For more than a century, classical neurology operated under the foundational premise of strict neocortical modularity. Canonical neuroscience posited that primary sensory cortices were hardwired, genetically determined computational processors dedicated exclusively to their corresponding peripheral sensory organs. Within this localizationist architecture, the primary visual cortex—situated within the calcarine sulcus of the occipital lobe—was understood to be unalterably dedicated to decoding photic signals transduced by the retina. Absence of visual afferent input was widely assumed to condemn these specialized occipital networks to functional silence, progressive transneuronal degeneration, or evolutionary atrophy. The visual cortex, stripped of its biological imperative, was viewed as a biological relic within the sensory-deprived brain.
This long-standing paradigm was fundamentally overturned in 1996 through the groundbreaking functional neuroimaging investigations led by Japanese neuroscientist Norihiro Sadato and his colleagues at the United States National Institutes of Health. Utilizing positron emission tomography (PET), Sadato demonstrated that early blind individuals actively engaged their primary visual cortex while reading tactile Braille script. Even more remarkably, sighted control participants performing the identical tactile discrimination tasks exhibited pronounced metabolic deactivation of the very same occipital regions. This profound observation was not merely a demonstration of passive cross-talk; subsequent transcranial magnetic stimulation experiments confirmed that this occipital activation was functionally necessary for tactile perception, delivering an empirical blow to rigid localizationism.
Sadato’s findings catalyzed an epistemological revolution across systems neuroscience, cognitive neuropsychology, and philosophy of mind. The revelation that primary visual cortices could be repurposed to execute complex somatosensory and linguistic computations revealed an unprecedented degree of neuroplasticity within the human brain. It spurred decades of research into developmental critical periods, cortico-cortical connectivity, subcortical rewiring, and the metamodal nature of the neocortex. By tracing the trajectory of Sadato’s discoveries from their historical antecedents to their modern clinical and neurotechnological applications, we unveil how sensory deprivation illuminates the profound capacity of the human central nervous system to reorganize its neural real estate in response to environmental demands.
1. Historical Paradigms of Cortical Specialization and Early Hypotheses on Blindness
1.1 Classical Localization Theory and Modality Specificity
The dawn of modern systems neuroscience was defined by the doctrine of cerebral localization, spearheaded by the phrenological concepts of Franz Joseph Gall and subsequently formalized through the rigorous clinico-pathological observations of Paul Broca, Carl Wernicke, and John Hughlings Jackson. As the nineteenth century transitioned into the twentieth, histological investigations sought to establish the physical architecture underlying this functional segregation. Most prominently, Korbinian Brodmann executed his monumental comparative cytoarchitectonic analyses, delineating the human neocortex into distinct anatomical fields based on laminar composition, neuronal density, and cell morphology. Within this framework, Brodmann Area 17—the striate cortex of the occipital pole—was identified as the anatomical seat of primary visual reception, distinguished by its prominent line of Gennari in layer IV.
Concurrently, the physiological tenets of Johannes Peter Müller’s “law of specific nerve energies” were projected directly onto the cerebral mantle. Müller had asserted that the mind perceives not external objects directly, but rather the states of sensory nerves, with each sensory nerve possessing an innate, immutable quality. Transposed to neocortical architecture, this dogma crystallized into the belief that primary sensory cortices were intrinsically modality-specific. Striate cortex was biologically predestined to process visual attributes: retinotopic space, luminance, wavelength, and spatial frequency. The functional specialization was understood to be structurally hardwired, leaving virtually no conceptual space for the notion that non-photic inputs could systematically engage early visual processing arrays.
Consequently, early clinical assumptions regarding visual deafferentation—whether congenital or acquired early in life—assumed that the deprived occipital cortex would suffer a fate of passive obsolescence. Classical neurologists hypothesized that without retinal input to drive synaptogenesis and maintain cortical tone, Area 17 would undergo severe retrograde and transneuronal degeneration, descending into quiescent atrophy. The conceptual barrier against cross-modal reassignment was immense; to propose that visual cortex could process tactile, somatosensory, or auditory signals was viewed as a biological impossibility that violated fundamental tenets of neuroanatomy.
1.2 Early Animal Models of Sensory Deprivation
Empirical challenges to the immutability of sensory architecture began to coalesce through mid-twentieth-century experimental neurophysiology, most notably through the pioneering work of David Hubel and Torsten Wiesel. Their systematic investigations into the visual cortex of kittens and monkeys revealed that the functional architecture of the primary visual cortex is not rigidly hardwired at birth, but rather shaped during post-natal life through environmental exposure. Hubel and Wiesel demonstrated that monocular deprivation during a brief developmental window—termed the “critical period”—resulted in a dramatic reorganization of ocular dominance columns in layer IV of the striate cortex, where non-deprived inputs expanded to capture territory vacated by the silent eye.
However, Hubel and Wiesel’s classical paradigms also highlighted severe pathological consequences following complete binocular deprivation. When animals were dark-reared or subjected to bilateral eyelid suture from birth, their visual cortices suffered catastrophic structural deficits: extensive synaptic pruning, arrested dendritic spine maturation, synaptic degeneration, and a near-total failure of orientation-selective receptive fields to properly differentiate. These findings were initially interpreted as confirmation that early sensory deprivation invariably causes sensory cortices to wither into functional incompetence, reinforcing the hypothesis of occipital atrophy in human blindness.
Yet, scattered through the literature of dark-reared feline and rodent models, anomalies began to emerge. Researchers observed that in neonatally enucleated or dark-reared animals, non-visual sensory projections began to manifest in traditionally visual structures. Tracers revealed aberrant ascending projections from the inferior colliculus to the visual thalamus, as well as somatosensory inputs infiltrating the lateral geniculate nucleus and the deep layers of the visual cortex. Despite these provocative neuroanatomical curiosities, a formidable theoretical gap persisted. Rodent and feline models could demonstrate anomalous, uncalibrated electrophysiological spikes in response to whiskers or auditory clicks, but these animal studies could not elucidate whether such structural plasticity could sustain higher-order, complex cognitive and linguistic operations—such as the human capacity to decipher symbolic, language-dense tactile scripts like Braille.
1.3 Pre-Sadato Clinical Observations in Human Blindness
Long before functional neuroimaging permitted non-invasive visualization of living human neural activity, behavioral neurologists and psychologists recorded striking sensory paradoxes in early blind populations. Anecdotal accounts and psychophysical studies repeatedly documented that blind individuals frequently demonstrated perceptual acuity superior to sighted individuals across non-visual domains. Blind subjects displayed hyper-refined tactile spatial acuity, exceptional auditory localization within peripheral soundscapes, superior pitch discrimination, and sophisticated linguistic memory capacity.
These observations triggered an intense debate between two competing theoretical frameworks. The first, rooted in conservative peripheral sensory tuning models, argued that sensory compensation was entirely attributable to behavioral training, strategic shifts in attention, and tactile hyper-practice. Proponents of this view posited that blind individuals simply learned to optimize peripheral receptor utilization—such as maximizing Merkel-cell neurite complex readouts at the fingertips—without invoking fundamental structural or functional reorganization within the cerebral neocortex. The opposing hypothesis, radical for its time, posited that central neuroplastic compensatory shifts were taking place: the sensory-deprived brain was actively reallocating cortical processing real estate, repurposing silent visual zones to augment the computational bandwidth of remaining modalities.
Preliminary electroencephalographic (EEG) investigations conducted in the 1960s and 1970s offered tantalizing, albeit ambiguous, support for central plasticity models. Researchers recording event-related potentials (ERPs) during tactile and acoustic tasks observed unexpected electrical anomalies, showing shortened latencies and enlarged negative-wave amplitudes over the posterior scalp electrodes overlying the occipital lobes of blind subjects. Nevertheless, the intrinsic spatial limitations of scalp electrophysiology, plagued by volume conduction and poor spatial localization, prevented definitive attribution of these signals to the striate cortex itself. Skeptics maintained that the recorded signals represented far-field potentials generated by the parietal somatosensory cortices or generalized non-specific cognitive arousal. To definitively resolve this mechanistic dispute, the scientific community required functional neuroimaging methodologies possessing both high spatial resolution and metabolic specificity.
2. Norihiro Sadato’s Seminal 1996 Discovery: Primary Visual Cortex Activation During Braille Reading
2.1 Experimental Architecture of the 1996 Nature Study
The definitive breakthrough arrived in April 1996, when Dr. Norihiro Sadato and an international team of collaborators published their landmark paper in Nature, titled “Activation of the primary visual cortex by Braille reading in blind subjects.” Conducted at the National Institute of Neurological Disorders and Stroke (NINDS) at the NIH, the study was designed to isolate the neural substrates mediating tactile reading through an elegant, rigorously controlled experimental architecture. Sadato established two carefully stratified cohorts: eight early blind individuals who had lost their sight at birth or in early infancy (all proficient in tactile Braille reading), and a control group of ten sighted individuals with normal or corrected-to-normal vision.
To eliminate confounding variables associated with language parsing, motor execution, and passive skin displacement, Sadato engineered a sophisticated hierarchy of active and passive tactile conditions. The primary experimental task required blind participants to execute lexical Braille reading using their dominant reading finger, sweeping across four-letter embossed words to perform lexical identification. To isolate the purely somatosensory, non-linguistic mechanical components of the task, the researchers implemented tactile discrimination control tasks. These included discriminating the spatial orientation of non-lexical embossed dot patterns (detecting whether a raised pattern formed a horizontal or vertical array) and evaluating textured surfaces lacking geometric configuration.
Crucially, sighted control participants were trained to perform identical tactile discrimination tasks, sweeping their fingertips across identical arrays of embossed dots and textured surfaces. Furthermore, passive tactile stimulation tasks—in which Braille-like tactile stimuli were applied directly to the stationary fingerpad—and purely motor tasks—in which subjects executed identical sweeping finger movements across completely smooth surfaces—were incorporated into the subtraction paradigms. This rigorous task design ensured that any metabolic fluctuations detected in the occipital lobes could be definitively parsed, isolating tactile spatial discrimination, linguistic decoding, sensorimotor mechanics, and non-specific somatosensory stimulation.
2.2 Functional Neuroimaging Observations in the Occipital Pole
The results generated by Sadato’s PET imaging were unambiguous, dramatic, and fundamentally unprecedented. In the early blind cohort, lexical Braille reading elicited robust, highly statistically significant bilateral activation within the primary visual cortex (Brodmann Area 17) and associative visual areas (Brodmann Areas 18 and 19). The activation was concentrated symmetrically around the calcarine sulcus, precisely within the anatomical regions historically defined as the retinotopic map of central and peripheral vision. When blind participants performed non-lexical tactile discrimination of embossed dots, Area 17 similarly displayed intense metabolic recruitment, demonstrating that the primary visual cortex was directly engaged in processing spatial tactile geometry independent of lexical-semantic processing.
Even more startling was the functional divergence revealed when analyzing the sighted control cohort. When sighted individuals performed the identical tactile discrimination tasks, their primary visual cortices did not simply remain at a neutral baseline; instead, they demonstrated pronounced, statistically significant deactivation. Regional cerebral blood flow (rCBF) within the calcarine sulcus of sighted subjects plummeted well below resting levels during tactile exploration. This striking cross-modal sensory gating highlighted a classic inhibitory mechanism: in the sighted brain, the central nervous system suppressed irrelevant sensory cortices (occipital visual networks) to allocate metabolic and attentional resources to the primary and secondary somatosensory cortices (S1 and S2) tasked with deciphering the fingertip sensations.
Quantitative analysis of rCBF alterations established a stark dichotomy between the two groups. In early blind participants, regional blood flow in the striate cortex climbed by upwards of 4 to 8 percent during tactile tasks relative to resting baselines, a magnitude of hemodynamic response comparable to that evoked by high-contrast, dynamic photic stimulation in sighted individuals. Furthermore, Sadato uncovered an intricate functional topography: while non-Braille tactile discrimination activated primary visual cortex, lexical Braille reading engaged a broader, highly integrated network that extended beyond Area 17 into associative visual regions, ventral stream extrastriate zones, and classical perisylvian language structures, including Broca’s area. Tactile input had not merely leaked into the visual cortex; it had fully colonized it.
2.3 Immediate Theoretical Shockwaves in Cognitive Neuroscience
The publication of Sadato’s 1996 findings sent immediate theoretical shockwaves through the international cognitive neuroscience community. The study provided the first definitive functional neuroimaging proof that human primary sensory cortices are fundamentally pluripotent, capable of breaking through classical modality-exclusivity under conditions of early sensory deprivation. It dealt a decisive blow to the classical view of immutable cytoarchitectonic specialization, forcing a radical paradigm shift toward large-scale cross-modal neuroplasticity.
Inevitably, the magnitude of these findings provoked profound academic skepticism. Critics immediately raised the “epiphenomenon critique.” Skeptical neurophysiologists argued that the observed occipital activation might represent nothing more than unmasked evolutionary noise, passive hemodynamic spillover from parietal association areas, or a functional spandrel. Under this counter-hypothesis, the primary visual cortex, devoid of normal retinal inputs, was merely experiencing non-functional synaptic cross-talk—a metabolic echo lacking any genuine computational utility. Opponents contended that actual tactile discrimination and lexical decoding were executed entirely within the intact somatosensory and temporal language cortices, rendering the occipital activation behaviorally irrelevant.
Far from retreating, Sadato and the burgeoning cross-modal plasticity community embraced this challenge. The scientific debate catalyzed an ambitious, highly focused global research agenda. To substantiate that the visual cortex was truly executing vital computations, neuroscientists needed to move beyond the correlational framework of hemodynamic neuroimaging and prove functional necessity. Sadato’s 1996 discovery catalyzed a decade of experimental inquiries aimed at establishing the causal relevance of the visual cortex in blindness, ultimately redefining our understanding of neocortical evolvability and developmental epigenetics.
3. Neuroimaging Methodologies: Positron Emission Tomography in Sadato’s Investigations
3.1 H215O Positron Emission Tomography Principles in Plasticity Research
The methodological vehicle that facilitated Sadato’s initial discovery was oxygen-15-labeled water (H215O) positron emission tomography. In an era when functional magnetic resonance imaging (fMRI) was still in its experimental infancy and plagued by severe acoustic noise and technical vulnerabilities to magnetic field inhomogeneities, H215O PET served as the gold standard for quantitative, whole-brain mapping of regional cerebral blood flow. The underlying physiological principle relies on the tight coupling between local neuronal activity, cellular energy metabolism, and arteriolar vasodilation—a phenomenon termed functional hyperemia. As synaptic ensembles within a specialized cortical column fire action potentials and process post-synaptic potentials, the local demand for glucose and oxygen surges, precipitating a localized inflow of oxygenated arterial blood.
The radiotracer H215O possesses unique kinetic properties ideally suited for neuroplasticity paradigms. With a physical half-life of approximately 122 seconds, 15O decays rapidly through positron emission. Upon emission, the positron travels a fractional millimeter before encountering an electron, resulting in a matter-antimatter annihilation event that yields two coincident 511 keV gamma-ray photons traveling at 180 degrees to one another. The circular detector arrays of the PET scanner identify these coincident events, allowing the reconstruction of three-dimensional tracer concentrations throughout the brain parenchyma. Because H215O is a freely diffusible tracer across the blood-brain barrier, the local accumulation of radioactivity within a 40-to-60-second acquisition window serves as a direct, highly linear mathematical proxy for local rCBF.
However, the technological constraints of PET dictated rigorous experimental paradigms. The temporal resolution was inherently coarse, measuring metabolic integration over nearly a minute, while spatial resolution was physically limited by positron range and detector geometry to approximately 6 to 8 millimeters full-width at half-maximum. To isolate fine-grained cognitive processes, Sadato leveraged rigid cognitive subtraction paradigms. By systematically subtracting the digitized rCBF volumes of control states (such as passive fingertip stimulation or smooth surface movement) from active states (such as geometric dot discrimination or active Braille reading), background physiological noise was mathematically subtracted away. Crucially, analytical protocols had to account for and eliminate subtle motion artifacts. Tactile exploration inherently requires steady rhythmic sweeps of the hand and wrist; any transmission of this physical kinetic force up through the kinetic chain to the skull could produce false-positive registration shifts. Sadato mitigated this through customized thermoplastic head-restraints and automated inter-scan image realignments.
3.2 Correlation Between Regional Cerebral Blood Flow and Tactile Performance
A vital dimension of Sadato’s neuroimaging investigations was establishing the quantitative relationship between metabolic demand in the calcarine cortex and behavioral task parameters. Utilizing linear regression models embedded within Statistical Parametric Mapping (SPM) software, Sadato and his team interrogated whether the intensity of rCBF within Area 17 scaled parametrically with the cognitive and physical complexity of the tactile discrimination. Their analyses confirmed a direct linear correlation: as the spatial discrimination tasks grew progressively more difficult—shifting from gross textural discrimination to high-density, geometrically complex embossed dot arrays—the magnitude of rCBF elevation within the occipital pole expanded accordingly.
Comparative analyses across cohorts revealed deeply stratified hemodynamic responses. While early blind subjects exhibited high, scalable metabolic surges in Area 17, late blind subjects (individuals who lost vision after adolescence) demonstrated markedly blunted or functionally variable rCBF responses, often failing to exhibit significant striate activation altogether. Sighted control subjects, conversely, showed a reliable, task-induced decrease in rCBF; the more intensely a sighted person concentrated on deciphering a tactile dot with their finger, the more deeply the primary visual cortex was metabolically silenced. This bidirectional divergence demonstrated that the metabolic response of Area 17 was structurally governed by developmental sensory history.
Furthermore, inter-subject variability within the early blind cohort illuminated a critical developmental metric: the age of Braille acquisition. Regression analyses demonstrated that early blind participants who acquired tactile reading proficiency in early childhood exhibited significantly greater occipital activation than those who learned Braille later in life, even if the onset of blindness was congenital in both groups. Statistical parametric mapping implementations, adapted for these cross-modal cohorts, utilized standardized stereotaxic space (Talairach and Tournoux coordinates) to overlay these functional clusters onto structural neuroanatomy. This proved conclusively that the peak coordinates of tactile-evoked rCBF in early blind individuals resided precisely within the banks of the calcarine sulcus, overlapping identically with the primary visual representations of the fovea and parafovea in sighted populations.
3.3 Subsequent Integration of Functional Magnetic Resonance Imaging (fMRI)
As functional magnetic resonance imaging matured throughout the late 1990s and early 2000s, Sadato and other investigators transitioned from PET to high-field fMRI, seeking to harness the superior spatial resolution and non-invasive, radiation-free nature of blood-oxygen-level-dependent (BOLD) contrast imaging. While PET had successfully established macro-level regional recruitment, fMRI enabled researchers to interrogate the fine-grained functional topography of the reorganized occipital mantle at sub-millimeter scales. Using high-field magnets (3 Tesla and later 7 Tesla), neuroscientists could examine whether the retinotopic architecture that classically defines V1 was preserved, erased, or fundamentally repurposed to serve tactile spatial coordinates.
High-resolution fMRI investigations revealed that the functional architecture of the early blind occipital cortex was far more organized than crude cross-modal spillover could ever account for. By replacing block designs with rapid event-related fMRI paradigms, researchers were able to decouple the transient sensory onset of a tactile touch from the sustained, continuous sensorimotor sweep of a reading finger. These event-related designs proved that V1 responded dynamically to discrete tactile transitions with hemodynamic response functions (HRFs) identical in latency and shape to classical visually evoked BOLD curves observed in sighted controls. The temporal fidelity of BOLD imaging firmly dissociated sensory decoding from general motor readiness or cognitive arousal.
Nevertheless, deploying fMRI to study tactile processing introduced immense technical challenges, most notably the heightened vulnerability to movement artifacts. In high-field magnetic environments, even sub-millimeter head displacements induced by continuous finger movements, wrist rotations, or postural shifts during Braille sweeps can introduce dramatic magnetic susceptibility artifacts, generating false activations along high-contrast cortical boundaries. To conquer these artifacts, Sadato and subsequent laboratories engineered specialized non-ferromagnetic tactile stimulation devices, custom-molded vacuum immobilization beds, and integrated advanced retro-cueing algorithms alongside independent component analysis (ICA) to strip away motion-induced noise, validating the persistent, robust presence of tactile-evoked BOLD responses deep within the calcarine cortex.
4. Demonstrating Functional Causality: Transcranial Magnetic Stimulation Experiments
4.1 The Epiphenomenon Critique in Neuroimaging
Despite the hemodynamic robustness demonstrated by PET and fMRI, functional neuroimaging is fundamentally limited by its correlational nature. Demonstrating that Brodmann Area 17 metabolizes oxygen and glucose during Braille reading does not prove that it is orchestrating the behavior. Throughout the late 1990s, critics championed the epiphenomenon critique: the activation could simply represent an unpruned, vestigial evolutionary spandrel. Because the visual cortex was deafferented, latent poly-synaptic pathways might leak metabolic energy into the occipital lobe without those networks executing any computationally relevant processes.
Under this skeptical framework, the primary somatosensory cortex (S1), secondary somatosensory cortex (S2), and the supramarginal gyrus of the parietal lobe were presumed to perform 100 percent of the perceptual and discriminative computations required to identify tactile characters. The visual cortex was categorized as a mere bystander, passively receiving secondary efference copies or attentional overflow. To dismantle this critique, cognitive neuroscience required an experimental methodology capable of inducing temporary, reversible “virtual lesions” in intact human subjects. If the primary visual cortex was truly an epiphenomenon, disrupting its neural activity would leave Braille reading completely unimpaired; conversely, if Area 17 was executing essential computations, transient disruption would induce catastrophic, quantifiable behavioral performance failures.
This critical empirical test was formulated through the application of Transcranial Magnetic Stimulation (TMS). By placing a focal, figure-eight electromagnetic coil over the scalp, researchers could discharge brief, high-intensity current pulses through the copper windings. These pulses generate localized, transient magnetic fields that penetrate the cranium unimpeded, inducing a targeted electrical current within the underlying cortical tissue. This exogenous electrical discharge depolarizes local neuronal populations simultaneously, disrupting the organized endogenous firing of pyramidal cells and interneurons, thereby creating a focal, non-invasive, millisecond-precise virtual lesion.
4.2 Sadato and Colleagues’ Functional Interference Studies
In a series of landmark studies initiated in the late 1990s—culminating in definitive work led by Sadato’s close collaborator Álvaro Pascual-Leone alongside Sadato and colleagues—focal repetitive and single-pulse TMS was systematically applied to the occipital, parietal, and frontal cortices of early blind subjects and sighted controls while they performed tactile character identification. Blind participants were instructed to scan tactile Braille characters embossed on a motorized drum or static tactile display, verbally identifying the letters while focal TMS pulses were delivered at precise temporal intervals following tactile onset.
The experimental outcomes delivered a decisive blow to the epiphenomenon hypothesis. When focal TMS pulses were discharged over the occipital pole (specifically targeting Brodmann Area 17 and Area 18) of early blind participants, their capacity to identify Braille letters was severely disrupted. Blind subjects committed significant tactile recognition errors, frequently identifying characters incorrectly or reporting complete tactile extinction. Fascinatingly, subjects did not merely fail to identify the characters; they reported that the tactile sensations under their fingertips felt fundamentally degraded, distorted, or completely eradicated. The smooth tactile perception of embossed dots was replaced by an unidentifiable, amorphous tactile sensation.
To establish rigorous anatomical specificity, the researchers contrasted occipital disruption with TMS delivered over the primary somatosensory cortex (S1). While S1 stimulation also caused tactile processing disruption, it manifested at completely distinct temporal windows. By utilizing single-pulse TMS at varying latencies post-stimulus (mapping the chronometry of recruitment), the investigators demonstrated that S1 disruption occurred early (around 20 to 40 milliseconds post-touch), reflecting initial primary thalamocortical somatosensory input. In contrast, occipital TMS maximally disrupted perception at a later window (between 50 and 80 milliseconds post-touch). This chronometric delay mapped the exact time course required for tactile signals to be routed from early somatosensory relays into the visual cortex, proving that occipital recruitment was an indispensable late component of the tactile computational chain.
4.3 Definitive Proof of Occipital Necessity in Tactile Processing
The definitive comparative milestone emerged when this precise TMS protocol was applied to sighted control participants. Sighted individuals were trained to perform basic tactile discrimination of raised dots. When TMS pulses were fired over the somatosensory cortex of sighted subjects, their tactile discrimination was severely degraded, precisely as expected. However, when high-intensity TMS pulses were targeted directly over the occipital pole (Area 17) of these sighted subjects, tactile performance was completely unaffected. Disrupting the primary visual cortex in a sighted brain did nothing to impede its ability to parse tactile inputs at the fingerpads, underscoring that the causal involvement of Area 17 in somatosensation is an exclusive biological hallmark of the reorganized blind brain.
This empirical divergence definitively separated “correlational activation” from “functional necessity.” The primary visual cortex was not merely enjoying a passive metabolic ride; it had become an indispensable, causally essential computational node within the somatosensory reading network. Without an intact, functionally operative calcarine cortex, an early blind individual cannot read Braille. This finding was further corroborated by rare, tragic clinical cases: when an early blind Braille-proficient reader subsequently suffered an ischemic stroke localized strictly to the bilateral occipital lobes, they developed complete tactile Braille agnosia—losing the ability to read Braille despite preserving completely normal peripheral sensation, joint proprioception, and motor control in their hands.
Beyond proving functional necessity, these TMS interference experiments ignited deep questions regarding the neurobiology of perceptual qualia. When an early blind subject’s visual cortex fires to decipher a tactile embossed dot, what is the nature of their conscious subjective experience? Because stimulating Area 17 with single-pulse TMS in sighted individuals evokes phosphenes (flashes of light), researchers asked whether occipital TMS in the blind evoked visual sensations. In early, congenitally blind subjects, occipital TMS never elicited phosphenes; instead, it induced tactile paresthesias, sensation shifts, or functional numbness referred directly to the fingers used for reading. The primary visual cortex had not only adapted its computational machinery to process tactile data, but it had also completely reorganized its phenomenological output: its functional firing now contributed directly to the subjective qualia of touch.
5. Early Versus Late-Onset Blindness: Defining the Critical Periods of Reorganization
5.1 Chronological Boundaries of Occipital Plasticity
The human brain’s capacity to reassign an entire sensory cortex to alternative modalities is not boundless; it is governed by temporal windows of heightened sensitivity termed critical periods. In subsequent investigations, Norihiro Sadato and his contemporaries focused extensively on delineating the chronological boundaries that dictate whether an occipital cortex can successfully execute cross-modal reorganization. By contrasting early blind cohorts (individuals blinded at birth or prior to the age of four) with late-blind cohorts (individuals who suffered total visual deafferentation post-puberty, typically after the age of 14 to 16), profound structural and functional distinctions were unveiled.
When late-blind individuals performed tactile Braille reading or geometric dot discrimination inside PET and fMRI scanners, their hemodynamic profiles diverged radically from those blinded early in life. Late-blind participants consistently failed to recruit primary visual area 17 into the tactile computational loop. Instead of the robust bilateral activation observed in early blind subjects, the striate cortex of late-blind individuals often continued to exhibit the task-induced deactivation characteristic of the sighted brain, or at best, displayed minimal, functionally erratic metabolic shifts. The primary visual cortex in the late-blind brain appeared fundamentally refractory to being repurposed into an active primary somatosensory processor.
These neuroimaging disparities translated into measurable psychophysical consequences. While late-blind individuals can certainly learn to read Braille through extensive training, they rarely achieve the reading velocity, automaticity, or spatial tactile acuity displayed by those who lost their vision congenitally or in early childhood. Quantitative parametric evaluations of Braille reading speeds plotted against the precise chronological age of visual loss demonstrated a clear inflection point: full functional repurposing of Brodmann Area 17 is largely restricted to individuals who suffer profound visual deafferentation prior to the age of roughly 12 to 14 years. Beyond the closure of this developmental window, the primary visual cortex appears largely locked into its unalterable architecture, incapable of forming the dense, functional cross-modal computational networks required for high-velocity tactile reading.
5.2 Synaptic Pruning and Competitive Elimination
The biological mechanisms underpinning this rigid chronological boundary reside in the developmental neurobiology of synaptogenesis, synaptic stabilization, and competitive elimination. During early human post-natal development, the cerebral neocortex undergoes a phase of explosive, exuberant synaptogenesis. In the first year of human life, synaptic density within Brodmann Area 17 surges to nearly 200 percent of adult levels. During this embryonic and infantile window, cortical connectivity is broad and poly-sensory; transient, exuberant cortico-cortical axonal projections link auditory and somatosensory cortices directly to the striate cortex, and subcortical pathways demonstrate marked poly-sensory promiscuity.
Under normal developmental trajectories, the onset of patterned visual experience drives a ruthless process of competitive synaptic pruning. Guided by synchronous, photic retinal activity operating under classical Hebbian plasticity principles (“neurons that fire together, wire together”), retinogeniculate and intracortical visual connections are selectively strengthened and stabilized. Concurrently, non-congruent, silent, or asynchronous non-visual inputs are pruned away through competitive elimination. Axons projecting from somatosensory regions are retracted, and the local inhibitory architecture—primarily governed by parvalbumin-positive GABAergic interneurons—matures, locking the primary visual cortex into an exclusively photic processing network and closing the critical window.
In individuals suffering from congenital or early-onset blindness, however, this competitive landscape is fundamentally inverted. In the complete absence of driven retinal input, the retinogeniculate afferents remain silent, failing to outcompete non-visual sensory inputs. As a direct consequence, the exuberant cross-modal projections arising from somatosensory and auditory structures are never competitively pruned. Instead, as the blind child explores their environment through tactile scanning and acoustic orientation, these alternative sensory inputs become the driving force behind synaptic stabilization within Area 17. The exuberant connections are stabilized, reinforced, and expanded, solidifying an alternative wiring diagram that persists into adulthood. Once the critical period closes, however, mature perineuronal nets (PNNs) wrap around inhibitory interneurons, and molecular plasticity brakes (such as Lynx1 and Nogo receptors) solidify the extracellular matrix, permanently arresting large-scale structural axonal rerouting.
5.3 Residual Plasticity in the Adult Late-Blind Brain
The post-pubertal closure of critical periods does not imply that the adult late-blind brain is entirely devoid of neuroplastic potential. Rather, it indicates that the locus and mechanistic nature of the compensatory reorganization change dramatically. While primary visual cortex (Area 17) remains structurally refractory to functional takeover in late blindness, associative extrastriate visual areas (specifically Brodmann Areas 18, 19, and area MT/V5) retain significant residual plasticity throughout adult life.
Functional neuroimaging investigations in late-blind cohorts repeatedly demonstrate that while Area 17 fails to activate during tactile discrimination, extrastriate visual areas—such as the lateral occipital complex (LOC) and the motion-sensitive MT/V5 complex—display robust cross-modal recruitment. When late-blind subjects read Braille or track moving tactile gratings, MT/V5 fires vigorously, providing a functional substrate for the perception of tactile motion and spatial directionality. Because higher-order associative cortices intrinsically process complex, multi-modal features in the sighted brain, their synaptic networks maintain a state of prolonged epigenetic flexibility, permitting adult-onset sensory inputs to colonize these zones without requiring the extensive primary rewiring prohibited by the closure of V1 critical periods.
Furthermore, late-blind individuals compensate by deploying alternative neurocognitive networks. Rather than relying on early sensory-cortex repurposing, late-blind readers exhibit hyper-activation across frontoparietal attentional networks, secondary somatosensory cortices, and superior parietal regions. These individuals rely heavily on conscious, top-down spatial working memory to reconstruct tactile dot geometries. This mechanistic distinction between early and late blindness fundamentally reshaped the theoretical understanding of critical periods: the question is no longer whether adult sensory cortices can undergo plasticity, but rather which hierarchical levels of the cortical processing stream remain open to reorganization, and what molecular mechanisms constrain the plastic potential of adult primary sensory networks.
6. Neuroarchitectural Mechanisms: Corticocortical Connections Versus Subcortical Rerouting
6.1 The Direct Corticocortical Feedback Model
The realization that the visual cortex executes tactile computations raised a fundamental neuroanatomical question: by what physical pathways do tactile signals travel from the fingertip to the occipital pole? Two primary, competing neuroarchitectural models were advanced to resolve this structural enigma: the direct corticocortical feedback model and the subcortical rerouting model. The direct corticocortical feedback model posits that cross-modal processing is mediated by pre-existing, latent cortico-cortical connections linking primary and secondary somatosensory areas, posterior parietal regions, and extrastriate visual cortices directly to the primary visual cortex.
Anatomical tracer studies in primates have long confirmed that sensory hierarchies are not strictly feedforward; dense networks of reciprocal, feedback, and horizontal connections span across traditionally segregated modalities. In the normally sighted brain, these cross-modal feedback projections are heavily suppressed by powerful GABAergic inhibitory interneuron networks, keeping them functionally silent or restricted to sub-threshold modulatory roles. Following visual deafferentation, however, the sudden loss of primary drive from the optic radiation induces a widespread, rapid downregulation of GABAergic inhibition across the calcarine cortex. This disinhibition unmasks latent corticocortical pathways, allowing somatosensory signals originating in Area 3b, 1, and 2, or routed via the posterior parietal cortex, to flow directly into the striate cortex.
Advanced diffusion tractography and resting-state functional connectivity analyses have provided substantial empirical backing for this model. Researchers have documented heightened functional and structural coherence along the parieto-occipital pathways of early blind individuals. These pathways execute continuous, bi-directional dialogues: tactile spatial signals assembled within the intraparietal sulcus and secondary somatosensory cortices are fed back horizontally into the retinotopic arrays of V1 and V2, which act as a high-resolution computational buffer to dissect the high spatial frequencies of the tactile patterns.
6.2 The Subcortical Rerouting Hypothesis
The alternative, though not mutually exclusive, framework is the subcortical rerouting hypothesis. This model posits that the critical cross-modal divergence occurs far earlier along the ascending sensory pathway, deep within subcortical thalamic relays and brainstem nuclei. In the intact sensory system, somatosensory afferents ascend the dorsal columns of the spinal cord to synapse within the cuneate and gracile nuclei of the medulla, projecting subsequently via the medial lemniscus to the ventral posterolateral (VPL) nucleus of the thalamus, which then projects strictly to S1. Visual inputs, meanwhile, flow from retinal ganglion cells through the optic tracts to synapse in the lateral geniculate nucleus (LGN) before terminating in layer IV of Area 17.
The subcortical rerouting model suggests that under conditions of early visual deprivation, ascending somatosensory projections sprout anomalous axon collaterals, invading the visual thalamus. Under this scheme, somatosensory inputs are hypothesized to divert into the LGN, or alternatively, to pass through the polysensory pulvinar nucleus or the thalamic reticular nucleus. Animal studies in neonatally enucleated rodents and surgically altered ferrets lend credence to this possibility: in classical experiments executed by Mriganka Sur and colleagues, surgically redirecting retinal projections to auditory thalamic nuclei caused the primary auditory cortex to develop visual receptive fields. By extension, researchers hypothesized that in congenital human blindness, tactile or auditory projections might sprout aberrant terminal branches within the deafferented LGN, establishing a subcortical route that ascends via the optic radiations directly to layer IV of the primary visual cortex.
Despite its theoretical elegance, human neuroimaging and post-mortem histology have revealed significant constraints on the subcortical rerouting model. High-resolution magnetic resonance morphometry consistently demonstrates that the lateral geniculate nucleus in early blind humans undergoes severe, profound transneuronal degeneration. Deprived of its massive retinal input (which constitutes over 90 percent of its afferent drive), the human LGN atrophies substantially, shrinking by up to 50 percent in volume with severe structural degeneration across both its magnocellular and parvocellular layers. While the pulvinar nucleus—a structure renowned for its multi-modal, polysensory integrative properties—maintains metabolic integrity and shows enhanced functional coupling to Area 17, the vast majority of current empirical evidence indicates that subcortical rewiring alone cannot account for the rich, high-fidelity tactile processing observed in the blind occipital pole, cementing the predominance of the corticocortical feedback model.
6.3 Long-Term Potentiation and Hebbian Synaptic Remodeling
Whether driven by unmasked corticocortical feedback or subcortical sprouting, the permanent stabilization of cross-modal tactile representations requires long-term, activity-dependent structural remodeling. At the microscopic synaptic level, this transformation is orchestrated through classical Long-Term Potentiation (LTP) and Hebbian synaptic remodeling within the calcarine cortex. When non-visual inputs systematically drive local pyramidal cells in the absence of retinal inputs, N-methyl-D-aspartate (NMDA) receptor complexes are activated, initiating downstream biochemical cascades that drive the insertion of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors into post-synaptic densities.
This persistent synaptic drive profoundly impacts the structural morphology of neurons in Brodmann Area 17. In normal animals and sighted humans, dark-rearing typically results in the widespread collapse and degradation of dendritic spines. In early blind humans who actively practice Braille and tactile exploration, however, post-mortem tissue and structural imaging reveal that dendritic arborization and spine densities within layers II/III and V of the striate cortex are maintained at high, robust levels. The continuous tactile input functions as an effective neurotrophic substitute, preventing the transneuronal synaptic collapse that would otherwise follow absolute sensory isolation.
Crucially, this synaptic stabilization is heavily dependent upon ascending neuromodulatory projections. Diffuse cholinergic inputs originating from the basal forebrain (nucleus basalis of Meynert) and noradrenergic afferents projecting from the locus coeruleus descend upon the occipital cortex, gating local plastic changes. In the early blind brain, every tactile sweep of a Braille character requires intense focal attention, triggering substantial releases of acetylcholine and norepinephrine across the calcarine sulcus. This neuromodulatory influx significantly lowers the threshold for LTP induction, solidifying the synaptic reorganization of the occipital mantle. The current consensus across systems neurobiology accepts a hybridized framework: pre-existing, latent corticocortical pathways provide the initial structural scaffolding, which is subsequently reinforced, pruned, and structurally solidified through lifelong, activity-dependent Hebbian plasticity.
7. Tactile Discrimination and Spatial Representation in the Deafferented Visual Cortex
7.1 Topographical Mapping Beyond Retinotopy
The primary visual cortex of the sighted human brain is famously organized according to a strict retinotopic map. Adjacent points on the retinal surface project to adjacent neuronal columns along the banks of the calcarine sulcus, with central foveal vision disproportionately magnified at the occipital pole and peripheral visual fields mapped along its anterior extent. When early blindness severs this retinal input, does the underlying spatial coordinate framework collapse into an unstructured, diffuse computational array, or does it preserve its topographical geometry to map tactile space?
Electrophysiological recordings and high-resolution fMRI mapping have revealed that the deafferented visual cortex fundamentally retains its intrinsic spatial processing framework. Rather than discarding its topographical architecture, the cortex repurposes it. Spatial tactile exploration requires the continuous decomposition of fine-grained spatial metrics: dot spacing, edge orientation, horizontal alignment, and surface contour. These computational demands are structurally analogous to the spatial frequency, line orientation, and contour extraction algorithms that Area 17 historically evolved to execute for optical images.
Remarkably, researchers have demonstrated that tactile spatial receptive fields within the blind occipital cortex exhibit orientation-selective tuning curves. Neurons in Area 17 fire preferentially when a blind subject scans an embossed grating oriented at specific angles (e.g., 45 degrees versus 90 degrees), mimicking the orientation columns described by Hubel and Wiesel in the visual system. Furthermore, emerging evidence from functional retinotopy paradigms adapted for tactile space indicates that the spatial layout of tactile displays is represented topographically across the occipital cortex. Even the reading hand itself manifests a form of somatotopic mapping: tactile inputs from distinct fingers used during reading display spatially segregated activations within the calcarine sulcus, demonstrating that the cortical coordinate system has translated retinotopy into somatotopically anchored spatial representation.
7.2 Discrimination of Dot Patterns Versus Abstract Tactile Textures
Sadato’s foundational 1996 experiments, along with subsequent parametric neuroimaging investigations, went to great lengths to dissociate the neural mechanisms mediating abstract tactile texture discrimination from those dedicated to geometrically organized dot configurations. This distinction is paramount: tactile textures (such as the roughness of sandpaper or the continuous friction of silk) are predominantly characterized by micro-spatial irregularities that are parsed through high-frequency mechanical vibrations transduced by Pacinian corpuscles and rapidly adapting Meissner’s corpuscles. Conversely, Braille reading relies on the discrete, deterministic spatial arrangement of static embossed dots, transduced primarily by slowly adapting type 1 (SA1) afferents terminating in Merkel-cell neurite complexes.
When early blind subjects interact with purely textured surfaces devoid of geometrical or lexical structure, functional imaging reveals that activation is largely confined to primary and secondary somatosensory cortices (S1 and S2), with minimal or blunted recruitment of the calcarine cortex. However, the moment the task requires the subject to extract spatial geometry—such as identifying the spacing, count, or orientation of raised dots—metabolic activity surges dramatically within Brodmann Area 17 and the Lateral Occipital Complex (LOC). The visual cortex is not simply responding to the general sensation of skin contact; it is selectively engaged to parse the spatial arrangement and geometric configuration of discrete tactile features.
This functional specialization highlights a computational convergence between visual and tactile shape recognition. In the sighted brain, the LOC serves as a critical structural node within the ventral visual pathway, responsible for extracting invariant object shapes and geometric contours from optical scenes. In early blind individuals, the LOC along with ventral occipitotemporal regions are co-opted to execute the exact same computation for touch. The blind reader uses their visual ventral “what” stream to synthesize disconnected cutaneous impressions into a unified mental representation of a geometric shape, demonstrating that these extrastriate areas are fundamentally task-specific computational modules (specialized for shape extraction) rather than modality-specific visual processors.
7.3 Active Tactile Exploration and Sensorimotor Loops
Reading Braille is intrinsically dynamic. It cannot be reduced to static touch; it requires an exquisitely coordinated active sensorimotor loop. The skilled Braillist executes rapid, rhythmic, fluid sweeping movements across the page, tightly regulating downward fingerpad pressure, scanning velocity, and horizontal trajectory. Alterations in scanning velocity directly alter the temporal frequency at which tactile dots deform the mechanoreceptors of the fingertip, requiring instantaneous computational adjustments within the central nervous system.
Functional neuroimaging studies focused on the kinematics of Braille reading have revealed intricate functional coupling between motor execution structures and the reorganized occipital mantle. During active tactile reading, robust functional connectivity is established between the primary motor cortex (M1), the supplementary motor area (SMA), posterior parietal reaching-and-grasping circuits, and Brodmann Area 17. The primary motor cortex sends continuous efference copies of the motor scanning command directly to the parietal and occipital cortices. This allows the visual cortex to anticipate the incoming sensory stream, compensating for fluctuations in sweeping speed and effectively transforming temporal sensory sequences into a coherent spatial map of the Braille line.
Moreover, the mechanics of active reading fundamentally alter sensory gating. In sighted individuals, active motor movement typically induces “sensory attenuation”—a physiological process where the primary somatosensory cortex dampens its responses to self-generated tactile stimuli. In proficient early blind Braillists, however, active scanning bypasses this attenuation; instead, the sensorimotor loop selectively amplifies relevant spatial signals while suppressing non-specific frictional drag. When blind readers utilize both hands simultaneously (a common reading strategy where the left hand tracks the start of a new line while the right hand completes the previous one), interhemispheric communication across the corpus callosum synchronizes the bilateral calcarine cortices, orchestrating a continuous, uninterrupted stream of cross-modal spatial processing.
8. Auditory Localization, Spatial Mapping, and Non-Tactile Cross-Modal Processing
8.1 Echolocation and Spatial Auditory Mapping in the Occipital Lobe
While Sadato’s foundational discoveries focused primarily on tactile Braille processing, subsequent research revealed that the pluripotency of the deafferented occipital cortex extends far beyond somatosensation. Cross-modal reorganization encompasses the auditory modality, demonstrating exceptional functional recruitment during acoustic spatial localization tasks. Early blind individuals frequently demonstrate an extraordinary capacity to map their surrounding physical environment through echolocation: emitting active mouth clicks, cane taps, or tongue snaps, and listening to the faint returning acoustic echoes reflected off walls, doorways, and objects.
Functional neuroimaging investigations of proficient human echolocators, pioneered by researchers such as Melvyn Goodale and Lore Thaler, have yielded astonishing insights. When blind echolocators listen to recorded playbacks of echolocation clicks containing spatial reflections inside an fMRI scanner, their primary visual cortices ignite with intense BOLD activation. In contrast, the auditory cortices of sighted controls process the sound, but their occipital lobes remain metabolically deactivated. Spatial acoustic echoes recruit the deafferented visual cortex to reconstruct the physical dimensions, distance, and shape of the physical space.
This phenomenon illustrates the profound repurposing of the visual dorsal stream. In the sighted brain, the visual dorsal pathway—spanning from V1 through V2 and V3 into the posterior parietal cortex—constitutes the classic “where” or “how” pathway, executing spatial coordinate transformations, motion tracking, and spatial guidance of physical actions. In the early blind individual, this entire dorsal stream is co-opted to process non-visual spatial parameters. Neurons within the occipital pole and parietal cortices display spatial tuning curves that fire selectively in response to binaural acoustic cues (interaural time differences and interaural level differences) and monaural spectral pinna cues. The visual dorsal stream is fundamentally transformed into a modality-independent spatial navigation engine.
8.2 Pitch, Timbre, and Auditory Object Recognition
Beyond spatial localization, non-tactile cross-modal reorganization extensively permeates the acoustic object domain. Early blind individuals frequently demonstrate superior psychophysical performance in pitch discrimination, musical timbre categorization, and auditory scene analysis. When exposed to complex acoustic environments, blind subjects can parse overlapping auditory objects—such as isolating a single conversation within an overwhelming acoustic soundscape—with greater efficiency than sighted peers.
Neuroimaging experiments isolating auditory feature analysis have revealed that this superior acuity is directly mirrored by cross-modal activation within the ventral occipitotemporal cortex. When blind individuals perform complex acoustic pitch discrimination or classify environmental sounds (e.g., distinguishing the sound of a closing door from the sound of an engine), metabolic increases are detected along the lateral extrastriate and inferior temporal zones—regions that historically house the visual ventral “what” stream. The visual cortex actively aids the primary auditory cortex (Heschl’s gyrus) and planum temporale, providing expanded computational surface area to process complex acoustic spectral features.
However, cognitive neuroscientists have uncovered explicit boundaries governing this cross-modal auditory recruitment. While spatial auditory tasks and complex acoustic object categorizations evoke intense occipital activation, passive exposure to simple, unstructured pure tones often fails to elicit robust responses within Area 17. The occipital cortex does not indiscriminately absorb all incoming acoustic signals; rather, it is selectively recruited when the auditory task demands high-level computational parsing—such as spatial metric calculation, fine-grained categorical distinction, or symbolic association. This task-dependent recruitment underscores that the deafferented occipital cortex preserves its identity as a high-level computational processor rather than a generic passive sensory sink.
8.3 Cross-Modal Processing of Olfactory and Gustatory Stimuli
The vast majority of research into cross-modal neuroplasticity has centered on the “spatial” senses: touch and hearing. A fundamental neurobiological question naturally followed: does the deafferented visual cortex also recruit the chemical senses—olfaction and gustation—which inherently lack high spatial resolution? Early behavioral studies indicated that blind individuals consistently outperform sighted controls in odor identification tests, olfactory threshold detection, and the spatial tracking of scent plumes.
Functional neuroimaging studies evaluating olfactory stimulation in early blind populations have yielded intriguing evidence of occipital engagement. When blind participants are exposed to distinct odorants during fMRI scanning, elevated BOLD responses are observed not only within the primary olfactory networks (the piriform cortex, amygdala, and entorhinal cortex) and secondary olfactory regions (the orbitofrontal cortex), but also within associative visual areas (Brodmann Areas 18 and 19) and, to a lesser degree, within the calcarine cortex itself. Olfactory-evoked occipital activity is particularly pronounced when tasks require subjects to identify, name, or mentally visualize the source of the odorant.
Nevertheless, a clear hierarchical hierarchy exists in cross-modal recruitment. The strength, reliability, and spatial extent of occipital activation evoked by chemical senses is markedly lower than the intense, whole-cortex hemodynamic responses elicited by tactile Braille and auditory echolocation. Somatosensory and auditory modalities interface directly with the occipital lobe because their computational demands—spatial mapping, edge extraction, frequency analysis, and symbolic communication—map directly onto the pre-existing columnar machinery of the visual system. Olfaction and gustation, by contrast, lack fine-grained spatial and topological metrics, limiting their functional footprint within the visual cortex. Cross-modal reorganization is thus demonstrated to be structurally constrained by computational compatibility between the sensory inputs and the recipient cortical matrix.
9. Linguistic and Semantic Processing Recruited into Occipital Networks
9.1 Higher-Order Language Processing in Area 17
Perhaps the most conceptually radical extension of Sadato’s initial discovery was the realization that the blind occipital cortex is not merely limited to low-level sensory processing; it is actively recruited into high-level, symbolic, abstract linguistic networks. While the 1996 study demonstrated V1 activation during Braille reading, critics initially attributed this to the sensory mechanics of tactile dot parsing. However, subsequent functional neuroimaging paradigms engineered by researchers such as Marina Bedny and Helen Neville dissociated low-level tactile sensation from high-level linguistic operations, uncovering a profound cognitive transformation.
When early blind subjects engage in purely auditory language tasks—listening to spoken sentences, performing verb-generation tasks in response to acoustic nouns, or evaluating complex syntactic structures—the primary visual cortex (Area 17) and associative visual areas (Areas 18 and 19) light up with robust, left-lateralized activation. Even when no tactile stimulus is present, V1 behaves like a classical perisylvian language area. The visual cortex responds parametrically to linguistic complexity: as sentences increase in syntactic hierarchy and semantic ambiguity (e.g., processing passive-voice, nested relative clauses), metabolic rates within the calcarine sulcus escalate proportionately.
Crucially, transcranial magnetic stimulation confirmed the functional necessity of this linguistic recruitment. When researchers applied focal TMS pulses over the left occipital pole of early blind subjects while they performed auditory verb-generation tasks, the subjects began committing frequent semantic errors or suffered complete speech arrest, identical to the deficits produced when TMS is targeted over Broca’s area in the left inferior frontal gyrus. TMS delivered over the same visual areas in sighted subjects performing identical auditory language tasks produced zero linguistic interference. The deafferented striate cortex of the early blind brain had successfully expanded its computational repertoire, evolving from a primary sensory receptor into an integral node of the central semantic and syntactic language network.
9.2 Working Memory and Executive Control within the Occipital Cortex
The functional colonization of the occipital cortex extends further into the domain of high-order executive function and working memory. Behavioral neuropsychology has documented that blind individuals frequently outperform sighted peers on tests of verbal working memory, such as forward and backward digit spans, serial word recall, and working memory manipulation tasks. To uncover the neural correlates of this enhanced memory bandwidth, neuroimaging studies evaluated early blind participants performing cognitively demanding n-back and serial retrieval paradigms.
These investigations demonstrated that the primary and associative visual cortices are robustly recruited during non-visual working memory maintenance and retrieval. When an early blind subject holds a long sequence of spoken words in short-term memory, Brodmann Area 17 displays sustained, continuous BOLD elevations that persist across the entire delay period of the task. Furthermore, functional connectivity analyses reveal that during these memory operations, the visual cortex enters a state of synchronized functional coupling with the dorsolateral prefrontal cortex (DLPFC) and the anterior cingulate cortex—the traditional command centers of executive control and working memory.
This observation necessitated a major conceptual expansion in cognitive neuroscience. Neuroplasticity in the sensory-deprived brain is not merely “cross-modal sensory plasticity” (swapping touch or sound for vision); it is also “cognitive cross-modal plasticity.” A cortical region historically defined as a low-level, modular sensory cortex is capable of housing high-level abstract cognitive operations, storing linguistic representations, and participating directly in executive working memory loops. The brain reclaims this computationally rich territory, repurposing its billions of uncommitted synapses to augment the overall cognitive and memory processing bandwidth of the individual.
9.3 Linguistic Plasticity Differences Between Braille and Spoken Language
The discovery that Area 17 participates in both tactile Braille reading and auditory spoken language processing spurred investigations into whether these two language forms utilize shared or distinct sub-networks within the reorganized occipital mantle. Tactile Braille requires a complex series of sequential operations: the mechanical sensation of skin deformation must be translated into a spatial geometric dot configuration, which must then be mapped onto an orthographic letter representation, converted into a phonological unit, and ultimately integrated into a lexical-semantic framework. Spoken language, by contrast, bypasses the tactile-to-orthographic conversion, flowing directly from acoustic spectrotemporal patterns into phonological and semantic representations.
High-resolution fMRI investigations designed to dissociate these operational streams have revealed an elegant spatial and functional architecture within the blind occipital lobe. While both tactile Braille and spoken language recruit the occipital cortex, they exhibit distinct functional subdivisions. Spoken language processing predominantly activates lateral and ventral extrastriate regions, displaying strong left-hemisphere lateralization that directly mirrors the classical perisylvian language network. In contrast, tactile Braille reading evokes massive, highly symmetrical bilateral activation of the central calcarine sulcus (Area 17), alongside the left-lateralized ventral linguistic stream.
Neurobiologists have traced this dual recruitment through distinct anatomical routes. The tactile orthographic stream relies heavily on parieto-occipital inputs delivering spatial geometric representations to the calcarine cortex, which then feeds forward into the left ventral visual word form area (VWFA) located within the fusiform gyrus. The spoken language stream, conversely, is mediated through direct cortico-cortical connections linking the superior temporal gyrus, middle temporal gyrus, and inferior frontal gyrus directly to extrastriate visual zones. Under conditions of early sensory divergence, the human language network expands opportunistically, routing diverse sensory pathways into the visual cortex to establish a flexible, multi-modal linguistic architecture unprecedented in sighted neuroanatomy.
10. Structural, Morphological, and Metabolic Alterations in the Occipital Cortex of the Blind
10.1 Cortical Thickness and Morphometric Paradoxes
The profound functional reorganization initiated by visual deprivation is mirrored by profound, paradoxical alterations in the structural and macroscopic morphology of the occipital mantle. When computational neuroanatomists first deployed Voxel-Based Morphometry (VBM) and surface-based cortical thickness algorithms to analyze the brains of early blind individuals, they anticipated finding substantial atrophy. Classical neurological theory predicted that an unutilized primary visual cortex would display reduced gray matter volume and thinned cortical ribbons, reflecting the absence of developmental sensory input.
The empirical findings revealed the exact opposite: a phenomenon now famously recognized as the “cortical thickness paradox.” High-resolution T1-weighted structural MRI revealed that early blind individuals possess a primary visual cortex that is significantly thicker than that of sighted controls. While the striate cortex of a normal sighted adult typically measures approximately 1.8 to 2.2 millimeters in thickness, the calcarine ribbon in early blind individuals regularly exceeds 2.5 to 2.8 millimeters. This structural expansion is largely restricted to primary visual area 17, with associative visual areas displaying variable or near-normal thicknesses.
The biological basis of this paradoxical thickening directly reflects the failure of normal developmental synaptic pruning. In the healthy sighted brain, post-natal visual maturation is characterized by the widespread competitive elimination of redundant synapses, axon terminals, and transient dendritic branches, which drives a progressive developmental thinning of the visual cortex until it reaches a stabilized adult thickness. In early blindness, the absence of patterned retinal input prevents the initiation of this competitive pruning program. Hyper-exuberant dendritic connections and synaptic densities persist into adulthood, leaving the cortical ribbon structurally thicker. In contrast, the underlying subcortical white matter tells an entirely different story: the optic radiations, geniculocalcarine tracts, and posterior thalamocortical white matter pathways display massive volume reductions and profound structural collapse, underscoring that cortical expansion occurs concurrently with subcortical input tract degeneration.
10.2 Microstructural White Matter Tractography
To quantify the microstructural integrity of white matter pathways connecting the reorganized occipital cortex to the rest of the brain, neuroscientists deploy Diffusion Tensor Imaging (DTI) and high-angular-resolution diffusion imaging (HARDI). By measuring the directional diffusivity of water molecules along axonal bundles, DTI provides quantitative metrics such as fractional anisotropy (FA), mean diffusivity (MD), and radial diffusivity (RD), which serve as sensitive in vivo markers of axonal density, myelination caliber, and directional coherence.
DTI investigations in early blind subjects demonstrate extreme, highly compartmentalized microstructural divergence. Within the geniculocalcarine tract (the optic radiation connecting the lateral geniculate nucleus to V1), fractional anisotropy is severely reduced, while radial diffusivity is markedly elevated. This profile provides clear histological evidence of disrupted visual myelination, extensive axonal diameter reductions, and secondary dysmyelination triggered by the lack of ascending retinal action potentials. The structural pipeline that normally feeds optical information into Area 17 is essentially degraded into an empty biological conduit.
Conversely, intra-cortical association tracts that link the occipital cortex to anterior cognitive regions display remarkable preservation, and in some cases, hyper-structural connectivity. The superior longitudinal fasciculus, the inferior fronto-occipital fasciculus (IFOF), and the vertical occipital fasciculus (VOF) demonstrate preserved or elevated fractional anisotropy in early blind readers. Furthermore, the microstructural integrity of these associative white matter tracts correlates positively with individual behavioral metrics: blind individuals displaying higher FA values within the fronto-occipital tracts achieve higher tactile Braille reading speeds and superior phonological processing scores. The white matter architecture of the blind brain actively prioritizes the structural highways linking its repurposed visual processor to the frontal, parietal, and temporal hubs of cognition.
10.3 Resting-State Functional Connectivity and Metabolic Homeostasis
The altered operational status of the blind visual cortex is fundamentally apparent even when the individual is resting quietly in the dark, executing no physical tasks. Resting-State Functional Magnetic Resonance Imaging (rs-fMRI)—which measures spontaneous, ultra-low-frequency (0.01 to 0.1 Hz) fluctuations in the BOLD signal—reveals a profound functional rewiring of resting-state networks in early blind individuals.
In the sighted brain, resting-state analyses reveal that the primary visual cortex is tightly functionally coupled with the primary motor, primary somatosensory, and auditory cortices, forming a coherent sensory-motor functional confederation. In the early blind brain, however, this architecture is fundamentally transformed. Area 17 becomes strikingly decoupled from other primary sensory cortices, reflecting the loss of shared sensory synchronization. Simultaneously, the visual cortex establishes novel, robust hyper-connectivity with the Default Mode Network (DMN) and frontoparietal control networks—systems deeply engaged in internal mentation, semantic retrieval, working memory, and cognitive control.
Metabolically, this functional status is reflected in continuous resting-state energy consumption. Early fluorodeoxyglucose (FDG) PET investigations executed in blind cohorts demonstrated a critical metabolic reality: resting glucose utilization within the deafferented striate cortex is not diminished. Despite receiving zero retinal input, the primary visual cortex of an early blind individual metabolizes glucose and consumes oxygen at rates that equal or even exceed the resting baseline of a sighted brain. The deafferented cortex does not settle into metabolic silence; it maintains a state of continuous, high-energy metabolic homeostasis, keeping its synaptic architecture energized, synchronized, and primed to execute high-demand cross-modal and cognitive computations.
11. Clinical, Neuroprosthetic, and Rehabilitative Implications of Sadato’s Discoveries
11.1 Challenges in Sight Restoration: The Post-Plasticity Dilemma
Sadato’s discoveries regarding cross-modal reorganization and critical periods carry profound, often sobering clinical implications for ophthalmology, neurosurgery, and vision restoration. For centuries, philosophers and scientists—dating back to William Molyneux’s famous 1688 query to John Locke—wondered whether a person born blind, suddenly granted sight as an adult, would be able to immediately recognize and name a sphere and a cube by vision alone. Modern medical advances, including surgical corneal transplants, dense congenital cataract extractions, and retinal gene therapy (such as voretigene neparvovec for RPE65 mutations), have transformed this thought experiment into an urgent clinical reality.
The post-surgical outcomes of individuals who gain sight in adulthood following lifelong early blindness reveal the profound consequences of cross-modal plasticity: the “post-plasticity dilemma.” When optical clarity is successfully restored to the eye, these individuals do not instantly perceive an organized, coherent visual world. Instead, they suffer from severe, debilitating visual agnosia. Patients are overwhelmed by a chaotic, incomprehensible cascade of color, luminance, and motion. They cannot distinguish shapes, visually recognize familiar faces, perceive depth, or even identify common everyday objects that they instantly identify by touch.
The neurobiological explanation for this tragic impairment resides in the functional colonization of Area 17. The primary visual cortex, having spent decades functioning as a tactile and auditory computational processor, is functionally unavailable to process the restored optical input. The synaptic networks, receptive fields, and orientation columns required to parse retinotopic arrays have been completely overwritten by non-visual connections. Furthermore, a fierce cross-modal competition ensues: the entrenched tactile and auditory representations actively resist visual reclamation. Because the developmental critical period for visual acquisition has long closed, the newly restored retinal inputs cannot easily drive synaptic restructuring, leaving the restored visual system functionally disconnected from higher-order semantic understanding.
11.2 Visual Prosthetics and Cortical Stimulation Implants
Parallel challenges confront the frontier of neuroengineering, specifically the design and deployment of visual neuroprosthetics. In cases where the optic nerve is permanently destroyed (such as severe end-stage glaucoma, bilateral optic neuropathy, or physical trauma), retinal implants (like the Argus II) are therapeutically useless. Engineers have consequently bypassed the peripheral visual pathways entirely, developing intracortical microelectrode arrays (such as the Utah array or Orion I) designed to be implanted directly into Brodmann Area 17 to evoke artificial phosphenes via electrical microstimulation.
Sadato’s findings establish rigid, essential selection criteria for candidates seeking these cortical visual prosthetics. If an intracortical array is implanted into the primary visual cortex of an early or congenitally blind individual, microstimulation fails to evoke useful, organized visual phosphenes. Instead, due to the cross-modal reorganization of the underlying tissue, stimulation frequently evokes tactile sensations, diffuse paresthesias, or unstructured, confusing sensations. The neural substrate is no longer configured to translate electrical pulses into a retinotopic perceptual map of the visual field.
Consequently, cortical visual prosthetics are largely restricted to late-blind individuals—patients who possessed normal visual experience throughout their childhood and adolescence and lost their sight late in life. In these individuals, Area 17 retains its dormant, un-reorganized retinotopic architecture, allowing patterned microstimulation to successfully evoke phosphenes that map predictably across the visual field. However, even in late-blind patients, neurorehabilitation algorithms must explicitly account for residual cross-modal plastic drift, implementing adaptive machine-learning calibration interfaces that dynamically adjust stimulation parameters to match the subtle neurochemical and physiological shifts that occur within the visually deafferented human brain.
11.3 Sensory Substitution Technologies (SST)
While invasive sight restoration faces substantial physiological barriers in early blind individuals, Sadato’s work provided the fundamental theoretical foundation for an alternative technological revolution: Sensory Substitution Technologies (SST). Instead of attempting to force visual signals into a reorganized, resistant visual cortex, SST seeks to harness those established cross-modal pathways, utilizing non-invasive algorithms to convert visual images captured by digital cameras into structured, high-resolution tactile or acoustic signals.
Foremost among these systems are tactile-to-visual conversion devices—such as the BrainPort, which translates high-contrast camera feeds into electrical tactile matrices displayed on the user’s tongue—and audio-to-visual algorithmic conversion systems, most notably The vOICe developed by Peter Meijer. The vOICe scans digital images from left to right, translating vertical pixel position into acoustic pitch, horizontal position into time, and brightness into volume. Proficient, long-term users of these devices can achieve remarkable spatial functional competence: they can navigate complex obstacles, identify objects, and locate items across a room purely through acoustic or lingual tactile input.
Functional neuroimaging studies of blind subjects utilizing SST systems have vindicated Sadato’s pluripotency paradigm. When proficient blind users interpret camera feeds via The vOICe or BrainPort, their primary visual cortex (Area 17) and ventral extrastriate networks (the lateral occipital complex) light up with robust, task-dependent BOLD activation. The human brain utilizes its reorganized visual cortex as a high-performance, non-modal digital signal processor. By transforming sensory inputs algorithmically, modern neurorehabilitation capitalizes on the plastic architecture that Sadato first illuminated, empowering blind individuals to reconstruct the physical spatial geometry of their world through non-invasive sensory substitution.
12. Epistemological Shifts in Cognitive Neuroscience: From Modularity to Pluripotency
12.1 Deconstructing the Pluripotent Neocortex
The trajectory of research initiated by Norihiro Sadato’s 1996 discovery catalyzed a profound epistemological transformation across modern cognitive neuroscience. It fundamentally deconstructed the classical, nineteenth-century dogma of rigid neocortical modularity, forcing an intellectual retreat from the view that the neocortex is composed of genetically predetermined, sensory-modality-specific computational modules. In its place arose the paradigm of the pluripotent, task-specific, constructivist brain.
This conceptual revolution was formalized by neuroscientists Álvaro Pascual-Leone and Roy Hamilton through the formulation of the “Metamodal Brain Hypothesis.” The metamodal hypothesis posits that the functional unit of neocortical computation is not organized according to the sensory modality of the input (visual, auditory, tactile), but rather according to the operational or computational task that the cortical region is engineered to execute. Brodmann Area 17, under this view, is not biologically a “visual” cortex; it is an exquisitely specialized, high-resolution spatial processing machine. Its computational architecture is optimized to execute fine-grained spatial discrimination, high spatial frequency analysis, edge extraction, and geometric orientation coordinate tracking.
In the typically developing brain, retinal ganglion cells provide the highest spatial resolution sensory inputs by orders of magnitude, monopolizing this occipital computational machinery and creating the illusion that Area 17 is inherently visual. However, when sight is lost early in life, the cortical machine does not disappear; instead, it accepts the next available high-resolution sensory stream capable of utilizing its spatial processing power: the spatial arrays of tactile Braille or the acoustic reflections of echolocation. The evolutionary advantage of this pluripotency is immense: it ensures that an organism’s most sophisticated, energy-expensive neural real estate never sits idle, dynamically reallocating its computational architecture to meet the specific ecological and environmental challenges confronting the developing individual.
12.2 Consciousness and Qualia in Cross-Modally Reorganized Cortex
Beyond systems neuroanatomy, Sadato’s discoveries penetrated deep into the philosophy of mind, confronting core debates regarding human consciousness, neural correlates of consciousness (NCC), and the nature of subjective phenomenological experience (qualia). If the primary visual cortex is universally acknowledged as the physical engine that generates visual qualia in sighted individuals (giving rise to the conscious subjective experience of light, color, and visual geometry), what happens to this phenomenological output when its inputs are switched to touch?
This debate touches directly on the classic neurophilosophical question of functional role versus physical substrate. Does Brodmann Area 17 generate visual qualia due to its intrinsic anatomical wiring, or are visual qualia an emergent property of the entire retinofugal network? When an early blind individual reads Braille and their calcarine cortex fires, do they experience an abstract tactile impression, or do they experience a cross-modal, phenomenological hybrid? Empirical and introspective investigations indicate that congenitally blind individuals do not experience tactile Braille as flashes of light or visual sensations; their conscious, subjective experience is purely, unmistakably somatosensory and spatial.
This provides an empirical resolution to long-standing philosophical inquiries regarding sensory qualia. The subjective qualia evoked by cortical activity are not dictated strictly by the local cytoarchitecture of the cortical patch itself, but rather by the entire functional network within which that cortical patch is integrated. When Area 17 is functionally wired into the somatosensory and linguistic loops of the brain, its firing contributes directly to the conscious qualia of tactile shape and symbolic meaning. Furthermore, in the context of Molyneux’s problem, these findings illuminate that sensory representation is structurally constructivist: spatial geometry is abstract, amodal, and unified within the central nervous system, capable of reaching conscious awareness through whatever sensory portal the brain successfully integrates into its cortical fabric.
12.3 The Legacy of Norihiro Sadato in Modern Neuroplasticity Research
The historical legacy of Norihiro Sadato within contemporary neuroscience is both monumental and enduring. His 1996 publication in Nature marked the definitive empirical turning point where human neuroplasticity ceased to be viewed as a fringe, exceptional, or pathological phenomenon, and was officially enshrined as an active, fundamental organizing principle of the human central nervous system. Sadato transformed the human primary visual cortex into the ultimate model system for interrogating the genetic, molecular, physiological, and behavioral dynamics of cortical evolvability.
Today, Sadato’s foundational concepts continue to shape cutting-edge research frontiers across multiple scientific disciplines. In computational neuroscience and deep learning, artificial neural network (ANN) architectures are increasingly adopting dynamic, task-based pluripotency, moving away from rigid, pre-programmed sensor-specific modules toward self-organizing, cross-modal computational nodes inspired by the deafferented visual cortex. In developmental biology and neurogenetics, investigators are utilizing transcriptomics and single-cell RNA sequencing to isolate the precise epigenetic triggers and non-coding RNAs that govern the opening and closure of the developmental critical periods that Sadato so precisely mapped.
Ultimately, Norihiro Sadato’s investigations provided a profound, inspiring testament to the dynamic adaptability of the human mind. The sensory-deprived brain does not wither; it adapts, creates, and thrives. In the silent, photic darkness of the early blind brain, the primary visual cortex finds a new, remarkable biological voice—reading the subtle contours of embossed dots, mapping acoustic soundscapes across physical space, and processing the intricate syntactic structures of human language. The visual cortex of the blind stands as a monumental biological testament to the magnificent, unending capacity of the human brain to redefine itself in response to the world it seeks to comprehend.
Conclusion
The discovery of visual cortex reorganization in blind individuals, spearheaded by the visionary neuroimaging and neurophysiological investigations of Norihiro Sadato, fundamentally revolutionized modern neuroscience. By dismantling the classical doctrine of rigid cortical modularity, Sadato demonstrated that primary sensory cortices are fundamentally flexible, capable of extensive functional reassignment in the wake of sensory deprivation. Through the precise deployment of H215O positron emission tomography, transcranial magnetic stimulation, and high-field functional magnetic resonance imaging, Sadato and his contemporaries proved that the occipital cortex in early blind readers is not merely an inactive bystander, but a causally necessary computational engine driving tactile Braille processing, spatial orientation, and complex linguistic comprehension.
These discoveries unveiled the profound developmental rules that govern the human brain, delineating the vital roles of critical periods, synaptic pruning, unmasked corticocortical feedback, and Hebbian plasticity. Simultaneously, this work provided crucial insights into the clinical challenges of sight restoration, the selection criteria for cortical visual prosthetics, and the revolutionary development of sensory substitution technologies. The human visual cortex, stripped of retinal input, does not succumb to passive atrophy; rather, it actively redeploys its formidable spatial computational architecture to elevate the capabilities of the remaining senses. In this profound demonstration of cross-modal neuroplasticity, Norihiro Sadato illuminated the true nature of the human brain: an extraordinary, constructivist, dynamically adaptive organ, capable of transforming its sensory architecture to illuminate the world from within.
References
- Boroojerdi, B., Bushara, K. O., Corwell, B., Immisch, I., Battaglia, F., Muellbacher, W., & Cohen, L. G. (2000). Enhanced excitability of the human visual cortex induced by short-term light deprivation. Cerebral Cortex, 10(5), 529-534. https://doi.org/10.1093/cercor/10.5.529
- Brodmann, K. (1909). Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues. Johann Ambrosius Barth.
- Buchel, C., Price, C., Frackowiak, R. S., & Friston, K. (1998). Different activation patterns in the occipital cortex of late and congenitally blind subjects. Brain, 121(3), 409-419. https://doi.org/10.1093/brain/121.3.409
- Cohen, L. G., Celnik, P., Pascual-Leone, A., Corwell, B., Faiz, L., Dambrosia, J., … & Hallett, M. (1997). Functional relevance of cross-modal plasticity in blind humans. Nature, 389(6647), 180-183. https://doi.org/10.1038/38278
- Hubel, D. H., & Wiesel, T. N. (1970). The period of susceptibility to the physiological effects of unilateral eye closure in kittens. The Journal of Physiology, 206(2), 419-436. https://doi.org/10.1113/jphysiol.1970.sp009022
- Kupers, R., Fumal, A., de Noordhout, A. M., Gjedde, A., Schoenen, J., & Ptito, M. (2006). Transcranial magnetic stimulation of the visual cortex induces somatotopically organized tactile sensations in the blind. Proceedings of the National Academy of Sciences, 103(35), 13256-13260. https://doi.org/10.1073/pnas.0602925103
- Noppeney, U. (2007). The effects of visual deprivation on functional and structural organization of the human brain. Neuroscience & Biobehavioral Reviews, 31(8), 1169-1180. https://doi.org/10.1016/j.neubiorev.2007.04.012
- Pascual-Leone, A., & Hamilton, R. (2001). The metamodal organization of the brain. Progress in Brain Research, 134, 427-445. https://doi.org/10.1016/S0079-6123(01)34028-1
- Pascual-Leone, A., Amedi, A., Fregni, F., & Merabet, L. B. (2005). The plastic human brain cortex. Annual Review of Neuroscience, 28, 377-401. https://doi.org/10.1146/annurev.neuro.27.070203.144216
- Sadato, N., Pascual-Leone, A., Grafman, J., Ibañez, V., Deiber, M. P., Dold, G., & Hallett, M. (1996). Activation of the primary visual cortex by Braille reading in blind subjects. Nature, 380(6574), 526-528. https://doi.org/10.1038/380526a0
- Sadato, N., Pascual-Leone, A., Grafman, J., Deiber, M. P., Ibañez, V., & Hallett, M. (1998). Neural networks for Braille reading by the blind. Brain, 121(7), 1213-1229. https://doi.org/10.1093/brain/121.7.1213
- Sadato, N., Okada, T., Honda, M., & Yonekura, Y. (2002). Critical period for cross-modal plasticity in blind humans: A functional MRI study. NeuroImage, 16(2), 389-400. https://doi.org/10.1006/nimg.2002.1111
- Thaler, L., Arnott, S. R., & Goodale, M. A. (2011). Neural correlates of natural human echolocation in early and late blind echolocators. PLoS ONE, 6(5), e20162. https://doi.org/10.1371/journal.pone.0020162
- Voss, P., Gougoux, F., Zatorre, R. J., Lassonde, M., & Lepore, F. (2008). Differential functional connectivity in early and late blind individuals during dynamic sound localization. NeuroImage, 40(2), 746-758. https://doi.org/10.1016/j.neuroimage.2007.11.054
- Wieser, M. J., & Brosch, T. (2012). Faces in context: A review and systematic categorization of multi-modal face processing. Frontiers in Psychology, 3, 299. https://doi.org/10.3389/fpsyg.2012.00299