RESEARCH DOSSIER
Neural Development and Sensitive-Period Plasticity
Abstract. Neural development establishes the cellular and circuit conditions through which experience modifies perception, learning, and behavior. This research monograph examines sensitive and critical periods as temporally structured changes in the effect of experience on developing neural systems. It synthesizes foundational deprivation experiments, cellular evidence on inhibition and plasticity stabilization, longitudinal human anatomical studies, human visual and language research, and contemporary proposals concerning hierarchical developmental windows. The methods include historical primary-study analysis, critical comparison of animal interventions with human observational and quasi-experimental designs, and appraisal of construct validity, causal inference, and generalizability. The central synthesis identifies multiple overlapping developmental schedules rather than a universal age threshold. Mechanistic evidence is strongest for specific animal sensory circuits; human associations between age, deprivation, and subsequent performance require domain-specific interpretation. The monograph evaluates recovery in later life, describes limits of clinical translation, and formulates testable priorities for longitudinal multimodal research.
Keywords: neural development; sensitive periods; critical periods; developmental neuroplasticity; experience-dependent maturation; ocular dominance; synaptic pruning; parvalbumin interneurons; GABA; Otx2; perineuronal nets; language acquisition; visual deprivation; cortical development; recovery
The Research Object: Developmental Organization of Neural Plasticity
Neural development is the temporally organized generation, differentiation, migration, connection, functional specialization, and stabilization of nervous-system cells and circuits. Sensitive-period plasticity designates intervals when defined features of experience have an especially strong influence on the development of a neural system and the behaviors it supports. The relevant scientific unit is therefore a relation among a circuit, a developmental state, an input, an intervention or history of experience, and an outcome measured after an appropriate interval. Neural changes can be structural, physiological, or representational, while their behavioral consequences may be immediate, delayed, reversible, or persistent. Knudsen (2004) provides a foundational framework for connecting developmental sensitivity in neural circuitry to observed changes in behavior.
The questions governing this monograph concern why a particular interval of heightened modifiability appears, what initiates and ends it, how different circuits follow different calendars, and what follows when expected input is absent or altered. The inquiry also investigates how much can be inferred about molecular mechanisms from behavioral outcomes in human children and about human psychological development from controlled experiments in cats, mice, and rats. The evidence encompasses visual receptive-field physiology, inhibitory circuit manipulation, extracellular matrix changes, microglial refinement, postmortem human tissue, longitudinal brain imaging, infant speech perception, and developmental interventions. These methods reveal different features of one broad developmental problem rather than interchangeable indicators of the same latent property Hensch (2005) Sydnor et al. (2026).
The synthesis follows original experimental studies first, using major reviews to map competing explanations and historical revisions. Its selection is a structured, purposive critical review: it does not claim exhaustive screening, a preregistered protocol, pooled effects, or the coverage of a systematic review. Sources are assessed by the match between the operational manipulation and the claimed outcome, by control conditions, and by the distance between measured phenomena and proposed psychological mechanisms. The monograph treats sensory critical-period biology as a well-developed experimental case, and other systems as partly converging but more heterogeneous research programs. An emphasis on explicit inference prevents the broad word plasticity from masking substantial differences among cell growth, learning, circuit reweighting, recovery, and developmental vulnerability Takesian & Hensch (2013).
Sensitive Period, Critical Period, and Plasticity: Operational Definitions
A sensitive period is defined by a time-dependent gradient in the effect of experience: a specified exposure or deprivation has a larger or different influence during one developmental interval than another. An operational test requires comparing outcomes after comparable manipulations administered at distinguishable ages, ideally while controlling exposure duration, baseline function, and maturational status. A critical period is often reserved for cases in which an experience is necessary for typical development and the missed opportunity has persistent consequences; Knudsen (2004) specifies this distinction in terms of essential information and lasting alteration. In practice the terminology varies by field, so experimental designs and outcome measures should accompany any categorical label. An estimated peak in age sensitivity and a hard biological endpoint are different hypotheses.
Plasticity itself is multidimensional. Synapses may change strength without extensive anatomical remodeling; neuronal arbors may refine while gross MRI volumes appear stable; and behavior may improve through compensatory strategies that differ from restoration of the original circuit. The same system can show high plasticity for one feature, such as ocular dominance, while retaining or acquiring other capabilities on different schedules. The human visual studies synthesized by Lewis & Maurer (2005) make this especially clear: periods during which normal input supports development, deprivation produces damage, or restoration enables recovery can have distinct timelines. Consequently, sensitive-period inference requires an outcome-specific definition of normal function and a principled account of what recovery actually recovers.
Chronological age, developmental stage, exposure history, and biological readiness should be measured independently whenever possible. An age-of-exposure effect can arise because one circuit has matured, because earlier learning changes the informational value of later input, because social opportunities change with age, or because the dose and type of stimulation differ. Werker & Hensch (2015) stresses that the speech-perception literature combines learning-based changes with candidate circuit mechanisms, but an infant's behavioral discrimination threshold is not a direct reading of inhibitory interneuron maturation. The strongest models specify how timing shifts if circuitry is delayed, prior input is absent, or the same stimulus is introduced at a later stage.
Historical Formation: Sensory Deprivation as a Causal Experiment
The decisive modern experimental tradition emerged from visual-system physiology. Recording single neurons in deprived kittens, Wiesel & Hubel (1963) and associated studies revealed dramatic changes in the cortical representation of input from an eye denied patterned vision. The procedure made specific predictions about competition between inputs that could be evaluated physiologically, overcoming the ambiguity of general claims that experience shapes the brain. It also established a demanding standard for later developmental science: the result involved a manipulated sensory pathway, a tissue-level measure, and an age-dependent functional consequence. A physiological shift in ocular dominance remains a distinct outcome from full-field visual perception in ordinary behavior.
The more explicit age-series study by Hubel & Wiesel (1970) compared monocular eye closure at different ages in kittens. The authors found a sharply bounded period of pronounced susceptibility in that species, with intense effects of brief closure in early weeks and far weaker effects after maturation. Their measurements gave quantitative meaning to a developmental window, but the boundary was a property of the tested feline circuit and deprivation procedure. Cross-species differences in eye opening, cortical maturation, ecological demands, and age scaling make direct translation to human ages inappropriate. The same paper documented limited return of cortical responsiveness after prolonged deprivation, while noting behavioral improvement under certain conditions, foreshadowing the distinction between neural recovery and functional compensation.
The historical value of these experiments lies also in their limitations. Lid closure altered form vision and input balance; it did not isolate one synaptic mechanism. Deprivation effects can reflect interocular competition, homeostatic shifts, changes in inhibitory regulation, and structural consolidation. Later studies examined how specific cellular manipulations alter the onset and offset of ocular dominance plasticity Fagiolini & Hensch (2000) Pizzorusso et al. (2002). The intellectual trajectory thus moves from the discovery of an age-dependent phenotype to experimentally discriminable causal accounts of when and why that phenotype emerges. Ethical and methodological conditions also limit the extent to which deprivation interventions can be repeated or generalized in humans.
Constructing a Nervous System: Developmental Preconditions
Sensitive periods presuppose that a circuit exists in a state capable of being reorganized by relevant input. In vertebrates this developmental sequence includes neurogenesis, migration, specification of cell identity, growth of axons and dendrites, synapse formation, glial interactions, and the emergence of patterned spontaneous and sensory-evoked activity. A circuit's receptivity to stimulation depends on its prior developmental architecture; the same external stimulus can have different effects before connections have formed, during refinement, and after consolidated communication patterns are established. These foundations clarify why sensitive periods are properties of developmental systems rather than discrete behavioral skills detached from biology Hensch (2005) Takesian & Hensch (2013).
Genetic programs and developmental signaling establish cellular populations and broad connectional constraints, while experience can select, stabilize, weaken, and rearrange among the configurations made possible by those programs. In the visual system, activity-sensitive refinement occurs at multiple levels, including thalamic and cortical circuitry. The experimental studies of complement signaling and retinogeniculate connectivity by Stevens et al. (2007) and Schafer et al. (2012) identify biologically specific routes of selective elimination. Such findings should be understood as mechanisms in studied preparations, rather than proof that the same molecular pathway governs every psychological sensitive period. A useful explanatory model identifies which features are developmentally prepared and which are instruction-sensitive within that preparation.
The schedule of developmental construction is heterogeneous across regions and cell types. Molecular maturation of inhibitory interneurons, emergence of myelinated connections, and refinement of recurrent cortical networks need not share a single temporal trajectory. Behavioral milestones emerge from interactions among several partly asynchronous subsystems, making age-specific performance change difficult to assign to one cellular event. Contemporary proposals that sensory and association cortices show different windows of environmental receptivity rest partly on this heterochrony; Sydnor et al. (2026) presents a hierarchical hypothesis for further human testing. The hypothesis becomes compelling when anatomically defined maturation predicts differential, manipulable behavioral susceptibility instead of merely tracking age-related imaging differences.
Synapse Formation, Pruning, and Regional Developmental Calendars
Synaptogenesis produces an initially flexible pattern of connections that may contain redundancies and competing inputs. Developmental refinement includes stabilization of efficacious connections, modification of synaptic weights, structural elimination, and growth of alternative pathways. Synapse number alone cannot specify learning capacity: functional potency depends on receptor properties, inhibitory control, connectivity, and the informational structure of activity. A rigorous interpretation distinguishes a net change in synapse density from turnover of individual synapses and from improvements in behavior. The classical framework of selective stabilization has been expanded by direct studies of molecular tagging and glial participation Stevens et al. (2007) Paolicelli et al. (2011).
Human postmortem research demonstrates that developmental timing differs across cortical regions. Huttenlocher & Dabholkar (1997) compared synapse density in auditory and prefrontal cortex and found different ages for peak density and later decline. The result supplies anatomical evidence for regional heterochrony, while interpretation is constrained by cross-sectional tissue availability, specimen handling, sampling of cortical layers, and the fact that tissue measures cannot directly reconstruct an individual's learning history. A high density of connections does not by itself identify the exact dates when a child is most susceptible to one auditory or executive experience. The anatomical observation is a premise for a mechanistic hypothesis that requires an independent functional test.
A separate study of human prefrontal cortical dendritic spines by Petanjek et al. (2011) found protracted remodeling extending beyond adolescence into early adulthood. This evidence challenges any simple claim that cortical pruning is finished at puberty and supports sustained developmental change in association regions. The study's histological strengths coexist with limitations of postmortem, non-longitudinal evidence, and spine counts cannot be equated directly with the onset or closure of a behavioral critical period. The broad synthesis is temporal plurality: different neural features and functions mature over different spans, and the magnitude of an experience effect must be studied directly rather than inferred from one anatomical curve.
Activity-Dependent Competition and Ocular Dominance
Ocular dominance plasticity is an experimentally tractable reorganization of the relative influence of the two eyes on neurons in binocular visual cortex. In juvenile mammals, altering patterned input from one eye shifts physiological responses and can transform the underlying competition between pathways. The classic deprivation findings Wiesel & Hubel (1963) Hubel & Wiesel (1970) establish the age-dependent phenomenon, whereas later work identifies several molecular mechanisms through which developmental conditions govern its extent. An ocular-dominance index is a valuable neural outcome because it can be assessed across neurons and conditions, yet it does not automatically quantify conscious vision, visual acuity, or all sensory functions.
Competition is influenced by the correlation and timing of activity across pathways. An active input can gain influence relative to an inactive input without a proportional increase in the overall capacity of the circuit. Conversely, changes in total activity can recruit homeostatic mechanisms that stabilize firing while altering synaptic strengths. This distinction matters for intervention design: the same change in a cortical response distribution could reflect a loss of deprived-eye influence, a gain by the other eye, or several opposing processes. Sato & Stryker (2008) shows that adult ocular-dominance plasticity has distinctive features, demonstrating why an apparently similar physiological outcome cannot be presumed to arise from the same juvenile mechanism.
The causal interpretation of ocular deprivation improves when duration, treatment age, eye-specific sensory exposure, recovery interval, and recording protocol are controlled. Individual neurons are nested in animals, and statistical analyses should respect the animal as a central independent experimental unit rather than treating every recorded cell as an unrelated replicate. Procedures that compare both eyes before and after manipulation strengthen inference about the direction of competition. Cross-species differences in visual ecology and the timing of inhibitory maturation limit the use of mouse or cat postnatal days as direct analogues of human months or years. The research value of the paradigm lies in mechanistic precision, not universal calendrical prediction Hensch (2005).
Inhibitory Maturation as a Developmental Gate
Development of inhibitory cortical circuitry is one of the best-supported mechanisms regulating the opening of specific sensitive periods. Fagiolini & Hensch (2000) manipulated inhibitory function in the mouse primary visual cortex and provided evidence that a threshold of inhibition is necessary for the usual onset of ocular-dominance plasticity. The important inference is causal and circuit-specific: it derives from experimental perturbation of an inhibitory system with a measurable deprivation response. A threshold account explains why elevated excitatory drive alone does not identify a window of maximal reorganizability and why the same input has different consequences before and after particular interneuron networks become functional.
Parvalbumin-expressing interneurons contribute powerful, precisely timed inhibition to local circuits. Their maturation participates in regulating cortical dynamics, rhythmic coordination, and the timing of visual plasticity, although inhibitory cell types differ in connectivity and function. In the model reviewed by Hensch (2005), inhibitory maturation is a gate that allows experience to induce competitive refinement during a bounded developmental stage. Inhibitory control subsequently contributes to stabilization, making the relation between inhibition and plasticity nonlinear: too little developmental inhibition may delay opening, while mature regulation can constrain the magnitude or form of later remodeling. The causal account therefore concerns an organized circuit state, rather than a simplistic equation between more GABA and less learning.
Inhibitory findings demand careful distinctions among molecular expression, cellular physiology, and behavior. The presence of a marker for interneurons cannot itself establish that a person is inside a critical period. Pharmacological manipulations affect multiple processes, and altered performance can arise through attention, arousal, or activity changes rather than reorganization of a developmental window. Mechanistic confidence is greatest when targeted interventions shift both the timing of physiology and the predicted functional consequence. The review by Takesian & Hensch (2013) integrates these developmental gates with mechanisms that stabilize later circuits, supporting a balance between opportunities for adaptation and the maintenance of established functions.
Otx2, Experience, and Molecular Timing of Cortical Readiness
The homeoprotein Otx2 provides a concrete example of developmental information reaching cortical inhibitory circuitry through a non-cell-autonomous pathway. In mouse visual cortex, Sugiyama et al. (2008) investigated the experience-dependent accumulation of Otx2 in parvalbumin-positive cells. Their interventions connected the protein's transfer to maturation of the inhibitory network and to altered timing of ocular-dominance plasticity. This result strengthens a developmental-clock interpretation because it concerns a specific molecular signal and measured perturbations, rather than a correlation between age and behavioral improvement. It also shows that experience can influence the biological conditions that subsequently determine sensitivity to experience.
A useful account of Otx2 separates the source of the signal, its transfer and accumulation, the cell population receiving it, and the downstream physiological state. Those components can be investigated through conditional manipulations and local interventions, each of which tests a narrower causal claim. The resulting architecture suggests feedback: experience regulates a permissive molecular process, which helps shape the duration and intensity of later experience-dependent reorganization. In such a system chronological time is partly mediated by prior input, making developmental sensitivity history-dependent. Sugiyama et al. (2008) supports this mechanism for the studied visual preparation; whether comparable pathways gate human language or complex social cognition remains a separate research question.
Molecularly anchored hypotheses help connect experimental neuroscience to longitudinal developmental research. Instead of selecting an arbitrary birthday as the onset of sensitivity, investigators can seek converging evidence for changes in inhibitory architecture, extracellular organization, and input-dependent functional adaptation. Noninvasive human imaging cannot directly sample every signaling component demonstrated in animals, so proposed markers require independent validation. A proxy for cortical maturation is only a useful sensitive-period marker if it predicts when manipulation or naturally varying experience has a different functional effect. The hierarchical human framework examined by Sydnor et al. (2026) makes this validation problem central.
Perineuronal Nets, Extracellular Matrix, and Closure Mechanisms
Perineuronal nets are specialized extracellular matrix structures associated with subsets of neurons, including mature parvalbumin interneurons. Their developmental organization can restrict structural and functional rearrangement of connections. In rat visual cortex, Pizzorusso et al. (2002) reported that enzymatic degradation of chondroitin sulfate proteoglycans in adulthood restored susceptibility to ocular-dominance shifts after monocular deprivation. This experiment provided mechanistic evidence that the extracellular environment contributes to closure, rather than merely correlating with maturation. It also established that a developmental boundary can be partially reopened by modifying local biological constraints.
Related evidence comes from the amygdala and learned fear. Gogolla et al. (2009) manipulated perineuronal nets and studied the developmental change in whether extinction training could erase previously established conditioned fear. Their results supported an extracellular-matrix contribution to the stabilization of fear memories in mice. The conceptual value is that mechanisms of consolidation and closure can appear in systems other than visual cortex, although the relevant learning tasks and neural substrates differ. Extinction, spontaneous recovery, and reconsolidation have specific experimental meanings; the claim that nets protect a particular animal fear memory cannot be expanded without further evidence into a universal account of human trauma or therapeutic outcomes.
A critical tension emerges between plasticity and stability. Stable representations enable efficient, reliable perception and action, and unrestrained change could degrade already established functions. Molecular constraints on modification may therefore be adaptive components of development rather than obstacles whose elimination always benefits performance. Takesian & Hensch (2013) and Bavelier et al. (2010) situate extracellular, inhibitory, and behavioral mechanisms within this wider balance. Translational proposals that aim to loosen adult constraints must evaluate the selectivity of reorganization, persistence of benefits, unforeseen consequences, and the distinction between restoration and compensatory learning.
Myelin, Glial Regulation, and Developmental Synapse Elimination
As circuits mature, myelination supports reliable long-range signaling and can modify the opportunity for anatomical reorganization. McGee et al. (2005) studied Nogo-receptor signaling in mice and reported that its disruption permitted ocular-dominance plasticity beyond the ordinary juvenile interval. The experiment identified myelin-associated signaling as a restriction on later sensory plasticity. Because changes in axonal conduction, anatomical stability, and synaptic competition operate at different spatial scales, white-matter growth measured with MRI should not be treated as a direct assay of Nogo-dependent closure. The mechanistic result is evidence for a specific brake, not an argument that all age-related declines in learning derive from myelination.
Refinement also involves active elimination of connections. Stevens et al. (2007) found complement involvement in retinogeniculate synapse elimination, including deficits in anatomical refinement when key complement components were absent. Schafer et al. (2012) demonstrated activity- and complement-dependent microglial engulfment of presynaptic elements in the developing retinogeniculate system. These studies connect electrical activity, immune-related signaling molecules, glial cells, and anatomical outcome through intervention. Paolicelli et al. (2011) supplied convergent evidence for the importance of microglial pruning in normal developmental connectivity. The combined work extends the explanatory focus beyond neurons alone, showing that circuit assembly is a multicellular developmental process.
Pruning is not a synonym for a loss of intelligence or learning potential. Selective elimination can improve fidelity by reducing redundant or poorly coordinated connections. At the same time, an observed reduction in density may arise from several mechanisms and is sensitive to regional and developmental context. Postmortem structural evidence in humans Huttenlocher & Dabholkar (1997) Petanjek et al. (2011) aligns with the general importance of refinement but does not establish identical cell-by-cell processes or exact behavioral windows. A strong integrative explanation specifies a chain from experience and activity to molecular selection, circuit organization, and measured behavioral adaptation, with the evidential gaps at each link made explicit.
Human Visual Development and the Consequences of Early Deprivation
Human vision offers unusually informative evidence because some children experience restricted patterned input from congenital or early cataracts, and the timing of treatment can be documented. Lewis & Maurer (2005) synthesized psychophysical studies of visual development following such deprivation, showing that acuity, global motion, binocular integration, and other capacities do not share one uniform sensitive period. Their framework differentiates intervals for ordinary development, vulnerability to abnormal input, and recovery after treatment. This differentiation is essential: the period in which deprivation does the greatest harm need not coincide exactly with the period when later experience can yield the greatest repair.
Evidence from restored vision must be interpreted in light of clinical selection. Duration of opacity, unilateral versus bilateral restriction, optical correction, ocular health, treatment access, posttreatment therapy, and the precision of developmental records can alter outcomes. Children who receive treatment earlier may also differ systematically in family resources and access to care. Such circumstances complicate causal estimates of a unique neural closing age, although convergent outcome patterns remain informative. The studies reviewed by Lewis & Maurer (2005) help establish that early patterned vision matters, while the exact developmental trajectory should be modeled separately for distinct tasks and physiological subsystems.
Animal ocular-dominance interventions offer mechanistic hypotheses for human outcomes, including input competition and inhibitory consolidation Hubel & Wiesel (1970) Hensch (2005). Human studies supply ecological relevance and longer developmental follow-up but generally lack the cellular interventions used to establish those mechanisms in rodents or cats. A persuasive translational account therefore triangulates animal physiology, human visual psychophysics, and anatomical measures without assuming that matching functional deficits identify an identical causal pathway. Improvement in acuity after treatment, for example, is a functional outcome; full normalization of binocular cortical organization is a stronger and separate claim.
Infant Speech Perception and Native-Language Attunement
Speech-perception research demonstrates how early experience shapes representational specialization during the first year. Werker & Tees (1984) compared discrimination of native and nonnative phonetic distinctions across developmental stages and reported perceptual reorganization influenced by language exposure. The finding helped replace a simple maturation-only account with a process in which the auditory and language environment systematically changes the organization of categories. The observed narrowing of discrimination on some nonnative contrasts should not be confused with a global decline in auditory ability. Native-language attunement often accompanies improved recognition and processing of frequently encountered distinctions, and trajectories vary by contrast, language, and bilingual experience Werker (2024).
In a direct experimental investigation, Kuhl et al. (2003) exposed American infants to Mandarin speech in short laboratory sessions and compared live social instruction with audiovisual presentation. The study's results indicated an important role for social interaction in phonetic learning under its particular conditions. Its interpretive scope is specific: it identifies effects of experimental exposure and interaction on measured phonetic discrimination, while it does not establish that any recorded social contact is necessary or sufficient for language acquisition in general. The claim also does not demonstrate an exact molecular switch in the developing human auditory cortex. A biologically plausible account still requires bridging evidence across neural and behavioral levels.
The review by Werker & Hensch (2015) relates infant phonetic learning to candidate sensitive-period mechanisms and emphasizes questions about timing and plasticity regulation. Kuhl (2004) outlines statistical, prosodic, and social components of speech development. Together the literature suggests that multiple learning mechanisms cooperate as early experience becomes integrated with prior representations. Social input can modify attention and the availability of informative exemplars, while statistical regularities help establish category boundaries. Developmental change in discrimination is therefore best analyzed as a coupled process involving sensory statistics, perceptual categories, neural maturation, and the ecology of interaction.
Statistical Learning, Neural Commitment, and Linguistic Timing
Language development supplies a particularly clear distinction between rapid learning mechanisms and sensitive-period inference. In the original experimental study by Saffran et al. (1996), eight-month-old infants extracted regularities from a brief stream of continuous speech, demonstrating that distributional information can support early word segmentation. This result identifies a learning process observable during infancy. A critical-period claim would additionally require showing that its effects or constraints change systematically with developmental stage under comparable conditions. The mere demonstration of rapid infant learning does not show that the adult brain lacks analogous statistical-learning capacities.
The neural commitment account discussed by Kuhl (2004) proposes that exposure to native-language regularities organizes representations that subsequently facilitate compatible patterns while constraining different categories. On this view, changing performance across age can partly reflect the cumulative consequences of learning, not solely the exhaustion of an intrinsically timed cellular mechanism. The distinction can be tested through bilingual histories, altered exposure schedules, interventions targeting specific phonemic contrasts, and comparisons of children and adults trained on matched materials. Werker (2024) reviews later work extending the original observations and explicitly examines whether perceptual attunement deserves classification as a critical-period phenomenon.
Second-language outcomes form an adjacent but separate research domain. Johnson & Newport (1989) examined age-related differences in English grammatical proficiency among second-language learners, placing maturational hypotheses into a measurable behavioral framework. However, age of acquisition is correlated with length and quality of input, motivation, educational opportunity, native-language structure, and social context. Even a robust age gradient would not by itself prove a single neural deadline, and grammatical outcomes cannot be substituted for phonetic discrimination. The most precise synthesis treats speech sound attunement, language-related statistical learning, and second-language attainment as partly connected processes with potentially different mechanisms and temporal profiles.
Cortical Growth, Association Systems, and Protracted Maturation
Longitudinal anatomical MRI allows repeated observation of brain morphology without exposure to ionizing radiation. Giedd et al. (1999) reported nonlinear gray-matter changes and more sustained increases in white matter during childhood and adolescence, helping establish the importance of developmental trajectories rather than one-time age comparisons. Lenroot & Giedd (2006) reviewed broader MRI evidence and emphasized regional variation and large between-person overlap. Morphometric measures reflect complex mixtures of cellular and tissue processes, so a decrease in cortical gray-matter volume or thickness cannot be read literally as a counted loss of synapses. A longitudinal change in morphology is evidence of development, while a sensitive period requires a demonstrated interaction between age and the effect of experience.
Postmortem measures provide a complementary scale. Huttenlocher & Dabholkar (1997) identified regionally staggered synaptic development; Petanjek et al. (2011) documented unusually protracted changes in prefrontal dendritic spines. These studies support a developmental ecology in which primary sensory and high-order association systems can have different structural opportunities for experience-dependent refinement. Yet their samples are not the same individuals followed through time, and structural markers from a small region cannot establish a universal schedule for executive control, social behavior, or affect regulation. Integrating imaging and histology is most productive when explicit predictions distinguish maturation from plasticity and are supported by developmental intervention data.
A contemporary synthesis by Sydnor et al. (2026) proposes that plasticity-related developmental changes may progress along a sensorimotor-to-association cortical hierarchy and highlights the possibility of later windows in association systems. The review explicitly calls for stronger research to establish the precise presence, timing, and behavioral consequences of human high-order critical periods. Its contribution is a structured hypothesis supported by converging clues, not a completed molecular assay of human adolescent executive or social plasticity. The approach should guide prospective multimodal studies that measure experience, maturation, and function at repeated ages in the same participants.
Individual Differences, Exposure History, and Developmental Ecologies
Developmental sensitivity is expressed within a distribution of individual histories. Baseline sensory access, sleep, health, nutrition, social interaction, language environment, educational opportunity, and exposure continuity can change the meaning of an age-specific performance score. Experimental models demonstrate the influence of defined input under laboratory control; human development occurs in environments where relevant factors co-vary. Claims about optimal windows must therefore specify the type and intensity of input, the developmental outcome, and the target population. Knudsen (2004) describes sensitivity as a property of neural circuitry responding to environmental information, making environmental characterization part of the mechanism rather than a secondary demographic adjustment.
An important distinction is between deprivation, enrichment, and ordinary experience. Severe sensory deprivation can produce large developmental deficits while modest variation in everyday learning opportunity produces smaller or differently organized effects. Evidence that the absence of a sensory input alters cortical development cannot automatically establish that unusually intense training of the same capacity during that period yields proportional benefits. Lewis & Maurer (2005) addresses deprivation and visual restoration, whereas Kuhl et al. (2003) manipulates short-term language exposure under controlled laboratory conditions. These represent different causal questions and have different relevance to educational intervention design.
Human research needs appropriate comparisons across sex, language background, socioeconomic access, disability status, and cultural context, with attention to both ethical representation and measurement equivalence. Cross-sectional age differences may conceal cohort effects; longitudinal samples can be affected by attrition and changing tasks. A theoretical claim about developmental opportunity becomes stronger when its effect replicates across circumstances and when the proposed neural mediator can be measured independently from the behavioral target. The hierarchical model of Sydnor et al. (2026) provides a framework for such tests, while the detailed human evidence on speech Werker & Hensch (2015) demonstrates why generalization across tasks should remain conditional.
Experimental Designs for Detecting Developmental Windows
A rigorous sensitive-period experiment compares the effects of a well-defined manipulation introduced at different developmental times while holding the causal contrast as constant as biological and ethical conditions permit. Factorial designs may cross age, exposure duration, deprivation intensity, and recovery conditions. The decisive statistical question is whether the treatment effect differs by developmental stage, not simply whether younger and older groups have different mean outcomes. Age-dependent baseline performance, measurement ceilings, and task familiarity can generate spurious interactions. Hubel & Wiesel (1970) illustrates the strength of a direct age-series deprivation design in a controlled animal preparation, although its estimated window is specific to that preparation.
Longitudinal designs can distinguish changes within an individual from cross-sectional differences among individuals. Multiple repeated measures before and after exposure provide information about prior trends, developmental slope, and delayed effects. However, repeated assessment can itself introduce practice and adaptation. Randomization of a deprivation procedure is generally inappropriate in human children; natural experiments, carefully defined clinical cohorts, observational records, and ethically provided interventions therefore carry much of the human evidence. Lewis & Maurer (2005) exemplifies how human visual histories can inform developmental sensitivity even when exposure assignment cannot be randomized.
A design capable of identifying a true developmental gate should investigate more than one behavioral endpoint and, where feasible, link it to a mechanistic mediator. For example, the association between age and MRI cortical thickness becomes evidence about critical-period mechanisms only when it predicts a differential response to relevant experience in a prespecified task. Sydnor et al. (2026) identifies this challenge in human association cortex. Experimental controls should distinguish lack of exposure from an absence of learning readiness, and recovery studies should compare changes to appropriate maturational trajectories rather than attributing every post-intervention gain to reopening of an early developmental program.
Multiscale Measurement and Cross-Species Inference
Intracellular and extracellular electrophysiology measures activity and circuit responsiveness at a scale unavailable in ordinary pediatric neuroimaging. Cellular tagging, genetic perturbation, biochemical assays, histological counts, and manipulation of extracellular matrix components permit strongly localized causal claims in animal models. These tools supported evidence about inhibitory onset Fagiolini & Hensch (2000), homeoprotein transfer Sugiyama et al. (2008), extracellular constraints Pizzorusso et al. (2002), and microglial synapse elimination Schafer et al. (2012). Their inferential strength depends on successful controls for specificity, developmentally appropriate dosing, and distinguishing direct effects on a circuit from systemic developmental disruption.
Human studies use distinct measurement families. Structural MRI estimates morphology and myelination-related properties indirectly; functional MRI indexes task-related hemodynamic responses; EEG and MEG reveal population-level timing; eye tracking and psychophysics characterize sensory and behavioral function. Developmental recordings have age-dependent noise and artifact profiles, including motion and changes in skull and tissue characteristics. A neural developmental marker is valid only for the biological property it reliably measures. Reviews such as Lenroot & Giedd (2006) provide foundational interpretation of anatomical trajectories, but these measures do not reconstruct single-neuron receptive fields or complement-mediated synaptic pruning.
Triangulation should be organized around an explicit mechanistic chain: a manipulation changes a molecular or cellular process; the change alters a circuit phenotype; the altered circuit predicts a functional outcome. A relevant human observation may support the last link without directly verifying the first. Species also differ in developmental scheduling, social environment, cortical organization, and behavioral demands. A cat ocular-dominance critical period does not identify the human age when an executive capacity becomes fixed. Knudsen (2004) offers the framework for treating sensitive periods as circuit properties, with general principles tested through system-specific assays.
Adult Plasticity and Reopening Developmental Constraints
The decline of an early developmental sensitivity does not eliminate learning or structural reorganization in adulthood. Sato & Stryker (2008) documented distinctive adult ocular-dominance plasticity in mice, demonstrating that juvenile and adult responses may differ in time course and mechanism. McGee et al. (2005) showed that interfering with myelin-associated Nogo receptor signaling allowed juvenile-like ocular-dominance susceptibility to persist into adult stages under the experimental conditions. These findings justify the search for multiple forms of adult plasticity but caution against a simple binary model of open versus closed brains. An adult response may be slower, weaker, differently regulated, or dependent on particular interventions.
Extracellular manipulation supplies especially strong evidence for reopening certain plasticity phenotypes. Pizzorusso et al. (2002) found that degrading chondroitin sulfate proteoglycans permitted an adult cortical response to monocular deprivation in rats. Sale et al. (2007) reported that environmental enrichment in adult amblyopic rats improved measured visual function and was accompanied by changes in inhibitory and extracellular markers. Causal inferences in these experiments concern the species, preparation, and intervention tested. The broader implication is the existence of modifiable constraints on later plasticity, rather than evidence of routine clinical reversibility of all childhood deprivation effects.
The integrative review by Bavelier et al. (2010) describes approaches for loosening restrictions on adult plasticity, from molecular manipulation to behavioral training. This agenda faces a central design problem: restoring some ability to reorganize circuitry does not guarantee that new learning will be appropriately directed or stable. Timing, content of training, functional specificity, and adverse effects of altered circuit stability all need evaluation. Translation also requires replicable gains in ecologically meaningful outcomes over clinically relevant periods. The existence of plasticity at the circuit level is a prerequisite for many therapies, but not a substitute for evidence that a particular intervention is effective and safe in humans.
Recovery, Amblyopia, and the Difference Between Restoration and Compensation
Amblyopia research combines developmental theory with clinical questions because early imbalance in visual input can lead to durable alterations in perceptual function. Its causes, presentation, and treatment responses are heterogeneous. The original animal deprivation findings Hubel & Wiesel (1970) demonstrated dramatic age-dependent sensitivity in cortical responses, while human studies of early cataracts Lewis & Maurer (2005) showed that several visual functions follow different schedules for vulnerability and recovery. There is no single visual critical period applicable to all visual functions and treatment contexts. A clinically meaningful statement should specify the visual phenotype and the intervention rather than invoking a general expiration date for brain development.
Behavioral training can produce improvements in perceptual performance after the classic early windows. Levi & Li (2009) reviews perceptual-learning research relevant to amblyopia and the conditions under which adult visual function may improve. Review evidence should not be represented as one definitive randomized result; training protocols, baseline deficits, controls, outcome selection, and durability vary. Improved discrimination on practiced stimuli may reflect task-specific learning, changes in decision strategies, or broader neural adaptation. Demonstrating restoration of a normal visual pathway requires evidence beyond an improvement on the trained test.
Clinical translation requires prospective controls, masking where feasible, standard visual outcome measures, consideration of adherence, and follow-up sufficient to assess persistence. Studies in rodents using molecular inhibition or enrichment Pizzorusso et al. (2002) Sale et al. (2007) provide explanatory hypotheses and intervention targets, but their interventions cannot be assumed effective or appropriate for children or adults with human amblyopia. The medical significance of sensitive periods includes timely access to assessment and evidence-based treatment, while precise clinical decisions remain governed by condition-specific professional guidance. This monograph analyzes the scientific evidence and does not prescribe individualized treatment.
Threats to Validity, Replication, and the Interpretation of Age Effects
Age-dependent developmental studies face several recurring biases. Restricted age sampling can make a smooth change appear discontinuous. Exposure histories may be measured retrospectively, and a single performance score can combine perception, attention, motivation, and motor demands. Cross-sectional differences may reflect cohort or social context rather than within-person change. Even experiments with random assignment to a treatment condition may lack random assignment to developmental age, leaving age correlated with baseline maturation. Claims about sharply delimited windows therefore require direct tests of alternative functional forms, including graded sensitivity and multiple overlapping windows Knudsen (2004).
Animal experiments can be threatened by small numbers of independent animals, unbalanced litters, sex restrictions, unreported exclusion criteria, and the temptation to treat many neuronal recordings from one preparation as independent replicates. Investigators should document randomization, blinding, unit of analysis, and replication across animals, laboratories, and genetic backgrounds. Human longitudinal studies require transparent retention statistics, age-adjusted measures, reliability checks, and assessment of selection. A beautiful developmental curve is not necessarily a causal model. The evidence base summarized by Hensch (2005) is strongest where the proposed molecular variable has been manipulated and the temporal response changes as predicted.
The modern human hierarchical hypothesis Sydnor et al. (2026) is scientifically productive because it differentiates evidence categories and identifies tests still needed. Correlated changes in cortical microstructure, age, and psychiatric symptoms cannot establish that a particular adolescent critical period caused vulnerability or resilience. Replication should examine prespecified brain regions and behaviors, environmental moderators, and independent cohorts with enough observations to estimate timing uncertainty. Claims about interventions should report effect sizes, confidence intervals, clinically relevant endpoints, and adverse outcomes where available. The evidential standard is convergence among design, mechanism, and function rather than the rhetorical force of the phrase critical window.
Comparative Theories of Timing: Maturation, Competition, and Experience
Explanatory models of sensitive periods vary according to what sets their temporal boundary. A maturation-centered model emphasizes intrinsic programs that alter circuit readiness as development proceeds. Competition-based models emphasize the balance and correlation of alternative inputs, showing how deprivation shifts their relative representation. History-dependent learning models propose that accumulated representations change which new regularities can be acquired efficiently. Circuit-gating models identify specific molecular transitions such as inhibitory maturation, Otx2 transfer, and extracellular stabilization Fagiolini & Hensch (2000) Sugiyama et al. (2008) Pizzorusso et al. (2002). These accounts need not be mutually exclusive: developmental programs can prepare circuits for activity-dependent refinement, and learning can change the very conditions under which future plasticity occurs.
The models become scientifically distinct when they generate different counterfactual predictions. A purely maturational account anticipates a predictable change even when sensory exposure is altered, whereas a history-sensitive account predicts shifts when experience itself is delayed or reorganized. Manipulating inhibition may separate a gate from mere accumulated training, and selective matrix degradation may test whether an adult restriction is structural. Classic experiments establish aspects of these predictions in visual systems Hubel & Wiesel (1970) McGee et al. (2005). However, a manipulation may simultaneously alter several biological processes, so the causal interpretation improves when independent interventions converge on the same proposed mediator.
Human speech and cortical-development findings resist reduction to a single developmental clock. Phonetic attunement depends on language-specific input Werker & Tees (1984), and social presentation affects learning in particular infant protocols Kuhl et al. (2003). Cortical structural maturation follows regional schedules Petanjek et al. (2011), while hierarchical critical periods in association cortex remain under active investigation Sydnor et al. (2026). A useful overarching theory must permit task-specific sensitivity functions, exposure-dependent shifts, distinct clocks for vulnerability and recovery, and later learning after an early period has diminished. Such a model is testable because it predicts how interventions at different ages should change particular functions.
Scientific Synthesis: A Multilevel Organization of Developmental Windows
The accumulated evidence supports a multilevel account. First, development establishes cellular and anatomical prerequisites. Second, activity and experience select and refine patterns within that substrate. Third, inhibitory and extracellular processes regulate when substantial reorganization is accessible. Fourth, the stabilized architecture supports increasingly specialized performance while retaining forms of adult adaptation. Strong evidence for this sequence derives from converging perturbation studies of visual cortex and retinogeniculate circuitry Fagiolini & Hensch (2000) Stevens et al. (2007) Pizzorusso et al. (2002) Schafer et al. (2012). The principle applies as a research architecture across systems; each specific neural or psychological domain must supply its own evidence.
Sensitive periods should be represented as functions rather than universal labels. A general research formulation can describe outcome change as a function of developmental state, exposure characteristics, baseline organization, and assessment time. Distinct parameters govern ordinary experience-driven acquisition, susceptibility to deprivation, and recovery after deprivation Lewis & Maurer (2005). Within a domain, overlapping subsystems may have different values for those parameters. A child can remain capable of learning a skill after sensitivity to one foundational component has declined, and the reverse may also occur when experience is provided before the supporting circuitry is ready. The scientific task is to map these relations empirically.
The resulting framework also clarifies evidence hierarchy. Mechanistic experiments establish causal contributions within controlled model systems. Human natural experiments and longitudinal assessments constrain developmental interpretation in real populations. Structural imaging suggests candidate markers and schedules but does not directly assay sensitive-period onset. Reviews offer theoretical integration and highlight conflicts across paradigms Takesian & Hensch (2013) Sydnor et al. (2026). An interpretation is robust when those independent strands agree on a specific outcome while remaining explicit about measurements they cannot substitute for one another.
Psychological and Interdisciplinary Implications in the Artificial Era
Understanding sensitive-period plasticity matters to psychology because perceptual categories, learning habits, attention, regulation, and adaptive behavioral repertoires develop in interaction with organized neural tissue. The evidence supports a theory of temporally conditioned learning opportunities, with domain-specific biological constraints and experience histories. It does not license universal educational deadlines or deterministic predictions about individual intellectual potential. Developmental research can help identify opportunities for appropriate sensory access and enriched environments when interventions are supported by outcome evidence, and it can motivate research on rehabilitation beyond childhood Bavelier et al. (2010) Levi & Li (2009).
The Artificial Era introduces a separate comparative problem: how to study time-dependent adaptation in nonbiological learning systems and in human–artificial environments. Artificial systems may display initialization-dependent training sensitivity, changing learning rates, interference, or stability–plasticity tradeoffs, but these operational phenomena do not entail neurogenesis, GABAergic inhibition, perineuronal nets, or human sensitive periods. A valid cross-system comparison must define the measured learning outcome, experience schedule, plasticity constraint, and intervention separately. The conceptual contribution of biological developmental research is an experimental discipline of time, mechanism, and history rather than an assumption of substrate identity.
Human–artificial interaction can also influence the environments through which people develop: access to language exposure, feedback, educational resources, and cognitive support may change the distribution of developmental experiences. Claims that these changes affect a sensitive period require studies with prospectively measured exposure, appropriate control groups, developmental outcomes, and follow-up. A model of distributed support should evaluate how interaction quality alters human learning while distinguishing association from intervention effects. These questions belong to future empirical research, not to the existing causal evidence for ocular-dominance plasticity Knudsen (2004) Sydnor et al. (2026).
Unresolved Problems and a Research Program for the Next Generation
One unresolved problem is the causal identification of sensitive periods in human association cortex. Developmental imaging can detect regional maturation, but the presence and timing of mechanism-defined critical periods for executive control, socioemotional learning, and higher cognition remain uncertain Sydnor et al. (2026). Progress requires longitudinal cohorts with multiple developmental intervals, harmonized exposure measures, independent behavioral endpoints, multimodal neural markers, and explicit models of within-person change. Natural variation and ethical interventions could help estimate age-by-experience effects; causal claims would remain qualified by their design and by the validity of the proposed mediator.
A second problem concerns how diverse mechanisms interact across time. Inhibitory maturation, Otx2 accumulation, perineuronal nets, myelin-associated signaling, and glial pruning are experimentally implicated in particular circuits Fagiolini & Hensch (2000) Sugiyama et al. (2008) Pizzorusso et al. (2002) McGee et al. (2005) Schafer et al. (2012). Their contribution to overlapping, hierarchical, or recurrent plasticity windows is incompletely understood. Research should compare temporal changes in these mechanisms within the same circuit, employ selective interventions, assess compensatory responses, and follow behavioral outcomes long enough to test stability. Whether apparent reopening in an adult circuit reproduces juvenile mechanisms or creates a distinct form of adaptation is another experimentally discriminable question.
A third problem concerns equitable translation. Human studies often include participants whose exposure history, treatment access, language background, and neurological development differ in ways that affect generalizability. Researchers should prespecify target populations, measure environmental confounds, protect participants from unjustified deprivation, and distinguish mechanistic exploratory results from validated clinical interventions. The review literature on adult plasticity Bavelier et al. (2010) and visual rehabilitation Levi & Li (2009) identifies promising avenues, but the evidence needed for broad intervention guidance includes randomized comparisons when feasible, long-term outcomes, harms, and replication. Claims about optimal timing should remain tied to an empirically defined function, context, and intervention.
Conclusion
Neural development and sensitive-period plasticity form a coherent scientific field centered on the changing power of experience to organize neural circuits. The historical discovery of ocular-dominance susceptibility established that comparable sensory perturbations can have substantially different effects at different ages Wiesel & Hubel (1963) Hubel & Wiesel (1970). Subsequent interventions identified key cellular and molecular contributors to opening and restricting plasticity, including inhibitory maturation, homeoprotein transfer, extracellular matrix organization, myelin-related signaling, and complement-dependent synaptic refinement Fagiolini & Hensch (2000) Sugiyama et al. (2008) Pizzorusso et al. (2002) McGee et al. (2005) Stevens et al. (2007). These findings support a temporal organization of circuit development in which experience and biological readiness continuously interact.
Human visual and language research extends the investigation into psychophysical and educationally meaningful phenomena while demonstrating that the timing and consequences of experience differ among functions Lewis & Maurer (2005) Werker & Hensch (2015). Anatomical and imaging evidence shows heterogeneous regional development, and contemporary accounts propose possible later windows in association cortex Huttenlocher & Dabholkar (1997) Petanjek et al. (2011) Sydnor et al. (2026). The strongest general conclusion is that developmental plasticity comprises multiple overlapping, partly history-dependent windows. Claims of an invariant closing age, an irreversible loss of all learning, or guaranteed clinical reopening exceed what these studies establish.
The scientific program now moves toward mechanism-linked human longitudinal research and carefully controlled intervention studies. Its central questions concern when particular circuits are most responsive, what experiences select their developing functions, how stabilization is achieved, which forms of later change remain available, and how those processes can be assessed without conflating anatomical maturation with behavioral opportunity. This synthesis places development within a general psychology of temporally organized adaptation while preserving the distinctive evidence required for biological and artificial systems. Future research should translate conceptual coherence into independently replicated, outcome-specific causal tests.
References
- Bavelier, D., Levi, D. M., Li, R. W., Dan, Y., & Hensch, T. K. (2010). Removing brakes on adult brain plasticity: From molecular to behavioral interventions. Journal of Neuroscience, 30(45), 14964–14971. https://doi.org/10.1523/JNEUROSCI.4812-10.2010
- Fagiolini, M., & Hensch, T. K. (2000). Inhibitory threshold for critical-period activation in primary visual cortex. Nature, 404, 183–186. https://doi.org/10.1038/35004582
- Giedd, J. N., Blumenthal, J., Jeffries, N. O., Castellanos, F. X., Liu, H., Zijdenbos, A., Paus, T., Evans, A. C., & Rapoport, J. L. (1999). Brain development during childhood and adolescence: A longitudinal MRI study. Nature Neuroscience, 2, 861–863. https://doi.org/10.1038/13158
- Gogolla, N., Caroni, P., Lüthi, A., & Herry, C. (2009). Perineuronal nets protect fear memories from erasure. Science, 325(5945), 1258–1261. https://doi.org/10.1126/science.1174146
- Hensch, T. K. (2005). Critical period plasticity in local cortical circuits. Nature Reviews Neuroscience, 6, 877–888. https://doi.org/10.1038/nrn1787
- Hubel, D. H., & Wiesel, T. N. (1970). The period of susceptibility to the physiological effects of unilateral eye closure in kittens. Journal of Physiology, 206(2), 419–436. https://doi.org/10.1113/jphysiol.1970.sp009022
- Huttenlocher, P. R., & Dabholkar, A. S. (1997). Regional differences in synaptogenesis in human cerebral cortex. Journal of Comparative Neurology, 387(2), 167–178. https://pubmed.ncbi.nlm.nih.gov/9336221/
- Johnson, J. S., & Newport, E. L. (1989). Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language. Cognitive Psychology, 21(1), 60–99. https://doi.org/10.1016/0010-0285(89)90003-0
- Knudsen, E. I. (2004). Sensitive periods in the development of the brain and behavior. Journal of Cognitive Neuroscience, 16(8), 1412–1425. https://doi.org/10.1162/0898929042304796
- Kuhl, P. K. (2004). Early language acquisition: Cracking the speech code. Nature Reviews Neuroscience, 5, 831–843. https://doi.org/10.1038/nrn1533
- Kuhl, P. K., Tsao, F.-M., & Liu, H.-M. (2003). Foreign-language experience in infancy: Effects of short-term exposure and social interaction on phonetic learning. Proceedings of the National Academy of Sciences, 100(15), 9096–9101. https://doi.org/10.1073/pnas.1532872100
- Lenroot, R. K., & Giedd, J. N. (2006). Brain development in children and adolescents: Insights from anatomical magnetic resonance imaging. Neuroscience & Biobehavioral Reviews, 30(6), 718–729. https://doi.org/10.1016/j.neubiorev.2006.06.001
- Levi, D. M., & Li, R. W. (2009). Perceptual learning as a potential treatment for amblyopia: A mini-review. Vision Research, 49(21), 2535–2549. https://doi.org/10.1016/j.visres.2009.02.010
- Lewis, T. L., & Maurer, D. (2005). Multiple sensitive periods in human visual development: Evidence from visually deprived children. Developmental Psychobiology, 46(3), 163–183. https://doi.org/10.1002/dev.20055
- McGee, A. W., Yang, Y., Fischer, Q. S., Daw, N. W., & Strittmatter, S. M. (2005). Experience-driven plasticity of visual cortex limited by myelin and Nogo receptor. Science, 309(5744), 2222–2226. https://doi.org/10.1126/science.1114362
- Paolicelli, R. C., Bolasco, G., Pagani, F., Maggi, L., Scianni, M., Panzanelli, P., Giustetto, M., Ferreira, T. A., Guiducci, E., Dumas, L., Ragozzino, D., & Gross, C. T. (2011). Synaptic pruning by microglia is necessary for normal brain development. Science, 333(6048), 1456–1458. https://doi.org/10.1126/science.1202529
- Petanjek, Z., Judaš, M., Šimić, G., Rašin, M. R., Uylings, H. B. M., Rakic, P., & Kostović, I. (2011). Extraordinary neoteny of synaptic spines in the human prefrontal cortex. Proceedings of the National Academy of Sciences, 108(32), 13281–13286. https://doi.org/10.1073/pnas.1105108108
- Pizzorusso, T., Medini, P., Berardi, N., Chierzi, S., Fawcett, J. W., & Maffei, L. (2002). Reactivation of ocular dominance plasticity in the adult visual cortex. Science, 298(5596), 1248–1251. https://doi.org/10.1126/science.1072699
- Saffran, J. R., Aslin, R. N., & Newport, E. L. (1996). Statistical learning by 8-month-old infants. Science, 274(5294), 1926–1928. https://doi.org/10.1126/science.274.5294.1926
- Sale, A., Maya Vetencourt, J. F., Medini, P., Cenni, M. C., Baroncelli, L., De Pasquale, R., & Maffei, L. (2007). Environmental enrichment in adulthood promotes amblyopia recovery through a reduction of intracortical inhibition. Nature Neuroscience, 10, 679–681. https://doi.org/10.1038/nn1899
- Sato, M., & Stryker, M. P. (2008). Distinctive features of adult ocular dominance plasticity. Journal of Neuroscience, 28(41), 10278–10286. https://doi.org/10.1523/JNEUROSCI.2451-08.2008
- Schafer, D. P., Lehrman, E. K., Kautzman, A. G., Koyama, R., Mardinly, A. R., Yamasaki, R., Ransohoff, R. M., Greenberg, M. E., Barres, B. A., & Stevens, B. (2012). Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Neuron, 74(4), 691–705. https://doi.org/10.1016/j.neuron.2012.03.026
- Stevens, B., Allen, N. J., Vazquez, L. E., Howell, G. R., Christopherson, K. S., Nouri, N., Micheva, K. D., Mehalow, A. K., Huberman, A. D., Stafford, B., Sher, A., Litke, A. M., Lambris, J. D., Smith, S. J., John, S. W. M., & Barres, B. A. (2007). The classical complement cascade mediates CNS synapse elimination. Cell, 131(6), 1164–1178. https://doi.org/10.1016/j.cell.2007.10.036
- Sugiyama, S., Di Nardo, A. A., Aizawa, S., Matsuo, I., Volovitch, M., Prochiantz, A., & Hensch, T. K. (2008). Experience-dependent transfer of Otx2 homeoprotein into the visual cortex activates postnatal plasticity. Cell, 134(3), 508–520. https://doi.org/10.1016/j.cell.2008.05.054
- Sydnor, V. J., Ojha, A., Larsen, B., Martinez, A., Calabro, F. J., & Luna, B. (2026). Investigating hierarchical critical periods in human neurodevelopment. Neuropsychopharmacology, 51, 67–85. https://doi.org/10.1038/s41386-025-02246-5
- Takesian, A. E., & Hensch, T. K. (2013). Balancing plasticity/stability across brain development. Progress in Brain Research, 207, 3–34. https://doi.org/10.1016/B978-0-444-63327-9.00001-1
- Werker, J. F. (2024). Phonetic perceptual reorganization across the first year of life: Looking back. Infant Behavior and Development, 75, 101935. https://doi.org/10.1016/j.infbeh.2024.101935
- Werker, J. F., & Hensch, T. K. (2015). Critical periods in speech perception: New directions. Annual Review of Psychology, 66, 173–196. https://doi.org/10.1146/annurev-psych-010814-015104
- 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. https://doi.org/10.1016/S0163-6383(84)80022-3
- Wiesel, T. N., & Hubel, D. H. (1963). Single-cell responses in striate cortex of kittens deprived of vision in one eye. Journal of Neurophysiology, 26(6), 1003–1017. https://doi.org/10.1152/jn.1963.26.6.1003