Look at your living room floor right now.
Toys everywhere. A wooden block abandoned mid-stack. A spoon that was briefly a drumstick. A sock, pulled off and examined like an archaeological artifact, then dropped. Your baby moves through objects like a hummingbird through flowers — landing, tasting, turning, dropping, moving on. Nothing holds their attention for more than a few seconds.
You might worry about this. You might wonder if something is wrong with their focus, their concentration, their ability to stick with one thing. You might try to redirect them back to a toy they abandoned. You might buy toys designed to hold attention longer.
Don't. What you're watching is not distraction. It's a curriculum.
The 9.8-Second Curriculum
Herzberg, Adolph, Tamis-LeMonda, and colleagues published a landmark study in Child Development in 2022 that changed how we understand infant play. They brought cameras into the homes of 40 infants between 13 and 23 months old and recorded everything — frame by frame.
The numbers are staggering. Across the study, they documented 10,015 separate object play bouts. The median duration of each bout: 9.8 seconds.
Nine point eight seconds. Then on to the next thing.
But here's what the researchers found when they looked at the pattern: object play consumed the majority of waking time. Infants cycled through dozens of different objects — toys, household items, things they found on the floor — in short, varied bursts distributed across the entire day. They didn't play with one thing for a long time. They played with everything for a little time, over and over.
Herzberg's team recognized this for what it is: distributed practice. The same learning strategy that decades of cognitive science has identified as optimal for skill acquisition in every domain — from language learning to athletic training to musical performance. Space out your practice. Vary your targets. Short sessions, many repetitions, distributed across time.
Your baby didn't read the research. They're doing it anyway.
Not How Much — How Well
But distributed practice alone isn't the whole story. If it were, every baby who scattered toys across the room would be on the same developmental trajectory. They're not.
Muentener, Herrig, and Schulz (2018, Frontiers in Psychology, N=130) tracked infants from about 12 months across nine months, then followed up again at age three. They measured something they called exploratory efficiency — essentially, how many different parts of an object an infant contacted relative to their total play time.
The results were striking. Exploratory efficiency was the only stable measure across infancy — it didn't fluctuate with mood or context. And it predicted outcomes years later. At age three, efficiency of exploration at 12 months correlated with IQ (r = 0.37). It distinguished at-risk infants from typically developing ones. Just five minutes of free play observation was enough to detect the signal.
This is not about how long they play. It's about how they explore within the chaos.
A cautionary complement comes from Liu and colleagues (2025, Infancy, N=20 preterm infants born at ≤29 weeks). In their sample, the frequency of object interaction bouts at 12 months was negatively associated with cognition at 30 months (B = −6.20, p = .027). The preterm infants' median bout duration was 5.67 seconds — shorter and faster than Herzberg's 9.8 seconds in typical infants. More rapid switching reflected less sustained engagement, not more distributed practice.
There's a Goldilocks zone. Distributed practice is powerful. But too-fragmented play — touching everything, engaging with nothing — is different. The signal is efficiency within the chaos, not chaos itself.
The Curious Amplifier
If distributed practice is the strategy and efficiency is the quality, curiosity is the engine.
De Boer and colleagues published a remarkable study in Developmental Science in 2026 (N=60) that tracked curiosity from infancy to early childhood. They measured curiosity at eight months using a Bayesian model of looking time — essentially, how sensitive was each infant to information gain? Did they look longer at things that were maximally informative?
Then they measured IQ at three and a half years.
The relationship wasn't linear. In the top third of most curious infants, curiosity at eight months predicted 34 percent of the variance in IQ at three and a half years (F = 10.26, p = .005). In the medium and low curiosity groups: no effect at all.
Curiosity acts as a cognitive amplifier. It's not just that curious babies explore more. It's that highly curious babies are driven by information gain — they seek out what they don't know, and that seeking compounds over years of development. The baby who picks up every object on the floor isn't randomly sampling. They're driven by what each object might teach them.
Hands Build Language
Here is where the science gets genuinely surprising.
Contino and Nelson (2025, Infant Behavior and Development, N=90) tracked what they call Manipulation Complexity — a score from 1 to 8 based on whether infants used one hand or two, synchronous or asynchronous movements, fingers or full grasp, single or multiple objects. They followed trajectories from 9 to 14 months.
Infants in the high manipulation complexity class had significantly greater language skills at two years. The hands were predicting the mouth.
The mechanism is beautifully concrete. West and Iverson (2017) showed that parents are more likely to provide verbal labels when infants are actively manipulating objects — not just holding them, but turning them, combining them, exploring them with complex hand movements. Contino's companion paper confirmed it: caregiver object labeling increases specifically during complex manipulation. The infant's hands invite language from the caregiver.
And then Suarez-Rivera, Linn, and Tamis-LeMonda (2022, Language Learning) completed the cascade. They coded frame-by-frame from two-hour home videos of 32 infant-caregiver dyads at 18–23 months. Objects that infants manipulated AND that mothers named were more likely to appear in infants' vocabularies and spontaneous speech. A separate analysis of 5,520 infants from the Wordbank database confirmed: words for objects that are typically manipulated AND named are acquired earlier across the population.
This is the full developmental cascade: complex hand manipulation → caregiver naming → word learning. The mess on the floor isn't just motor practice. It's language class.
The Symbolic Leap
Watch your 12-month-old with a toy phone. They might hold it to their ear. They might press the buttons. They're using the object the way it's supposed to be used — what researchers call functional play.
This is the precursor to something enormous.
Over the next several months, functional play will evolve into pretend play. The phone will become a banana held to the ear. The block will become a car. The spoon will become a microphone. Your baby will begin substituting one object for another — and this substitution requires the same cognitive architecture as language itself.
A meta-analysis spanning 35 studies and 6,848 children found a robust correlation between symbolic play and language ability (r = 0.35). Lyytinen and colleagues (1999, N=171) showed that symbolic play at 14 months uniquely predicted language and cognitive skills at 24 months. Creaghe and colleagues (2021) demonstrated that symbolic play contexts elicit more conversational turns at 18 months, which in turn predict vocabulary and grammar at 24 months.
When your baby pretends — when they hold a block to their ear and say "hewwo" — they're not just playing. They're demonstrating that they understand one thing can stand for another. That is the foundation of all symbolic thought. Words, numbers, letters, metaphors — all of them rest on this insight, and it begins here, on the living room floor, with a banana and an imaginary phone call.
The Environment as Co-Teacher
The space your baby plays in isn't neutral. It teaches.
Van Liempd, Oudgenoeg-Paz, and Leseman (2025, Child Development, N=61, ages 11–48 months) studied how the physical and social environment shapes object exploration. They found that solitary and parallel play were positively associated with complex exploration — especially at child-height surfaces. Floor space, activity centers, and child-height tables were the three primary spatial components that supported exploration. Low visual boundaries and recognizable play areas helped infants organize their own play.
Wijeakumar and colleagues (2025, Infant Behavior and Development, N=91, ages 6–10 months) added a critical detail: the type of objects matters. Multi-component objects — stacking boxes, sorting towers, nesting cups — elicited more scaffolding from caregivers and more engagement from infants compared to single objects. And caregiver scaffolding during multi-component object play specifically predicted infant visual working memory. Single-object play showed no such effect.
The implication is practical: multi-component objects create a feedback loop. They're harder, so the baby engages more intently. The caregiver sees the engagement and offers more help. The help scaffolds the baby's working memory. The improved working memory enables more complex play. The cycle repeats.
What This Means for Your Floor
You don't need to engineer your baby's play. But you can support the curriculum they're already running:
Don't over-organize. A perfectly sorted toy shelf looks good in photographs. But infants learn through foraging — encountering objects in varied, unpredictable contexts. Some mess is pedagogically optimal.
Vary the objects. Rotate toys. Include household items that are safe to explore — wooden spoons, measuring cups, fabric scraps, cardboard tubes. Multi-component objects (stacking rings, nesting cups, shape sorters) invite more complex manipulation and more caregiver interaction than single-purpose toys.
Join the floor. When you're down at their level and they're manipulating objects, you naturally narrate what they're doing. That narration — "You're stacking the red one on top!" — is the caregiver labeling that bridges hands to language. You don't need a script. Just be present and describe what you see.
Follow their lead. When they abandon a toy after eight seconds, don't redirect them back. They got what they needed. The switch is the strategy, not a deficit. Trust the curriculum.
Watch for functional play. When your baby holds a toy phone to their ear or pushes a toy car along the floor, they're showing you the first flickers of symbolic thought. Respond to it — pick up another "phone" and answer. You're not just playing along. You're scaffolding the cognitive leap that will underpin language, imagination, and abstract thought for years to come.
The Lesson Plan on the Floor
I keep returning to Herzberg's number: 9.8 seconds. It sounds like nothing. It sounds like the attention span of a goldfish. It sounds, to anxious parents in an age of screens and distraction, like something to worry about.
But 9.8 seconds, repeated across 10,015 bouts, distributed across waking hours, with dozens of different objects, is not a short attention span. It is a learning strategy so effective that cognitive scientists have spent decades proving it works for adults. Your baby arrived knowing.
The chaos on your living room floor — the scattered blocks, the abandoned spoons, the sock examined and discarded — is not entropy. It is a curriculum. Each object touched is a data point. Each switch is a spacing effect. Each complex manipulation is an invitation for language. Each moment of curiosity-driven exploration is an amplifier for cognition.
Your baby is not distracted. They are studying. And the classroom is everywhere they look.