The organizing ideaTwo kinds of plasticity, not one
Infancy compresses more physical and neural change into a shorter span than any later chapter of the lifespan, and the temptation is to treat it as a checklist of milestones and growth charts. The more useful move is to find the single idea that organizes the whole unit, and here it is a distinction drawn from the neuroscience of brain development: the difference between the connections the brain builds because it expects a certain kind of experience and the connections it builds because a particular experience actually happened (Greenough et al., 1987). Synapses bloom in vast excess and are then pruned back, and this distinction explains why the brain is built to work that way.
The overproduce-then-prune pattern looks wasteful until you see its logic. Rather than specifying every connection in the genome — an impossible amount of information — development lays down far more synapses than will survive and lets experience finish the wiring (Huttenlocher, 1979). The environment does not merely fill a pre-built brain with content; it selects, from an overabundant starting network, which circuits are kept and strengthened. That is why the same early experiences that seem unremarkable to adults — being spoken to, seeing faces in ordinary light, being handled and moved — are the literal raw material of cortical architecture, and why their absence in a sensitive window can leave a permanent mark.
Experience-expectant plasticity: the brain overproduces synapses and then waits for the ordinary experiences every human infant reliably gets — patterned light, voices, faces, movement. Those expected inputs select which synapses survive; the rest get pruned. This is why a baby deprived of visual input in a sensitive window (e.g., untreated cataracts) suffers permanent deficits — the expected experience never arrived to shape the circuit. Experience-dependent plasticity is different: it's the lifelong ability to grow new connections in response to idiosyncratic experiences unique to you — learning violin, a second language, the layout of your city. One is a use-it-or-lose-it window; the other is open for life. Hold this pair and pruning stops looking like damage.
Intuitively, a denser brain sounds like a better brain, and pruning sounds like something going wrong. Both are backwards. Peak synaptic density occurs in early childhood, not adulthood — a two-year-old has roughly twice the synapses of an adult. Pruning is optimization, not attrition: the brain sculpts an efficient network by eliminating the connections experience didn't use, exactly as a sculptor removes marble. Conditions marked by too little pruning are associated with atypical development, not genius. The headline: the developing brain gets smarter by getting sparser.
GrowthBody and brain on their steepest climb
The infant body grows faster in the first two years than at any point after birth. Birth weight roughly doubles by five months and triples by the first birthday, while length increases by about half, and the changes are not uniform across the body. Growth follows two organizing gradients that also govern motor control: the cephalocaudal trend runs head-downward, so the head and trunk mature ahead of the legs and feet, and the proximodistal trend runs center-outward, so control of the shoulders and hips precedes control of the hands and fingers. A newborn is therefore top-heavy by design, its comparatively enormous head reflecting how far ahead the nervous system runs of the rest of the body.
Nowhere is that head start more literal than in the brain itself. The brain is nearer to its adult size at birth than any other organ and reaches roughly 80 to 90 percent of adult volume by age two, driven less by new neurons — most of which are generated prenatally — than by the proliferation of synapses, the growth of dendrites, and the myelination of axons (Gilmore et al., 2018). Synapse formation is explosive in the first postnatal months and peaks, depending on the cortical region, somewhere between the first and third years, at densities well above adult levels (Huttenlocher, 1979). Myelination, the wrapping of axons in fatty sheaths that speeds neural transmission, begins prenatally and continues for decades, but its fastest phase coincides with the first two years, which is part of why infant behavior becomes so rapidly more coordinated.
Crucially, this construction is not a self-contained maturational program that experience merely decorates. The overproduced synapses of early infancy are precisely what make the young brain so sensitive to input, and the timetable of building and pruning differs by region — sensory and motor areas mature earlier, association areas governing planning and self-control far later (Gilmore et al., 2018). The result is a brain that is at once remarkably plastic and remarkably vulnerable during the first years: plastic because so much wiring is still up for negotiation, vulnerable because the circuits laid down now become the scaffold on which everything later is built.
A public-health winHow researchers cut SIDS in half
Some of developmental science's clearest victories are not new theories but changed behaviors, and the story of sudden infant death is the field's best example. Sudden infant death syndrome (SIDS) is the sudden, unexplained death of an infant under one year that remains unexplained after a full investigation; it sits within the broader category of sudden unexpected infant death (SUID), which also includes accidental suffocation and entrapment. For decades SIDS was the leading cause of death between one month and one year in many wealthy countries, and its apparent randomness made it terrifying to parents and baffling to clinicians.
The most useful framework for understanding it is the triple-risk model: a death tends to occur when a vulnerable infant — for instance one with subtle abnormalities in the brainstem circuits that regulate breathing and arousal — passes through a critical developmental period and encounters an external stressor such as prone sleeping or overheating (American Academy of Pediatrics Task Force on Sudden Infant Death Syndrome, 2016). The power of the model is practical: caregivers cannot change an infant's underlying vulnerability, but they can remove the external stressors, and doing so is exactly what the public-health campaigns targeted.
After epidemiology linked prone (stomach) sleeping to sudden infant death, the American Academy of Pediatrics launched the "Back to Sleep" campaign in 1994 (rebranded "Safe to Sleep" in 2012). Putting babies to sleep on their backs drove SIDS rates down by roughly half in the following years. Current guidance goes further: alone, on the back, on a firm flat surface, in the parents' room but not the parents' bed (room-sharing, not bed-sharing), no soft bedding, bumpers, or pillows, and offering a pacifier at sleep. A cheap behavioral change, informed by developmental research, saved tens of thousands of infants — the kind of result that justifies the whole field.
SleepWhy infants sleep so differently — and why it matters
Newborn sleep is not simply more abundant than adult sleep; it is structurally different. Infants sleep around sixteen hours a day, but in short bouts distributed around the clock, and a strikingly large share of that time is spent in active (REM) sleep, the stage associated in adults with dreaming. Where adults spend roughly a fifth of the night in REM, newborns spend about half of their total sleep in it, a proportion that declines steadily across the first years (Roffwarg et al., 1966). The classic interpretation is the autostimulation hypothesis: REM's intense, self-generated neural activity may supply stimulation that helps build and refine the visual and other sensory systems at a time when waking experience is limited — an internal source of the very kind of patterned input that experience-expectant circuits are waiting for.
Over the first year, sleep also consolidates. The scattered short bouts of the newborn gradually reorganize into longer nighttime stretches as circadian rhythms mature and the day-night cycle takes hold, which is why "sleeping through the night" is a developmental achievement rather than a matter of parental technique alone. Sleep matters here for more than parental sanity: the same regulatory systems that govern arousal from sleep are implicated in the vulnerability that the triple-risk model places at the center of sudden infant death, which is why the safe-sleep recommendations are fundamentally recommendations about the sleep environment (American Academy of Pediatrics Task Force on Sudden Infant Death Syndrome, 2016).
How developed is each sense at birth? Tap one to see where newborns start.
The sensesBuilding a perceptual world from the first day
The infant does not wait passively for the senses to switch on; perception is active and, in several channels, remarkably capable from birth. Vision is the exception that proves the rule. Newborn visual acuity is poor — on the order of 20/400 — and contrast sensitivity and color vision take months to sharpen, yet even newborns are not indifferent to what they see. Presented with paired patterns, they look longest at complex, high-contrast, curved, face-like arrangements rather than blank fields, an early bias documented in the preferential-looking studies that founded modern infant-perception research (Fantz, 1961). That looking preference is also the field's central method: because infants cannot report what they perceive, researchers infer discrimination from where and how long they look, and from whether attention recovers when a stimulus changes.
Depth perception has its own iconic demonstration. On the visual cliff — a glass-covered table with a shallow side and an apparent drop-off — crawling infants of about six months will cross the shallow side but hesitate at the deep side, evidence that they perceive depth and connect it to the danger of falling (Gibson & Walk, 1960). Whether the perception of depth is innate or the fear of it is learned became clearer when heart-rate measures showed that even pre-crawling infants discriminate the two sides — their hearts decelerate in interest over the deep side rather than accelerating in fear — which suggests that depth is perceived early but that wariness of heights develops with locomotor experience (Campos et al., 1970).
Hearing, by contrast, is functional well before birth, and the newborn arrives already tuned to the human voice. Given control over a recording by sucking on a special nipple, newborns will adjust their sucking to hear their own mother's voice over a stranger's, a preference that must have been established prenatally through the muffled sound conducted to the womb (DeCasper & Fifer, 1980). Speech perception then undergoes a striking narrowing across the first year. Young infants discriminate the phonetic contrasts of every language, including ones their caregivers never use, but by roughly ten to twelve months they lose sensitivity to non-native contrasts as their perception commits to the sounds of the language around them (Werker & Tees, 1984). This perceptual reorganization is a textbook case of experience-expectant plasticity in action, and the statistical learning that drives it — infants tracking the distributional patterns of the speech they hear — has become a model system for how experience sculpts the brain (Kuhl, 2004).
Taste and smell are keen from the start and, like hearing, begin their tuning before birth. Newborns prefer sweet tastes and orient toward the odor of their own mother's milk, and flavors from the maternal diet cross into amniotic fluid and breast milk, so that infants come to accept foods they were "introduced" to prenatally more readily at weaning (Mennella et al., 2001). Perhaps most tellingly, perception is integrated across the senses from very early on rather than assembled channel by channel. Month-old infants who have sucked on a particular shape of pacifier without seeing it will, when shown two shapes, look longer at the one they felt — a cross-modal match between touch and vision that implies a unified perceptual world nearly from the outset (Meltzoff & Borton, 1979).
ReflexesThe newborn's built-in starter software
Before infants control much of anything voluntarily, they arrive equipped with reflexes — automatic, stereotyped responses to specific stimuli. Some are plainly adaptive survival reflexes: the rooting reflex turns the head toward a cheek that is touched and the sucking reflex organizes feeding, together ensuring the newborn can locate and extract milk. Others, often called primitive reflexes, have less obvious present-day function and are thought to be evolutionary remnants — the Moro (startle) reflex flings the arms outward and back at a sudden loss of support, the palmar grasp closes the fingers around anything pressing the palm, and the Babinski reflex fans the toes when the sole is stroked.
Reflexes earn their place in an assessment because their timing is diagnostic. Most primitive reflexes appear on a predictable schedule and then disappear over the first year as voluntary cortical control comes online and inhibits them; a reflex that is absent when it should be present, or that persists long past its usual window, can signal a neurological problem. But the reflexes are not merely things to be outgrown. The stepping reflex — a newborn held upright over a surface makes coordinated stepping movements — was long explained as a reflex that "disappears" through cortical inhibition, until a simpler, embodied account showed that infants stop stepping mainly because their legs grow heavier faster than they grow stronger; submerge the same "non-stepping" babies in water to unweight the legs and the movements return (Thelen & Fisher, 1982). That reinterpretation is a preview of the dynamic-systems view of motor development that the milestones only appear to obey.
Milestones aren't a clockCulture speeds up (and reshapes) motor development
Charts that announce "sits at six months, walks at twelve" imply a fixed maturational timetable ticking off pre-programmed steps. The cross-cultural and observational data tell a more interesting story: motor milestones are not switches thrown by a genetic clock but solutions the whole system assembles from the infant's changing body, the task at hand, and the surrounding environment — the core claim of the dynamic-systems approach to development (Thelen, 1995).
Consider how much of walking is practice rather than maturation. Toddlers learning to walk are astonishingly industrious, averaging on the order of 2,000 steps and dozens of falls per hour of free play and accumulating the equivalent of the length of several football fields daily; walking improves not on a schedule but with the sheer volume of variable, self-directed experience (Adolph et al., 2012). Where caregivers actively exercise infants — daily massage, joint stretching, propping them to sit, holding them to "step" — milestones arrive weeks or months ahead of Western norms, as documented in the classic studies of infants in some East African communities (Super, 1976). Conversely, where infants are carried much of the day and set down on their backs, stages the charts treat as universal, such as crawling, may be delayed or skipped entirely, yet walking still arrives within the normal range.
The embodied, enculturated character of motor development is the through-line of the modern synthesis: milestones are shaped by body proportions, everyday caregiving routines, and cultural practices, so the "normal" ranges printed in pediatric charts are best read as local norms rather than biological constants (Adolph & Hoch, 2019). The sequence — control spreading head to toe and center to periphery — is far more reliable than the calendar dates attached to it.
Infants in cultures that actively exercise them — daily massage, stretching, propping them to sit, holding them upright to "step" — reach motor milestones weeks to months earlier than Western norms. Meanwhile some cultures skip stages the charts treat as universal: where babies are carried constantly and set down on their backs, crawling can be rare or absent, yet these infants walk on time. Milestones are solutions a whole system finds — body, goal, and environment together (Thelen's dynamic systems) — not steps unlocked by a genetic timer. The "normal" ranges in the charts are local norms dressed up as biology.
The usual gross-motor sequence — approximate, and culturally elastic
Read these ages as a rough route, not a train timetable. A wide window is normal; the sequence is more reliable than the dates.
NutritionFeeding the fastest growth of the lifespan
All of this building — tripling body weight, quadrupling brain volume, myelinating tracts — runs on nutrition, and infancy is the one stretch of life when a few months of inadequate intake can leave lasting marks. Human milk is uniquely matched to the task, and the evidence for its benefits is now large and consistent: pooled across dozens of studies, breastfeeding is associated with fewer gastrointestinal and respiratory infections, reduced infant mortality, lower later rates of overweight and type 2 diabetes, and modest gains in measured intelligence (Victora et al., 2016). Major health bodies accordingly recommend exclusive breastfeeding for roughly the first six months, with continued breastfeeding alongside complementary foods thereafter, while recognizing that formula is a safe and adequate alternative where breastfeeding is not possible.
Feeding is also one of the infant's first curricula in what to like. Because flavors from the maternal diet reach amniotic fluid and breast milk, infants arrive already biased toward the foods of their culture and accept them more readily at weaning — a reminder that even nutrition is partly a matter of experience-driven learning rather than pure physiology (Mennella et al., 2001). The stakes of getting early nutrition right are highest precisely because the brain's steepest growth and the body's fastest gains coincide: undernutrition during this window is associated with stunting and cognitive deficits that are difficult to fully reverse later, which is why infant feeding is treated as a global-health priority and not merely a private parenting choice.
The social brainInteraction is not a bonus — it's the input
Sensory capacities get most of the attention, but the richest input the developing brain expects is another responsive human. The reciprocal, contingent exchanges between infant and caregiver are not enrichment layered on top of "real" development; they are among its primary engines — the proximal processes that a bioecological account places at the center of what actually drives development (Bronfenbrenner & Morris, 2006). An infant's brain, in other words, expects to be met, answered, and repaired, and it builds its circuits for attention, language, and emotion regulation out of those loops.
The clearest evidence that infants are active partners rather than passive recipients comes from what happens when the responsiveness stops. Infants do not merely react to a caregiver's emotional signals; they help co-regulate the interaction, working to restore it when it breaks down and drawing on the exchange to manage their own arousal (Tronick, 1989). Two ideas capture the machinery.
Serve and return
Development is built through thousands of tiny loops: the baby "serves" (a coo, a reach, a gaze), the caregiver "returns" (mirrors it, names it, responds). These reciprocal exchanges are the proximal processes Bronfenbrenner called the real engines of development — and they wire the circuits for attention, language, and emotion regulation. It's why a talked-to, responded-to baby thrives on more than nutrition.
The still-face paradigm (Tronick et al., 1978)
Have a mother go suddenly blank-faced and unresponsive for two minutes. The infant first tries harder to re-engage — smiling, pointing, reaching — then, when the "return" doesn't come, grows distressed and withdraws. It's a striking demonstration that even young infants expect contingent responsiveness and are actively regulating the interaction, not passively receiving it. Repair comes fast once the parent re-engages — which is the reassuring part.
A modern consideration: the American Academy of Pediatrics recommends avoiding screen media for children younger than about 18–24 months, apart from live video chat (American Academy of Pediatrics Council on Communications and Media, 2016). The reason connects straight back to serve-and-return — infants learn language and social contingency from responsive humans, and before roughly age two they transfer strikingly little from a screen to real life, a robust finding known as the "video deficit" (Anderson & Pempek, 2005). The concern is less the screen itself than the responsive interaction it displaces.
SynthesisOne organism, many timetables
Step back and the pieces of this unit resolve into a single argument. The infant is not a passive vessel being filled in on a fixed maturational schedule but an active organism whose body, brain, and behavior are assembled through a constant conversation with the environment. The brain overbuilds and then prunes because it is designed to let experience finish the wiring (Greenough et al., 1987; Huttenlocher, 1979); the senses come online early and actively, narrowing to the particular language and flavors and faces the infant actually encounters (Werker & Tees, 1984; Kuhl, 2004); motor milestones emerge as negotiated solutions rather than unlocked stages (Thelen, 1995; Adolph & Hoch, 2019); and the social exchanges that look like mere affection are the very input the brain expects (Bronfenbrenner & Morris, 2006; Tronick, 1989).
That framing also explains why the field's proudest achievement is behavioral rather than theoretical. Recognizing that a modifiable external stressor sat at the meeting point of vulnerability and a critical period made it possible to halve sudden infant death by changing one thing — how babies are laid down to sleep (American Academy of Pediatrics Task Force on Sudden Infant Death Syndrome, 2016). Get the ordinary, expected inputs right — responsive interaction, adequate nutrition, a safe place to sleep, room to move — and the extraordinary machinery of infant development largely takes care of the rest.
SourcesCited in APA 7
Adolph, K. E., Cole, W. G., Komati, M., Garciaguirre, J. S., Badaly, D., Lingeman, J. M., Chan, G. L. Y., & Sotsky, R. B. (2012). How do you learn to walk? Thousands of steps and dozens of falls per day. Psychological Science, 23(11), 1387–1394.
Adolph, K. E., & Hoch, J. E. (2019). Motor development: Embodied, embedded, enculturated, and enabling. Annual Review of Psychology, 70, 141–164.
American Academy of Pediatrics Council on Communications and Media. (2016). Media and young minds. Pediatrics, 138(5), e20162591.
American Academy of Pediatrics Task Force on Sudden Infant Death Syndrome. (2016). SIDS and other sleep-related infant deaths: Updated 2016 recommendations for a safe infant sleeping environment. Pediatrics, 138(5), e20162938.
Anderson, D. R., & Pempek, T. A. (2005). Television and very young children. American Behavioral Scientist, 48(5), 505–522.
Bronfenbrenner, U., & Morris, P. A. (2006). The bioecological model of human development. In R. M. Lerner (Ed.), Handbook of child psychology: Vol. 1. Theoretical models of human development (6th ed., pp. 793–828). Wiley.
Campos, J. J., Langer, A., & Krowitz, A. (1970). Cardiac responses on the visual cliff in prelocomotor human infants. Science, 170(3954), 196–197.
DeCasper, A. J., & Fifer, W. P. (1980). Of human bonding: Newborns prefer their mothers' voices. Science, 208(4448), 1174–1176.
Fantz, R. L. (1961). The origin of form perception. Scientific American, 204(5), 66–72.
Gibson, E. J., & Walk, R. D. (1960). The "visual cliff." Scientific American, 202(4), 64–71.
Gilmore, J. H., Knickmeyer, R. C., & Gao, W. (2018). Imaging structural and functional brain development in early childhood. Nature Reviews Neuroscience, 19(3), 123–137.
Greenough, W. T., Black, J. E., & Wallace, C. S. (1987). Experience and brain development. Child Development, 58(3), 539–559.
Huttenlocher, P. R. (1979). Synaptic density in human frontal cortex — Developmental changes and effects of aging. Brain Research, 163(2), 195–205.
Kuhl, P. K. (2004). Early language acquisition: Cracking the speech code. Nature Reviews Neuroscience, 5(11), 831–843.
Meltzoff, A. N., & Borton, R. W. (1979). Intermodal matching by human neonates. Nature, 282(5737), 403–404.
Mennella, J. A., Jagnow, C. P., & Beauchamp, G. K. (2001). Prenatal and postnatal flavor learning by human infants. Pediatrics, 107(6), e88.
Roffwarg, H. P., Muzio, J. N., & Dement, W. C. (1966). Ontogenetic development of the human sleep-dream cycle. Science, 152(3722), 604–619.
Super, C. M. (1976). Environmental effects on motor development: The case of "African infant precocity." Developmental Medicine & Child Neurology, 18(5), 561–567.
Thelen, E. (1995). Motor development: A new synthesis. American Psychologist, 50(2), 79–95.
Thelen, E., & Fisher, D. M. (1982). Newborn stepping: An explanation for a "disappearing" reflex. Developmental Psychology, 18(5), 760–775.
Tronick, E. Z. (1989). Emotions and emotional communication in infants. American Psychologist, 44(2), 112–119.
Tronick, E., Als, H., Adamson, L., Wise, S., & Brazelton, T. B. (1978). The infant's response to entrapment between contradictory messages in face-to-face interaction. Journal of the American Academy of Child Psychiatry, 17(1), 1–13.
Victora, C. G., Bahl, R., Barros, A. J. D., França, G. V. A., Horton, S., Krasevec, J., Murch, S., Sankar, M. J., Walker, N., & Rollins, N. C. (2016). Breastfeeding in the 21st century: Epidemiology, mechanisms, and lifelong effect. The Lancet, 387(10017), 475–490.
Werker, J. F., & Tees, R. C. (1984). Cross-language speech perception: Evidence for perceptual reorganization during the first year of life. Infant Behavior and Development, 7(1), 49–63.