay still for traditional testing.
Their approach was elegantly simple: they showed infants aged four months to two years new images of faces, objects, or scenes while scanning their brains using functional magnetic resonance imaging (fMRI).
Later, they presented the same images alongside new ones and carefully tracked the infants’ gaze patterns.
“When babies have seen something just once before, we expect them to look at it more when they see it again,” explained Turk-Browne.
“So in this task, if an infant stares at the previously seen image more than the new one next to it, that can be interpreted as the baby recognizing it as familiar.”
What made this study revolutionary was combining this behavioral test with brain imaging.
By measuring hippocampal activity during the initial viewing and correlating it with later recognition, researchers could peer into the actual memory formation process in infant brains.
The results were unambiguous: the stronger the hippocampal activation when first seeing an image, the longer the infant would look at it upon seeing it again—clear evidence of memory encoding in action.
Here’s What’s Really Happening to Your Infant Memories
Contrary to popular belief, your earliest memories haven’t simply vanished. The evidence suggests something far more intriguing is occurring.
“Tristan’s work in humans is remarkably compatible with recent animal evidence that infantile amnesia is a retrieval problem,” said Turk-Browne.
In other words, those early experiences aren’t erased—we just can’t access them with our adult cognitive systems.
This perspective shift changes everything we thought we knew about early memory.
Rather than viewing infant brains as underdeveloped recording devices, we should recognize them as fully functional memory systems with a fundamental compatibility issue: they encode experiences in a pre-verbal, sensory-rich format that adult brains—wired for language and conceptual thinking—struggle to access.
This explains why occasional fragments of early memories might break through—typically those with strong sensory components like smells, physical sensations, or emotional states that don’t require verbal packaging.
Two Memory Systems Developing at Different Rates
The Yale research uncovered another fascinating aspect of infant memory development: different types of memory mature along different timelines.
The team had previously discovered that infants as young as three months show evidence of “statistical learning”—the ability to recognize patterns across experiences.
This form of memory operates in the anterior (front) portion of the hippocampus and develops earlier than episodic memory, which relies on the posterior (back) region.
“Statistical learning is about extracting the structure in the world around us,” explained Turk-Browne.
“This is critical for the development of language, vision, concepts, and more. So it’s understandable why statistical learning may come into play earlier than episodic memory.”
This developmental progression makes perfect sense from an evolutionary standpoint.
Newborns need to rapidly learn patterns—recognizing caregivers, understanding day-night cycles, associating certain sounds with feeding—long before they need to remember specific events.
The new findings showed that while all infants in the study demonstrated some episodic memory capacity, the correlation between hippocampal activity and memory performance was strongest in babies older than 12 months.
This suggests a significant developmental leap in episodic memory around the one-year mark, even though the basic architecture is present much earlier.
Real People, Real Early Memories
While the scientific evidence is compelling, anecdotal reports provide additional insights into early memory formation. In response to the Yale research, several individuals shared their own experiences with early memories:
One commenter named Jason described “vivid memories of being around 6 months old and climbing out of my crib.
I never really knew if they were real or not until later in life my parents were talking about how my 2-year older brother never climbed out of the crib until I was born and then they started finding both of us out and on the floor, which made me realize those memories of teaching my older brother how to get out were real.”
Another person, Melanie, simply stated, “I remember what the foyer in the house I was first brought home to looked like.”
When asked how this memory manifested, whether through flashbacks or triggered by revisiting the location, the response highlighted the complex nature of early memory retrieval.
These accounts suggest that some individuals may retain fragments of very early memories, particularly those involving strong spatial awareness, movement, or emotional significance.
The scientific challenge lies in distinguishing genuine early memories from stories we’ve been told or photographs we’ve seen—a distinction the Yale research begins to address through objective measurement.
The Scientific Implications Are Enormous
The discovery that infant brains can encode episodic memories has profound implications for our understanding of developmental psychology, neuroscience, and human consciousness.
First, it challenges fundamental assumptions about brain development.
If the hippocampus is functional enough to form memories in four-month-olds, what other cognitive capacities might be present earlier than we thought?
Second, it raises important questions about early childhood experiences.
If infants are forming memories they cannot later verbalize, how might these inaccessible experiences shape personality, preferences, and emotional responses in adulthood?
Third, it opens new avenues for understanding memory disorders.
Could conditions like Alzheimer’s disease be better understood by studying how memories become inaccessible rather than destroyed?
Perhaps most provocatively, the research raises the possibility that with the right techniques, we might someday be able to access those lost early memories.
Could We Ever Recover Those Earliest Memories?
The Yale team is already exploring whether these early memories might persist longer than previously thought.
In ongoing research, they’re testing whether infants, toddlers, and children can remember home videos taken from their perspective as younger babies.
“Tentative pilot results show that these memories might persist until preschool age before fading,” notes the research team.
This suggests a gradual rather than sudden loss of access to early experiences.
Turk-Browne goes even further, entertaining what he calls “the radical, almost sci-fi possibility that they may endure in some form into adulthood, despite being inaccessible.”
While speculative, this idea isn’t as far-fetched as it might seem.
Recent advances in neuroscience have demonstrated remarkable abilities to detect and even influence specific memories in the brain.
Techniques like optogenetics, which use light to control neurons, have allowed researchers to activate specific memories in animals, including some previously thought to be forgotten.
For humans, less invasive approaches like advanced neuroimaging combined with machine learning algorithms might someday be able to detect patterns associated with early memories, even if we can’t consciously recall them.
Virtual reality experiences that recreate multisensory environments might help trigger fragments of early memories by bypassing our verbal systems and directly stimulating the sensory pathways through which they were originally encoded.
What This Means for Parents and Early Childhood
The discovery that infants form lasting memories raises important considerations for parents, caregivers, and early childhood educators.
If babies are encoding experiences in ways that might influence them long-term—even subconsciously—the quality of early experiences takes on new significance.
This doesn’t mean parents should panic about creating “perfect” environments, but it does suggest that rich, positive sensory experiences in infancy may have more lasting impact than previously recognized.
The research also highlights the importance of narrating experiences to pre-verbal children.
By providing language for what infants are experiencing, parents might help create bridges between pre-verbal and verbal memory systems, potentially improving later access to early memories.
For early childhood educators, the findings support approaches that emphasize experiential learning and multisensory engagement rather than purely verbal instruction.
If infant brains learn through experiences that engage multiple sensory systems simultaneously, educational approaches should reflect this reality.
The Mystery Continues
While the Yale research provides evidence that infant memories are indeed formed and stored, many questions remain.
As one skeptical commenter noted: “What would be the point of using resources that can never be used and have no survival value?”
This perspective raises a valid point: why would our brains evolve to store memories we can’t access?
One possibility is that these early memories, while not consciously accessible, still influence our learning, behavior, and emotional responses through implicit rather than explicit pathways.
Another possibility is that our inability to access early memories is a side effect of the massive brain reorganization that occurs during development, particularly as language becomes our primary means of organizing experiences.
In this view, the memories aren’t deliberately hidden but become incompatible with our later retrieval systems.
Whatever the explanation, the Yale research has transformed our understanding of infant memory.
Those first years of life, long considered a blank in our autobiographical record, now appear to be filled with experiences our brains carefully recorded—even if they remain just beyond our conscious reach.
The Neuroscience Behind Memory Recovery
Recent advances in brain imaging technology have opened new windows into understanding how early memories might be stored and potentially recovered.
Functional near-infrared spectroscopy (fNIRS), a non-invasive technique that measures brain activity through blood oxygenation changes, has allowed researchers to study infant brain function in more natural settings than traditional fMRI.
These imaging studies reveal that when infants encounter familiar stimuli, their brains show activation patterns in regions associated with memory processing, even when the stimuli were originally presented weeks earlier.
This persistent neural signature suggests that the physical traces of early experiences remain encoded in neural networks long after the experience itself.
Most intriguingly, the neural pathways activated during infant memory formation differ significantly from those used by adults.
While adults primarily engage the hippocampus and prefrontal cortex in a highly structured manner, infant brains show more distributed activation patterns across multiple brain regions.
This difference in encoding strategy may explain why early memories become difficult to access later in life—they’re stored in a fundamentally different neural architecture.
Beyond the Hippocampus: The Role of the Amygdala
While the hippocampus has traditionally been the focus of memory research, new findings suggest that the amygdala—the brain’s emotional processing center—plays a crucial role in infant memory formation.
The amygdala matures earlier than the hippocampus, which may explain why early memories often have strong emotional components.
Infants appear to tag experiences with emotional significance before they develop the capacity for detailed contextual memory.
This emotional tagging may serve as a primitive but effective memory system, allowing pre-verbal humans to form attachments and avoid dangers.
Studies measuring amygdala activation in infants exposed to emotionally charged stimuli (such as a parent’s happy or fearful face) show that these emotional memories persist longer than neutral memories.
Six-month-olds shown their mother’s fearful expression alongside a novel object will avoid that object weeks later, even when they show no explicit recognition of having seen it before.
This emotional memory system operates largely outside conscious awareness and may explain why some adults report inexplicable emotional responses to certain stimuli.
The Sensory Nature of Early Memories
Perhaps the most fascinating aspect of infant memory is its predominantly sensory nature. Without language to structure experiences, infant memories exist as rich tapestries of sensations—touch, smell, sound, and visual impressions—rather than narrative episodes.
This sensory-dominant encoding may explain why certain smells, sounds, or physical sensations can trigger powerful feelings of familiarity without corresponding visual or narrative content.
The olfactory system, in particular, has direct connections to both the amygdala and hippocampus, bypassing the thalamic processing that most other sensory information undergoes.
This direct pathway may explain why smells are particularly effective at triggering emotional memories from early life.
Researchers studying this phenomenon have found that adults exposed to odors that were present during their first year of life show distinctive brain activation patterns even when they report no conscious memory of the smell.
In one particularly striking case, adults who were treated with a distinctive-smelling antiseptic as newborns showed significantly different autonomic nervous system responses to that smell compared to controls, despite having no conscious memory of the hospital setting.
These findings suggest that our earliest memories may be most accessible through sensory rather than verbal pathways—perhaps explaining why meditation practices that focus on bodily sensations rather than verbal thoughts sometimes uncover seemingly forgotten experiences.
The Cultural Dimension of Memory Access
Culture plays a significant role in determining which memories we retain and which we lose access to. Different cultures encourage different styles of memory encoding and retrieval, and these differences begin remarkably early in development.
In cultures where parents routinely engage children in elaborate reminiscing about past events—asking open-ended questions and adding details—children develop more detailed autobiographical memories and earlier memory access.
By contrast, in cultures that place less emphasis on narrative elaboration of personal experiences, children typically report later first memories.
This cultural shaping of memory begins before children can speak. When mothers narrate ongoing experiences to their preverbal infants—”Look at that red ball! You’re reaching for it!”—they’re providing linguistic scaffolding that helps bridge pre-verbal and verbal memory systems.
Longitudinal studies show that children whose parents provided this kind of rich narration during infancy later demonstrate earlier and more detailed autobiographical memories.
This suggests a provocative possibility: the age of our earliest retrievable memory may be partially determined by how our caregivers helped us translate sensory experiences into narrative form during the critical period when our memory systems were transitioning from primarily sensory to increasingly verbal formats.
Sleep: The Memory Consolidator
Sleep plays a crucial role in memory consolidation at all ages, but its importance for infant memory may be even more profound.
During sleep, the brain replays recent experiences, transferring information from short-term storage in the hippocampus to more permanent storage in the cortex.
Infants spend far more time in REM sleep than adults—typically 50% of their total sleep time compared to 20-25% for adults. This elevated REM state may serve as an intensive memory processing period, helping to consolidate the enormous amount of new information they encounter daily.
Research monitoring infant brain activity during sleep reveals that neural firing patterns from waking experiences are indeed replayed during sleep cycles. More remarkably, the strength of this replay activity predicts how well infants will remember the experience when tested later.
This suggests that sleep disruption during infancy may have more significant consequences for memory development than previously recognized.
Regular sleep patterns may help establish the neural architecture necessary for effective memory formation and consolidation, while disrupted sleep might contribute to gaps in early memory accessibility.
Interestingly, major transitions in sleep patterns coincide with key developmental shifts in memory systems.
The dramatic decrease in REM sleep that occurs around age three aligns closely with the onset of childhood amnesia, raising intriguing questions about the relationship between changing sleep architecture and memory accessibility.
Trauma and the Persistence of Early Memories
While ordinary early memories may fade from conscious access, traumatic experiences appear to follow different rules.
Overwhelming emotional experiences can create memories that persist despite occurring before the typical age of earliest memories.
This phenomenon, sometimes called “infant PTSD,” challenges conventional understanding of memory development.
Infants who experience medical procedures, abuse, or other traumatic events may not form explicit narrative memories of these experiences, but their brains and bodies remember nonetheless.
These memories often manifest as somatic (body-based) responses rather than conscious recollections.
A child who experienced painful medical procedures as an infant may develop inexplicable fears of white coats or alcohol swabs without consciously knowing why. The body remembers what the conscious mind cannot access.
The neural mechanism behind this phenomenon appears to involve the amygdala’s ability to form strong emotional memories independently of the hippocampus.
When an experience triggers an intense fear response, the amygdala creates powerful associative memories that can persist even when hippocampal-dependent contextual details are lost.
This has significant implications for understanding and treating early childhood trauma. Traditional talk therapy, which relies on accessing narrative memories, may be ineffective for addressing preverbal traumatic experiences.
Instead, approaches that engage the body and work directly with physical sensations may better access these early imprints.
The Language Bridge: How Words Transform Memory
The acquisition of language represents a fundamental shift in how humans encode and retrieve memories.
As children develop linguistic abilities between ages 1-3, they gain not just a communication tool but an entirely new way of organizing experience.
This transition creates what researchers call the “language bridge”—a critical developmental period when preverbal experiences are either successfully translated into linguistic form or become increasingly difficult to access through conscious retrieval.
The process appears to work in both directions. When parents provide language for their infants’ experiences in real-time (“You’re feeling frustrated that you can’t reach that toy”), they’re helping create verbal tags for emotional and sensory states.
Later, when children begin acquiring language themselves, they may use these verbal labels to access previously preverbal memories.
Conversely, experiences that never receive verbal labels may remain locked in preverbal format, accessible only through indirect means such as emotional responses or bodily sensations.
This may explain why directed reminiscing—the practice of regularly discussing past events with young children—has been shown to strengthen early memory access.
The language bridge concept suggests that our earliest memories aren’t necessarily lost but may exist in a format incompatible with our later retrieval systems—like trying to open a digital file without the correct software. Those experiences that successfully “cross the bridge” by acquiring verbal labels during the critical window become part of our accessible autobiographical narrative.
Implications for Early Education and Childcare
The emerging understanding of infant memory has profound implications for early childhood education and childcare practices.
If infants are forming lasting memories from their earliest days—memories that may influence them at subconscious levels throughout life—the quality of early environments takes on new significance.
Traditional approaches to infant care have sometimes prioritized physical needs over cognitive and emotional engagement, assuming that babies won’t remember specific interactions anyway. The new research suggests this assumption is deeply flawed.
Instead, the evidence points toward the importance of creating rich, responsive environments that engage multiple sensory systems simultaneously. Infants appear to learn best through multisensory experiences that combine touch, movement, vision, and sound—precisely the kind of integrated sensory experiences that their memory systems are optimized to encode.
This doesn’t mean bombarding infants with stimulation or turning childcare into a series of instructional activities. Rather, it suggests the importance of attuned caregiving that responds to infants’ natural interests and provides narration that helps connect sensory experiences to emerging language.
Perhaps most importantly, it suggests that the emotional quality of early care may be inscribed in memory systems even when specific events are forgotten.
The sense of security and emotional regulation that comes from responsive caregiving appears to create enduring patterns in both explicit and implicit memory systems.
The Future of Early Memory Research
As technology advances, our ability to study the mysteries of infant memory continues to improve.
Emerging techniques like magnetoencephalography (MEG) offer the possibility of tracking memory formation in real-time with millisecond precision, while advances in portable neuroimaging make it possible to study infants in more natural settings.
Meanwhile, artificial intelligence approaches to analyzing brain activity patterns are beginning to decode the neural signatures of specific memories.
These machine learning systems can already identify which image an adult is viewing based solely on brain activation patterns.
Applied to infant research, similar techniques might eventually allow us to detect the presence of specific memories even when babies can’t tell us what they remember.
The most tantalizing frontier may be the possibility of accessing our own early memories. While current technology can’t retrieve specific content from inaccessible memories, therapeutic approaches that combine sensory activation with advanced neuroimaging feedback might someday help bridge the gap between our conscious awareness and our earliest experiences.
Such abilities would raise profound questions about the nature of identity and consciousness. If we could access memories from our first year of life, how might that change our understanding of ourselves?
Would recovering these memories feel like remembering something we’ve forgotten, or more like meeting a different version of ourselves—a preverbal being whose experiences are organized in fundamentally different ways from our adult consciousness?
These questions remain speculative, but the accelerating pace of neuroscience research suggests they may not remain so for long.
As we continue to unravel the mysteries of infant memory, we may discover that our earliest experiences remain more accessible than we ever imagined—not lost to time, but simply waiting in different formats for the right keys to unlock them.
