RESEARCH DOSSIER
Vestibular Perception and Self-Motion
Abstract. Vestibular perception and self-motion are investigated as the construction of gravity-relative orientation, angular rotation, linear translation, heading, and experienced movement from uncertain biological sensor measurements and their convergence with visual, proprioceptive, and motor-prediction signals. This fundamental critical narrative monograph reconstructs classical experimental vestibular science, examines semicircular-canal and otolith physiology, analyzes the physical tilt–translation ambiguity and its central resolution, and evaluates psychophysical motion thresholds, subjective vertical, visual vection, multisensory heading, reliability-weighted inference, and coordinate transformations. Primary behavioral and macaque-neuroscience experiments are considered alongside computational models, natural self-generated movement, vestibular reafference, route reconstruction, spatial navigation, clinical perceptions of dizziness, and human factors in virtual environments. The synthesis distinguishes sensory encoding, reflexes, conscious judgments, clinical outcomes, and model-based explanations. It identifies experimental and translational priorities in causal neural tests, repeated individual psychophysics, ecological active-motion paradigms, and constrained comparison with artificial sensor-fusion systems. This is a scholarly synthesis of published research, not a registered systematic review, intervention guideline, or original experimental report.
Keywords: vestibular perception; self-motion perception; vestibular system; semicircular canals; otoliths; gravity; heading perception; tilt–translation ambiguity; optic flow; vection; visual–vestibular integration; cue reliability; Bayesian inference; subjective vertical; vestibular psychophysics; galvanic vestibular stimulation; reafference; path integration; motion sickness; persistent postural-perceptual dizziness; Artificial Era
Scientific Scope and Evidence Strategy
Vestibular perception concerns the inference of head and body motion, orientation relative to gravity, and spatial stability from the signals of the inner-ear labyrinth and their integration with vision, proprioception, motor commands, and learned expectations. Self-motion includes angular rotation, linear translation, and changes of orientation; heading is the direction of translation, while subjective vertical is an estimate of gravity-relative orientation. These quantities are related but require different stimuli and response measures. A person can report motion without moving, and can move while perceiving an approximately stable world. The explanandum is therefore the organization of perceptual estimates rather than a single vestibular sensation. Angelaki & Cullen (2008) Cullen (2019)
This monograph is a critical narrative synthesis of historical physiological experiments, human psychophysics, behavioral multisensory studies, primate neural recordings, computational models, clinical observations, and contemporary reviews. Studies are interpreted according to the signals manipulated, the task performed, and the measured outcome. A heading-discrimination threshold, the direction of a galvanically induced illusion, vestibulo-ocular reflex gain, and postural sway quantify different functional products; they cannot be treated as interchangeable indices of perceptual accuracy. The literature includes informative convergences across methods, alongside unresolved differences between tightly controlled passive-motion experiments and voluntary movement in natural settings. Fetsch et al. (2009) Fitzpatrick & Day (2004) Karmali et al. (2014)
The research questions are how physical motion is transformed into afferent activity, how ambiguous cues support estimates of inertial and gravitational motion, how the nervous system integrates visual and nonvisual evidence, and how task, development, pathology, and context affect those estimates. No new human experiment or quantitative meta-analysis is claimed. Where a source reports a particular number, the applicable population and method are retained; explanatory models are evaluated as models. Related psychological topics such as proprioceptive body representation, motion sickness, and spatial memory enter when they clarify vestibular perception, while keeping the scientific object centered on self-motion and orientation. Laurens & Droulez (2007) Lopez & Blanke (2011) Riecke et al. (2023)
The Labyrinth as a Physical Motion Sensor
The vestibular labyrinth contains three semicircular canals and two otolith organs, the utricle and saccule, on each side of the head. Their receptor hair cells transduce mechanical displacement into changes of vestibular nerve activity. Canals are arranged in approximately orthogonal planes and respond to the dynamics of angular head motion; otolith organs register the gravito-inertial force produced by gravity and translational acceleration. The sensory geometry is indispensable for a psychological interpretation because central estimates are reconstructed from measurements made in head-centered coordinates. A motion signal does not by itself specify where the body will travel or how the eyes should move. Angelaki & Cullen (2008) Cullen (2012)
The utricular and saccular maculae differ in orientation and contain hair bundles with varied directional sensitivities. The semicircular canal cupula is displaced by endolymph dynamics when head rotation changes, whereas sustained constant rotation is not signaled indefinitely as an unchanging acceleration. Opponent populations of afferents and the mechanics of the canals shape frequency-dependent response properties. Consequently two physical rotations with identical peak velocity but different time courses can evoke different neural and perceptual responses. Experimental rotation profiles must be specified before a threshold or illusion is compared across studies. Angelaki & Cullen (2008) Fitzpatrick & Day (2004) Cullen (2019)
Vestibular inputs reach brainstem nuclei, cerebellar circuits, ocular and spinal pathways, thalamus, and distributed cortical networks. A single input is transformed differently for gaze stabilization, posture, spatial perception, and action prediction. Early convergence with visual, neck proprioceptive, and efference-copy signals means that an isolated pure vestibular sensation is unusual in everyday experience. This architecture also explains why normal eye reflexes do not guarantee an accurate conscious motion estimate, or why perceptual disturbance can exist even when some peripheral tests are unremarkable. Multimodal organization is fundamental to vestibular psychology from its first central processing stages. Lopez & Blanke (2011) Cullen (2012) Cullen (2019)
From Experimental Vertigo to Vestibular Science
The experimental study of the labyrinth developed through interventions that connected inner-ear anatomy to observable behavior. Flourens investigated semicircular canal damage in animals during the nineteenth century, revealing plane-related disturbances of movement and equilibrium. Later, Breuer, Mach, and Crum Brown articulated hydrodynamic accounts relating canal stimulation to rotation. Their work established a mechanistic question: how does fluid motion in the labyrinth generate directional signals whose consequences include nystagmus and perceived movement? Historical contributions should be read through the instruments and interpretations available to their authors, rather than through later terminology imposed retroactively. Wiest (2015) Angelaki & Cullen (2008)
Robert Bárány's rotation and caloric investigations connected vestibular stimulation, eye movements, and clinical examination. His Nobel lecture records the earlier animal interventions and the distinct lines of reasoning through which nineteenth-century investigators converged on canal function. The reported physiology eventually supported quantitative descriptions of reflex dynamics, yet subjective reports of rotation and equilibrium disturbances retained scientific importance. The history of vestibular inquiry is therefore a joint history of perceptual observation, experimental perturbation, and physiological explanation. The same distinctions remain relevant when a modern laboratory compares neural coding against a two-choice perceptual judgment. Bárány (1916) Wiest (2015)
Twentieth-century research expanded from canals and oculomotor tests toward gravitational orientation, self-motion illusions, multisensory conflict, and navigational behavior. Mittelstaedt's account of subjective vertical proposed an internal combination of gravitational and body-related information. Oman formulated an internal-model account of motion sickness in which sensory discrepancies have dynamic consequences. Later vestibular research introduced high-precision motion platforms, signal detection methods, single-cell recordings, and computational modeling to separate hypotheses that earlier descriptive observations could not distinguish. Historical priority establishes a research lineage; contemporary experiments determine whether the mechanisms inferred from that lineage remain persuasive. Mittelstaedt (1983) Oman (1982) Fetsch et al. (2009)
Angular Motion and Canal Dynamics
Angular self-motion perception begins with a distinction among angular position, velocity, and acceleration. A canal is mechanically sensitive to changes in rotation, and neural processing extends or transforms this information into useful estimates of velocity and orientation. Psychophysical experiments employ controlled yaw, pitch, and roll movements, often in darkness, to measure when an observer can identify the direction of a small rotation. The resulting threshold reflects combined sensory noise, temporal dynamics, decision strategy, and response procedure. It is not a direct single-cell threshold, and it varies across stimulus frequencies and axes. Angelaki & Cullen (2008) Karmali et al. (2014)
In a rotating chair, canal response dynamics and central processing may sustain sensations of movement after physical rotation changes, sometimes creating postrotatory illusions and nystagmus. The concept of velocity storage refers to central processes that prolong and spatially organize rotational signals beyond the short peripheral response. Laurens and Angelaki examined how these dynamics interact with gravity and with natural motion constraints. A perceptual report after rotation must therefore be interpreted against both the imposed kinematics and the observer's changing internal estimate. Such responses challenge any model in which momentary afferent discharge maps one-to-one onto a conscious velocity judgment. Laurens & Angelaki (2011) Bárány (1916)
The experimental distinction between yaw, roll, and pitch is important because gravity interacts differently with the same head-centered canal signal depending on orientation. During combined rotations and head tilts, a person may perceive an altered route or axis of self-motion. An electrical vestibular perturbation, for example, can yield different inferred body rotations as the head is pitched, reflecting transformations from skull coordinates into gravity-related coordinates. This is evidence that perceptual space is constructed by computations over signals and reference frames. Generalization requires attention to stimulus amplitude, baseline orientation, available visual cues, and the response demanded by an experiment. Day & Fitzpatrick (2005) Lopez & Blanke (2011)
Otolith Signals and the Gravito-Inertial Problem
Otolith organs respond to specific force, which combines linear acceleration with the effect of gravity. A forward acceleration and an appropriate backward head tilt can create similar otolith patterns even though the physical causes differ. This tilt–translation ambiguity is a property of the sensor and its physical environment, not an accidental failure of observation. If a motion task relies solely on otolith input, an unambiguous inference about translation or gravity generally requires assumptions about the motion being experienced. Psychology must explain those assumptions and their dynamic interaction with available sensory cues. Angelaki & Yakusheva (2009) Laurens & Droulez (2007)
Semicircular canals supply angular velocity information that can help update an estimate of gravity direction as the head rotates. Comparing the estimated changing direction of gravity with current otolith measurements makes it possible to infer inertial translation. The computation can be described through physical motion equations and realized through distributed vestibular-cerebellar and brainstem processing. Evidence from neural recording does not imply that an observer consciously computes a numerical vector; it shows that neural responses possess signatures compatible with the transformations required for perception and control. Relevant distinctions include afferent versus central representation, steady versus dynamic motion, and head versus body reference frames. Angelaki & Cullen (2008) Angelaki & Yakusheva (2009)
Research on roll tilt and translation is especially informative because the balance of canal and otolith evidence changes with frequency. Low-frequency tilt may depend differently on prior expectations and extravestibular reference cues than faster movements. The psychophysical measurement of orientation in the dark includes uncertainty about head posture, trunk orientation, and the experienced vertical. Models that succeed for small controlled motions may require extensions for large accelerations, unusual head postures, or weightless conditions. The empirical goal is to specify which combination of inputs and assumptions explains responses across these changing conditions. Mittelstaedt (1983) Karmali et al. (2014) Laurens & Droulez (2007)
Tilt–Translation Ambiguity in Neural Circuits
Comparative physiology examines how vestibular neurons represent tilt and translation when canal and otolith stimuli are experimentally combined. Angelaki and Yakusheva summarized responses in vestibular nuclei, cerebellum, and thalamus showing that a peripheral ambiguity can be transformed into central signals appropriate to distinguish motion causes under selected conditions. These findings support the existence of circuitry capable of constructing more informative motion variables from individually ambiguous inputs. The evidence remains tied to the species, recording site, motion axis, and stimulus combinations tested; not every vestibular neuron is expected to carry the same representation. Angelaki & Yakusheva (2009)
A useful computational account must specify its coordinate system and what it represents. A gravity estimate may be maintained in head-centered coordinates while the head rotates, whereas a translation estimate must account for accelerations attributable to movement through space. The canal contribution updates a gravitational prediction; otolith activity constrains the total gravito-inertial signal. Additional proprioceptive and visual information can correct accumulated error. This framing allows prediction of both successful discriminations and specific illusions when sensory cues become inconsistent. It also supports direct tests by changing the relative timing or geometry of tilt and translation. Laurens & Droulez (2007) Angelaki & Cullen (2008)
Neural correlates are most informative when coupled to behavioral discriminations and perturbations. A recorded neuron may respond as predicted by an internal-model equation while serving several motor or autonomic functions. To establish a perceptual mechanism, experiments must connect its activity to judgment, causal manipulation, or carefully matched population-level signatures. Similarly, a model's fit to a limited rotation trajectory leaves open whether a simpler phenomenological filter could fit the same data. Scientific synthesis assigns strongest support to the shared physical constraints and the observed transformations, while treating the precise neural implementation of conscious perception as an ongoing research problem. Cullen (2012) Angelaki & Yakusheva (2009) Cullen (2019)
Gravity, Verticality, and the Upright Body
Subjective visual vertical tasks measure the orientation that an observer judges to align with gravity, commonly by adjusting a visible line. Subjective postural vertical and perceived body tilt probe related but distinct estimates. In ordinary life, an upright observer combines labyrinthine, visual, somatic, and learned body-axis information; in darkness or when environmental structure conflicts with gravity, weights can change. Mittelstaedt's formulation emphasized that apparent verticality can reflect an organized internal estimate rather than a transparent reading of one sensor. A rigorous study therefore specifies whether it measures line orientation, felt body inclination, or balance behavior. Mittelstaedt (1983)
The gravitational reference has consequences for eye movement, route estimation, body orientation, and spatial cognition. Perception of pitch or roll can be biased when the body lies sideways, when a visual scene is tilted, or when visual and somatosensory cues are discordant. The direction and size of a bias depend on experimental geometry and baseline adaptation. A single phrase such as vertical illusion may describe different mechanisms unless the dependent variable is identified. Comparisons across visual-line tasks, posturography, and motion platforms should attend to what each paradigm makes available to the participant. Mittelstaedt (1983) Lopez & Blanke (2011)
Weightlessness exposes the role of gravitational assumptions by altering the normal physical relationship between otolith activity and an external vertical. Reports of unexpected upright or inverted orientation in spaceflight suggest that body-axis biases and learned expectations may become more visible when gravitational information changes. Such reports are informative yet often based on selected participants and unusual conditions; their explanatory value lies in testing the interaction of vestibular and somatic graviception. They should not be generalized mechanically to clinical dizziness or ordinary terrestrial perception. The larger conclusion is that experienced orientation depends on a continuously updated estimate of body, head, and world relations. Mittelstaedt & Glasauer (1993) Laurens & Droulez (2007)
Optic Flow and the Visual Description of Motion
Self-motion through a structured environment changes the pattern of retinal motion, producing optic flow whose expansion, contraction, and rotation can carry information about heading and trajectory. Flow depends on scene depth, eye rotations, moving objects, and the observer's path; it is therefore informative without constituting an infallible description of motion. Vestibular signals supply independent constraints on the head's movement in physical space and can help determine when a retinal pattern originates from self-motion or an independently moving object. Heading judgments that combine these signals operationalize an important aspect of perceptual self-location. Fetsch et al. (2010) Fetsch et al. (2007)
In a dark room, vestibular heading discrimination can be tested through physical translation, while optic flow can simulate travel in a visual display without moving the body. In combined conditions a motion platform and display present compatible or subtly discrepant directions. Psychometric curves summarize the relationship between actual heading and the observer's left-right choices. Their slope represents sensitivity under the chosen model, and their midpoint indicates bias. Because visual motion depends on display structure and vestibular motion depends on platform kinematics, careful temporal alignment and stimulus calibration are essential to causal interpretation. Gu et al. (2008) Fetsch et al. (2009)
Visual motion and vestibular motion have distinct physical origins and are subject to different errors. Optic flow can be degraded by sparse texture, moving traffic, or eye movements; vestibular information becomes weak or ambiguous during slowly changing translations or unusual gravitational circumstances. These differences create a natural test of multisensory inference. When both channels are available, perception can exploit one modality to compensate for uncertainty in the other, although context-dependent biases can persist. A theoretical account must explain the contribution of each modality to the combined judgment, rather than merely showing that two cues activate the same neural region. Morgan et al. (2008) Fetsch et al. (2010)
Vection and Illusory Self-Motion
Vection is a compelling experience of self-motion induced when the body is physically stationary, often by large-field visual movement. The classic observation of apparently moving with a neighboring train illustrates the role of contextual visual motion in attributing movement to the observer rather than the environment. In laboratory studies, investigators measure onset latency, duration, apparent speed, direction, or subjective strength; these are distinct dependent variables. Different display designs and instructions may change which aspect of vection a report reflects. The most informative interpretation specifies the stimulus's relation to eye and vestibular signals. Riecke et al. (2023) Probst et al. (1985)
A rotating optokinetic surround can create circular vection even while the vestibular organs signal little actual angular movement. When visual and real vestibular rotations are combined, experiments can examine which directional estimate dominates and under what acceleration conditions. Probst and colleagues used combined visual, vestibular, and somatosensory inputs to study such ambiguities. These findings established that visually induced motion and physically evoked motion interact in the construction of perceived self-motion. They do not by themselves establish a single fixed hierarchy among senses, because the outcome varies with stimulus dynamics, visual field, task, and the observer's sensory history. Probst et al. (1985) Riecke et al. (2023)
Contemporary vection studies also examine auditory, tactile, biomechanical, and vestibular contributions to an experience often labeled visual. Riecke and colleagues reviewed how such nonvisual information can supplement, weaken, or modify apparent travel. Virtual reality is scientifically useful because it permits controlled manipulation of visual environments while measuring conflicts with actual inertial input. Yet the fidelity of a simulated route cannot be inferred solely from reported immersion. Robust study separates objective physical movement, phenomenological strength of vection, postural adjustment, and symptoms of visually induced discomfort. Riecke et al. (2023) Oman (1982)
Reliability-Weighted Visual–Vestibular Integration
When two measurements estimate the same quantity, a normative model can combine them according to their uncertainty. Under an independence assumption and approximately Gaussian noise, the estimate assigning higher weight to the more precise signal minimizes variance. A psychophysical prediction is improved discrimination in congruent bimodal conditions and systematic shifts toward the more reliable cue when a small conflict is introduced. This model is informative because reliability can be manipulated experimentally, for example by changing the coherence of optic flow while maintaining a constant physical translation. Its assumptions must be tested against actual sensory covariance and observer behavior. Fetsch et al. (2010) Karmali et al. (2014)
Fetsch and colleagues studied human observers and macaques judging heading when visual and vestibular signals were presented separately or together. By varying visual reliability and creating small cue conflicts, they estimated cue weights from perceptual biases. Their 2009 experiments supported rapid reliability-based reweighting in macaques, while also documenting cases in which observers placed greater weight on vestibular evidence than simple optimal models predicted. Such deviations matter theoretically: integration is adaptable but neither perfectly optimal nor fixed in a single modality hierarchy. The experiment directly measures trial-level judgment under controlled conditions, not overall navigation accuracy in an uncontrolled environment. Fetsch et al. (2009)
Karmali and colleagues independently examined frequency-dependent perceptual thresholds for visual and vestibular motion and the benefits of combining cues. Their results reinforce the point that cue reliability is a property of stimulus frequency, method, and observer as well as sensory modality. A universal statement that visual information dominates balance or vestibular information dominates heading is therefore misleading. Weighting can change across tasks and temporal regimes, with individual variation requiring further explanation. Computational models must account for both improved precision and departures from their assumptions, including correlated noise, prior expectations, decision biases, and unmodeled sources of sensory information. Karmali et al. (2014) Fetsch et al. (2010)
Neural Population Codes for Heading
A major experimental challenge is to identify neural populations whose responses can support the statistical combination of visual and vestibular information. Neurons in the dorsal medial superior temporal area of macaque cortex respond to patterns of optic flow and to inertial motion during heading tasks. Gu, Angelaki, and DeAngelis recorded these populations while animals discriminated headings from single cues and cue combinations. Cells whose tuning directions agreed across modalities exhibited improved sensitivity under congruent stimulation, whereas other cells showed different patterns. This establishes the importance of population structure for an account of multisensory perceptual gain. Gu et al. (2008)
The activity of individual neurons provides a mechanistic candidate for the computation but cannot directly reveal subjective experience. Choice probability and neurometric sensitivity connect fluctuations in firing rates with behavioral judgments, yet a positive association may reflect shared inputs or downstream feedback as well as direct causal contribution. In later work, Fetsch and colleagues related macaque population responses to reliability-based cue weighting and found behaviorally meaningful correspondence. The strongest interpretation is that MSTd contains candidate signals capable of contributing to perceptual inference under controlled heading tasks. Causal sufficiency and the role of other cortical and subcortical populations require complementary intervention experiments. Fetsch et al. (2012) Gu et al. (2008)
The cortical organization of vestibular information includes interconnected parietal, insular, temporal, and thalamic territories rather than a single dedicated cortical relay. Neuroanatomical reviews document that vestibular signals are distributed and transformed along several parallel pathways. A perceptual judgment of forward travel, a reflexive eye response, and a conscious judgment of body tilt therefore need not depend on identical neural populations. The relevant neural code must be specified in relation to the movement variable, coordinate frame, active task, and behavioral response. Lopez & Blanke (2011) Cullen (2019)
Reference Frames and Coordinate Transformations
A fundamental obstacle to combining visual and vestibular evidence is that the two modalities initially describe the world in different coordinates. Vestibular acceleration is measured relative to the head; visual optic flow is projected on the retina, where eye position and movement alter the mapping from world movement to sensory displacement. A combined heading estimate requires transformations that incorporate head-on-body position, gaze direction, and gravity-relative orientation. The precise pattern of neuronal tuning can change according to whether a task samples head-centered, eye-centered, or world-related representations. Fetsch et al. (2007) Lopez & Blanke (2011)
Fetsch and colleagues investigated spatial reference frames in macaque MSTd and reported that visual, vestibular, and combined heading responses can use different coordinate organizations. Such results complicate any simplistic picture in which two fixed spatial vectors are merely added. Multisensory integration must also solve correspondence: which retinal motion belongs to a world-stationary surface, which belongs to an independently moving object, and which reflects the movement of the observer? These questions become acute during head turns or when a visual scene contains its own motion. Fetsch et al. (2007) Fetsch et al. (2010)
Coordinate transformation is also expressed in a human experiment using galvanic vestibular stimulation. Day and Fitzpatrick found that the reported illusory body-rotation direction depended systematically on head-pitch angle, consistent with a gravity-related transformation of a head-centered vestibular vector. This method isolates particular computations through predictable changes in orientation, although electrical stimulation activates a complex peripheral input and does not duplicate every property of natural angular acceleration. Across neural and perceptual studies, the common requirement is a representation of motion that remains useful when the head, eyes, and body occupy changing configurations. Day & Fitzpatrick (2005) Fitzpatrick & Day (2004)
Cue Conflict, Causal Inference, and Perceptual Attribution
When visual and inertial cues disagree, the observer must decide how much they refer to the same event. A small discrepancy may be reconciled by weighted fusion, whereas a large discrepancy may favor separate causes, reinterpreted object movement, or increased uncertainty. The experimental question is not merely which sense wins, but what evidence the observer treats as belonging to self-motion. Causal-inference models offer a framework for distinguishing cue combination from cue segregation, although a successful fit does not uniquely identify the implemented neural algorithm. Vestibular studies can test the framework by independently manipulating conflict magnitude, reliability, and temporal coincidence. Fetsch et al. (2010) Riecke et al. (2023)
A moving object adds a particularly revealing complication to optic flow. Its motion introduces image displacement that can bias the apparent heading of an observer. Vestibular measurements can help partition this flow into components consistent with self-motion and those attributable to the object. Neural and behavioral work on multisensory heading supports the value of independent inertial evidence in such situations, but direct inference about complex real-world navigation requires experiments with richer scenes and natural movement. What appears as noise in one laboratory paradigm may serve as a meaningful signal about environmental dynamics in another. Fetsch et al. (2007) Cullen (2019)
Cue conflict also has consequences for comfort, postural control, and the sense of a stable world. A mismatch between expected and observed signals can alter subjective motion without producing an objectively incorrect judgment by a single modality. Therefore experiments should separately measure bias, discrimination precision, symptom reports, and response latency. Theoretical progress depends on identifying conditions under which models make distinct predictions, particularly when fusion, switching, adaptation, and prior expectations can otherwise explain similar behavioral averages. Qualitative descriptions of conflict are the beginning of a model, not its validation. Oman (1982) Fetsch et al. (2009)
Bayesian Accounts and Their Testable Assumptions
Bayesian treatments of vestibular inference combine likelihood functions representing uncertain sensory evidence with priors representing expected motion or orientation. Laurens and Droulez developed a formal account of vestibular information processing in which the interpretation of ambiguous gravito-inertial signals depends on the physical dynamics and probabilistic assumptions of the observer. Such models predict systematic biases as well as precision, providing a way to connect physical sensory limitations with apparently irrational illusions. They also make explicit the distinction between a statistically optimal computation under specified assumptions and a claim of universally accurate human behavior. Laurens & Droulez (2007)
The uncertainty assigned to a cue can be estimated from psychometric functions, but an observed threshold depends on the noise model and psychophysical procedure. Priors and likelihood widths may trade off in a fit, creating questions of identifiability. Model comparison should involve held-out stimuli, distinct motion frequencies, altered baseline orientation, and direct manipulation of cue reliability. A fitted curve from one tightly constrained discrimination task establishes compatibility with a model rather than unique proof of a neural Bayesian algorithm. Successful predictions across conditions are more demanding evidence than retrospective descriptions of known illusions. Karmali et al. (2014) Fetsch et al. (2009)
Reliability-weighted integration, the compensation of gravitational ambiguity, and active-movement prediction are compatible with a broad family of probabilistic state-estimation models. Yet the relevant task differs across perception and motor control. A reflex might prioritize rapid stabilization, while an explicit judgment incorporates decision uncertainty and prior beliefs. The scientific use of Bayesian vocabulary is strongest when a model specifies its states, observations, dynamics, priors, likelihoods, and falsifiable predictions. If assumptions change between conditions, those changes must be justified rather than concealed in adjustable parameters. Laurens & Droulez (2007) Cullen (2012) Fetsch et al. (2012)
Psychophysical Thresholds and Motion Platforms
Vestibular psychophysics measures how reliably observers discriminate motion directions, amplitudes, or orientations under specified conditions. In a typical direction-recognition task a motorized platform delivers small controlled movements, and a participant reports whether rotation or translation was toward one side. Repeated trials at multiple stimulus levels permit the estimation of a psychometric function. Its location can represent systematic directional bias while its width summarizes discrimination uncertainty. Because head support, auditory cues, seat vibration, and response expectations may reveal the direction independently of vestibular sensation, experimental apparatus and control trials are integral parts of construct validity. Karmali et al. (2014) Mardirossian et al. (2014)
Small-movement thresholds depend on the trajectory and frequency of stimulation, the axis tested, whether an experiment measures detection or direction, and whether vestibular input is combined with vision. Karmali and colleagues reported variation in visual and vestibular precision across frequencies, making a fixed universal sensitivity value inappropriate. Mardirossian and colleagues examined roll-tilt perceptual thresholds across different cue conditions and illustrated variation between observers. A comparison between studies is meaningful only after translating the dependent variables into common physical units and matching relevant stimulus geometry. Karmali et al. (2014) Mardirossian et al. (2014)
Adaptive staircases and Bayesian psychometric estimators may reduce the number of trials required to measure a threshold, but they can also introduce design-specific dependencies. Lapses, response bias, fatigue, learning, and head restraint should be considered explicitly. Repeated measurements are valuable because between-person differences could reflect stable sensory properties, temporary fluctuations, or measurement noise. The correct scientific conclusion is conditional: a reported threshold characterizes performance for a given task and protocol. Mechanistic interpretation requires additional evidence linking that performance to vestibular pathways and multisensory estimates. Karmali et al. (2014) Cullen (2019)
Perception, Reflexes, and Separate Outcome Variables
The vestibulo-ocular reflex stabilizes images during head movement and can be measured objectively through eye kinematics. It is an important indicator of vestibular function, yet it is not identical to perceived rotation or translation. Ocular responses follow sensorimotor transformations optimized for gaze, whereas perceptual judgments involve estimation, decision, attention, and response. Clinically normal reflex measurements do not automatically imply that every perceptual computation is normal. Conversely, experimentally altered motion perception may coexist with useful reflexive stabilization. Vestibular research gains accuracy when its dependent variables are separated before drawing an inference about sensory function. Cullen (2012) Demer (1995)
Caloric tests, video head impulse testing, rotation testing, vestibular-evoked myogenic potentials, and perceptual thresholds sample overlapping but differentiated functions. Their sensitivity to canals, otolith pathways, stimulus frequencies, and central transformations differs. The question of what a test detects must precede the interpretation of its result. Experimental galvanic vestibular stimulation adds another form of perturbation, supplying a controlled electrical input that can change posture and motion judgments. It creates a useful causal manipulation while requiring caution because the imposed electrical signal differs from natural motion-driven receptor mechanics. Fitzpatrick & Day (2004) Wardman et al. (2003)
A strong multimethod design may combine eye movements, subjective direction judgments, postural measurements, and physiological recordings during the same controlled motion profile. Agreement strengthens the hypothesis of a shared input, while divergence can reveal pathway-specific computations rather than simple measurement failure. Individual symptoms introduce further dimensions: discomfort, perceived motion, instability, and fear can change together or separately. The scientific task is to establish which mechanism predicts each outcome and whether its effects are replicated across laboratories and populations. This principle also limits the transfer of animal neural data directly into claims about human conscious experience. Cullen (2019) Staab et al. (2017) Lopez & Blanke (2011)
Active Motion and Vestibular Reafference
Self-generated movement activates the vestibular sensors just as externally imposed movement does, but its consequences are partly predictable from motor commands. Reafference denotes sensory input resulting from one's own actions; exafference denotes externally imposed or unexpected input. A system that treated both indiscriminately would have difficulty distinguishing intended head turns from perturbations requiring correction. Experiments in alert macaques have compared active and passive head movement to determine where the distinction emerges in vestibular pathways. Primary vestibular afferents carry motion-related activity, while particular central neurons show marked context dependence. Cullen et al. (2009) Cullen et al. (2011)
Cullen and colleagues found that some vestibular nucleus neurons attenuate responses to self-generated movement when predicted proprioceptive feedback corresponds to the actual consequences of a motor command. This attenuation is selective rather than a global suppression of all vestibular activity, and externally imposed components of motion remain represented. The result fits an internal forward-model account in which predicted sensory consequences are compared against actual afferent input. The evidence is causal in the sense that task conditions and movement sources are manipulated, but the proposed precise internal circuitry must still be tested with targeted interventions. Cullen et al. (2011) Cullen (2019)
The distinction matters for human perception because everyday motion occurs while the observer walks, turns, reaches, and moves the eyes. Passive laboratory translation removes many predictive and proprioceptive components of naturally generated movement. A model fitted only to a seated subject may therefore estimate how the system responds under one constrained action policy rather than describe all self-motion experience. Generalizing requires comparing matched active and passive kinematics, controlling efference copy and neck feedback, and observing whether motion bias and sensitivity change systematically. Such studies can explain why moving intentionally often feels stable while an unexpected displacement is salient. Cullen et al. (2009) Cullen (2019)
Neck Proprioception and the Organization of the Moving Body
The vestibular organs measure the head's motion in space, while the body's route depends on how the head moves relative to the trunk. Neck proprioception provides information about this relation, allowing head-centered inputs to contribute to body-centered control and orientation. A person turning the head while walking cannot reconstruct the body path from the labyrinth alone without information about head-on-body motion. Vestibular and proprioceptive convergence is therefore a central component of self-motion inference rather than a secondary correction added after perception has been completed. Angelaki & Cullen (2008) Cullen (2012)
Electrical vestibular stimulation illustrates this principle under controlled conditions. Day and Fitzpatrick elicited perceived route changes that depended on head pitch even though the stimulation originated in head-centered vestibular pathways. The spatial relation between skull and gravity determined how the signal was interpreted in terms of body movement. During active motion, expected neck feedback can additionally modify central vestibular responses. The interaction of transformation and prediction provides a mechanistic framework linking neck proprioception to motion perception. Day & Fitzpatrick (2005) Cullen et al. (2011)
The boundary between this topic and proprioceptive body representation is scientifically useful. Proprioception is often studied through limb-position matching, tendon vibration, and body ownership, whereas vestibular perception concentrates on gravity, head dynamics, heading, and spatial orientation. Both systems contribute to embodied self-location, and their convergence is a legitimate focus of experimental research. Research designs that remove or distort one input can estimate its marginal contribution, but sensory reweighting means the resulting estimate also depends on the availability of other information. A single deprivation condition rarely reveals a context-free absolute contribution. Lopez & Blanke (2011) Angelaki & Cullen (2008)
Locomotion and the Perceived Travelled Route
Walking involves a changing relation among expected motor commands, sensory feedback, bodily orientation, and environmental landmarks. A vestibular signal constrains turning and acceleration, but accumulated position depends on integrating movement over time and accounting for the orientation of the head. Errors in this integration can alter the apparent location of a previously visited point. Controlled route-reconstruction tasks ask observers to indicate their starting point or reproduce their trajectory, providing a behavioral measure that connects short-lived motion sensations with longer-term spatial estimation. Day & Fitzpatrick (2005) Cullen (2019)
Galvanic stimulation has been used to produce systematic deviations of walking trajectory and perceived travelled path. Such results are informative precisely because motion perception and motor behavior can diverge: the electrical input may modify inferred orientation, postural control, and movement itself. An observed walking deviation alone cannot establish the subjective direction or the cortical origin of the perceived change. Designs that collect independent pointing or route reports help disambiguate these pathways. They also emphasize why path integration should be studied with controlled landmarks, comparable routes, and explicit movement histories. Fitzpatrick & Day (2004) Wardman et al. (2003)
Naturalistic experiments must also account for rapid eye movements, gaze stabilization, head motion relative to the trunk, and ongoing visual sampling. A walking observer inhabits a multisensory stream whose causal structure differs from a passive platform translation. Models built from instantaneous heading estimates need mechanisms for temporal accumulation, error correction, and reorientation when landmarks reappear. The importance of vestibular information is consequently both momentary and sequential: it affects judgments about current movement and the maintenance of an estimate of where the observer has traveled. Cullen (2019) Brandt et al. (2005)
Vestibular Contributions to Spatial Memory
Spatial navigation depends on representations of direction, route, and location that extend beyond immediate sensations of motion. Vestibular inputs contribute information about rotations and accelerations that can support such representations during movement. Brandt and colleagues examined a small clinical group with chronic bilateral vestibular loss, reporting differences in hippocampal volume and performance on a virtual spatial navigation task relative to controls. These observations are notable because the patients performed the virtual task without corresponding real-world self-motion, suggesting consequences of vestibular history for spatial processing. The study's observational design and limited sample constrain causal conclusions. Brandt et al. (2005)
The hippocampal result should be interpreted alongside animal experiments and broader neural accounts of vestibular–spatial interaction. A correlation between vestibular loss, regional anatomy, and task performance can arise through multiple pathways, including altered navigation experience and compensatory behavior. It does not show that a single perceptual signal uniquely causes hippocampal change. The stronger research program combines longitudinal measurement, carefully defined sensory loss, comparable vestibular and nonvestibular controls, and tasks that distinguish route learning from spatial memory and decision strategy. Brandt et al. (2005) Lopez & Blanke (2011)
For self-motion psychology, navigation evidence establishes the temporal depth of vestibular processing. The nervous system must use current motion estimates to maintain spatial beliefs when landmarks disappear and correct accumulated error when reliable cues return. These operations may occur without a conscious feeling corresponding to each calculation. Vestibular perception, spatial memory, and map-based reasoning therefore overlap functionally while remaining separable constructs. A complete account of route perception must specify which stage is being measured: immediate directional discrimination, accumulated displacement, landmark-based reorientation, or later memory for traveled space. Cullen (2019) Brandt et al. (2005)
Postural Control and Perceived Orientation
Balance involves coordinated estimates of the body's position relative to gravity and the support surface. The contribution of vestibular information changes with stance, availability of vision, surface compliance, and the task's stability demands. Posturography can quantify sway or responses to controlled perturbation, but sway is a motor outcome resulting from multiple feedback loops. It must be distinguished from what the participant reports experiencing. Combining perceived vertical, motion judgments, and postural measures provides a better test of whether a proposed orientation estimate supports multiple behavioral outputs. Bronstein (1995) Fitzpatrick & Day (2004)
Galvanic vestibular stimulation is particularly useful in this domain because it can alter posture in predictable directions. Wardman, Day, and Fitzpatrick examined temporal components of postural responses under controlled electrical stimulation, illustrating that vestibularly evoked behavior contains multiple dynamic processes. However, the direction of postural sway need not equal the direction of an explicit motion illusion. A stimulus that changes perceived reference orientation can generate compensatory action, directly recruit postural pathways, or do both. Experimental interpretations require simultaneous or separate measures of these possibilities. Wardman et al. (2003) Day & Fitzpatrick (2005)
Visual dependence may become especially evident when full-field motion destabilizes a person who otherwise stands adequately under static viewing conditions. Bronstein's observations of people with visual vertigo connected visually induced symptoms with heterogeneous vestibular, central, and oculomotor histories. Such findings argue for attention to patient subgroup and stimulus configuration rather than reducing balance to a single sensory weighting parameter. The psychological importance of posture lies in the interaction between perceived upright, expected action outcomes, and behavioral stabilization. Bronstein (1995) Staab et al. (2017)
Sensory Conflict and Motion Sickness
Motion sickness offers a consequential test of vestibular inference because symptoms often emerge when sensory information about movement is unusual or inconsistent with expectations. Oman developed a mathematical internal-model account in which sensory conflict is generated during the central estimation of orientation and movement, with dynamically evolving consequences for discomfort. The framework links perception to prediction, adaptation, and bodily responses without assuming that every self-motion illusion is accompanied by illness. Experimental outcomes must distinguish nausea, motion discomfort, postural instability, and perceived motion; each may follow a different time course and depend on prior exposure. Oman (1982)
Visual simulations can create a vivid sensation of travel while the otolith organs register no corresponding acceleration. Conversely, movement in a vehicle may activate vestibular pathways while the available visual scene fails to specify the same trajectory. These configurations reveal sensory contingencies that normal locomotion keeps largely consistent. However, describing sensory conflict as a cause of all symptoms is insufficient without specifying the expected sensory state, the source of its estimate, and the mechanisms linking discrepancy to autonomic or behavioral responses. A model can be confronted with manipulations of head orientation, visual dynamics, and controlled inertial movement. Oman (1982) Riecke et al. (2023)
Repeated exposure can change discomfort and response, but behavioral habituation, sensory recalibration, and changes in task expectancy are empirically distinct possibilities. Research programs should measure them independently and describe when a simulator or laboratory protocol generalizes to cars, boats, aircraft, or head-mounted displays. Susceptibility also varies between people and situations, potentially changing both the expression of symptoms and their attribution. The central scientific point is that perceptual self-motion and bodily adaptation belong to an integrated process whose components must be separated analytically before their relations can be established experimentally. Oman (1982) Cullen (2019)
Human Factors, Simulation, and Virtual Environments
Flight simulation, vehicle design, and immersive displays manipulate the coupling among optic flow, inertial cues, and motor behavior. A simulation may reproduce visual heading while omitting gravitational or translational acceleration, or a motion platform may approximate some inertial components within mechanical limits. Its perceptual fidelity consequently depends on the task and the sensory dimensions being evaluated. A system that supports accurate directional judgment may still produce weak vection or considerable discomfort, while compelling vection may occur without accurate path reconstruction. These possibilities are central to human-factors experiments and require specific outcome measures. Riecke et al. (2023) Fetsch et al. (2010)
The evidence on multisensory cue combination suggests that adding consistent inertial input can improve a controlled heading judgment when it offers reliable independent information. Yet equipment specifications and psychophysical results must be matched: the latency, frequency response, range of platform movement, visual field of view, and head-tracking behavior can change cue congruency. A visually detailed scene is not equivalent to a physically accurate multisensory experience. Engineering evaluations should therefore connect measurable display and motion properties to experimentally observed heading bias, vection, spatial orientation, and symptom responses. Karmali et al. (2014) Fetsch et al. (2009)
Naturalistic head movement creates further demands. If a virtual environment responds to a user's head turn, visual flow and vestibular activity should exhibit the relation expected from active movement. Delays or inaccurate transformations can disrupt this relation even if each component looks plausible when presented alone. Vestibular reafference models explain why action contingency matters and why passive demonstration cannot fully predict interactive performance. A rigorous research program combines controlled psychophysics with active locomotion or head-turn tasks, reports individual variation, and evaluates ecological transfer rather than relying only on immersion ratings. Cullen et al. (2011) Cullen (2019) Riecke et al. (2023)
Clinical Disturbances of Perceived Motion
Disorders of vestibular function can produce spinning vertigo, disequilibrium, visually provoked dizziness, and difficulties with spatial orientation. These phenomena differ in their causes and should be related to objective sensory findings and to the specific perceptual complaints reported. Bronstein described a heterogeneous group experiencing visual vertigo, including peripheral and central conditions, with variable responses to full-field visual motion. The findings point toward altered sensory dependence in selected patients while underscoring that visual discomfort is not a diagnosis of one peripheral deficit. Clinical interpretation requires an appropriate differential examination and time course. Bronstein (1995)
Persistent postural-perceptual dizziness, or PPPD, has published consensus criteria established by the Bárány Society. Its core clinical pattern concerns persistent dizziness, unsteadiness, or nonspinning vertigo, commonly aggravated by upright posture, motion, and complex visual scenes. The consensus characterizes PPPD as a chronic functional vestibular disorder and acknowledges coexistence with other precipitating conditions. The diagnosis depends on clinical assessment rather than a single vestibular reflex result or a psychological questionnaire alone. Its mechanisms continue to be studied through multisensory processing, threat-related expectations, and postural control. Staab et al. (2017)
The relation between self-motion psychophysics and clinical diagnosis remains an active translational question. A group difference in a motion threshold or visual-dependence test can suggest a mechanism without proving a diagnostic biomarker for an individual patient. Small samples, medication or comorbidity differences, selection bias, and task learning can alter observed effects. The appropriate synthesis is to use controlled perception studies to generate and constrain mechanistic hypotheses while applying clinical classifications according to their actual criteria. This monograph offers scientific analysis and does not prescribe personal medical interventions. Staab et al. (2017) Mardirossian et al. (2014) Bronstein (1995)
Adaptation, Compensation, and Development
Vestibular estimates are shaped by the changing relation between sensory evidence and successful action. After altered input, the organism may learn different weights, use new strategies, or depend more heavily on visual and somatosensory information. These possibilities can produce similar average performance, so compensation requires evidence concerning the processes that changed. A return to acceptable walking performance does not imply complete restoration of every peripheral signal or threshold. Investigators should compare perceptual judgments, reflexes, and functional behavior over repeated sessions and identify the sensory context in which the apparent recovery occurs. Cullen (2012) Cullen (2019)
Developmental and aging differences complicate the interpretation of laboratory vestibular results. Changes in sensory precision, locomotor experience, attention, and confidence may influence motion judgments, while selection effects can make older or clinically affected groups unlike younger controls in more than vestibular function. Research questions should distinguish maturation of the peripheral sensors from learning to integrate and interpret their signals. Age, motor experience, visual acuity, and task familiarity must be measured where relevant. The same experimental environment may supply very different effective sensory information to two participants. Karmali et al. (2014) Riecke et al. (2023)
Adaptation is not unrestricted flexibility. A sensor's physical ambiguity persists even as the brain develops better priors and additional sources of information become available. Likewise, a clinical compensation strategy may work in a predictable room while failing in a busy visual environment. The research opportunity is to identify the conditions under which recalibration transfers between environments, motion frequencies, and active versus passive movement. Longitudinal designs with reliable within-person measures would provide stronger evidence about individual change than cross-sectional comparisons alone. Laurens & Droulez (2007) Staab et al. (2017) Bronstein (1995)
Evidence Quality, Replication, and Model Discrimination
Vestibular science combines unusually strong physical control of stimuli with difficult measurement of conscious perception. Motion platforms can specify angular and linear trajectories precisely, but subtle vibration, audition, anticipation, or restrained posture can introduce unwanted cues. Neural studies permit direct observation of defined populations in nonhuman animals, but their results depend on task training and species-specific recording access. Clinical observations identify meaningful impairments but often include heterogeneous histories and multiple concurrent mechanisms. Evidence synthesis therefore requires a common vocabulary for physical stimulus, sensory mechanism, response variable, population, and uncertainty. Gu et al. (2008) Brandt et al. (2005) Karmali et al. (2014)
Many influential findings establish that visual and vestibular cues interact, yet competing accounts may predict similar cue weights or biases over a limited stimulus range. Reliability-weighted fusion, prior-driven estimation, causal segregation, and adaptive control must be distinguished through experiments that vary frequency, delay, reliability, and conflict independently. A robust study reports psychometric function fitting, uncertainty intervals, exclusions, repeated observations, and methods for controlling nonvestibular information. Replication should test the explanatory parameterization and not merely reproduce one qualitative direction of effect. Scientific progress depends on models that survive altered circumstances rather than on a single convincing simulation. Fetsch et al. (2009) Fetsch et al. (2012) Laurens & Droulez (2007)
Claims about practical significance additionally require an outcome beyond laboratory discrimination. An improved heading threshold may not translate into fewer navigation errors, safer vehicle operation, or reduced discomfort without direct tests of those endpoints. Likewise, associations between vestibular loss and spatial memory cannot alone specify whether the relation is causal, developmental, or mediated by experience. Cautious inference is particularly important for proposed biomarkers and for subjective phenomena such as vection whose rating methods vary. Independent replication, cross-method convergence, transparent analysis, and longitudinal designs are priorities for the next phase of the field. Brandt et al. (2005) Staab et al. (2017) Riecke et al. (2023)
Artificial Era and Computational Research
Vestibular perception provides an instructive scientific model for how an embodied system can infer hidden world states from incomplete sensor measurements. An artificial agent equipped with inertial sensors faces analogous engineering ambiguities among acceleration, gravity, coordinate frames, and motion dynamics. In robotics, estimation may combine accelerometers, gyroscopes, visual flow, proprioceptive encoders, and state predictions. Such functional analogies can inform the Psychology of Artificial as an interdisciplinary research direction when the sensors, algorithms, and evaluation conditions are made explicit. They support comparisons of inference architecture without requiring conclusions about machine subjective experience. Laurens & Droulez (2007) Angelaki & Cullen (2008)
A useful comparative research program asks whether an artificial motion-estimation system and a biological vestibular system respond similarly to carefully matched stimulus ambiguities. Similar output can arise through different mechanisms; inspection of internal representations and controlled perturbation therefore matters. Biological findings about active versus passive self-motion motivate tests of prediction-error handling in artificial agents, while human multisensory psychophysics suggests benchmarks in which sensor reliability and cue conflict vary systematically. Comparisons should be reported as operational performance and model equivalence under specified tasks. Cullen (2019) Fetsch et al. (2009)
Human–Artificial Relations introduce a related practical question: how motion simulations, mobile robots, or adaptive interfaces alter the human user's spatial orientation and motion expectations. A human interacting with a visually moving virtual environment experiences consequences of that artificial environment through biological sensory pathways. This relation can be studied empirically through vection, balance, task performance, discomfort, and changes in agency judgments. The Artificial Era supplies a broader research setting for designing systems that respect perceptual constraints; its philosophical interpretation and its experimentally testable claims should remain explicitly distinguishable. Riecke et al. (2023) Fitzpatrick & Day (2004) Cullen (2019)
Conclusion
Vestibular perception is the dynamic construction of estimates of self-motion and orientation from peripheral inertial signals and their convergence with vision, proprioception, and prediction. Semicircular canals and otoliths provide essential measurements whose physical properties determine both perceptual possibilities and unavoidable ambiguities. Central computations integrate angular motion with gravito-inertial information, transform signals across reference frames, and coordinate perceptual judgments with gaze and posture. The strongest evidence arises when physics, behavioral psychophysics, and neural recording constrain the same model rather than when one method is treated as a sufficient account of perception. Angelaki & Cullen (2008) Angelaki & Yakusheva (2009) Cullen (2019)
The empirical literature demonstrates reliability-dependent visual–vestibular integration in heading tasks, neural populations with plausible cue-combination properties, and context-dependent changes between self-generated and externally imposed motion. Vection, galvanic illusions, route-reconstruction experiments, and clinical observations reveal how inferred movement can diverge from physical movement and from individual reflex or motor measures. These results jointly justify a model of vestibular perception as multimodal state estimation, with mechanisms operating at multiple levels and timescales. They also identify limits on claims of universal optimality, single-region coding, and simple translation from laboratory thresholds to everyday function. Fetsch et al. (2009) Gu et al. (2008) Day & Fitzpatrick (2005)
The next research priorities are stronger model comparisons across motion frequencies and sensory conflicts, repeated individual psychophysics, active naturalistic experiments, causal tests of neural representations, and prospective work connecting perceptual estimates with spatial behavior and clinical outcomes. Such research will refine our understanding of how a moving organism establishes the direction of gravity, the geometry of its own trajectory, and the stability of the surrounding world. The central accomplishment of the field is a disciplined connection among physical motion, sensory encoding, computational inference, and experienced self-motion. Laurens & Droulez (2007) Lopez & Blanke (2011) Riecke et al. (2023)
References
- Angelaki, D. E., & Cullen, K. E. (2008). Vestibular system: The many facets of a multimodal sense. Annual Review of Neuroscience, 31, 125–150. https://doi.org/10.1146/annurev.neuro.31.060407.125555
- Angelaki, D. E., & Yakusheva, T. A. (2009). How vestibular neurons solve the tilt/translation ambiguity: Comparison of brainstem, cerebellum, and thalamus. Annals of the New York Academy of Sciences, 1164, 19–28. https://doi.org/10.1111/j.1749-6632.2009.03939.x
- Bárány, R. (1916, September 11). Some new methods for functional testing of the vestibular apparatus and the cerebellum [Nobel lecture for the 1914 Nobel Prize]. NobelPrize.org. https://www.nobelprize.org/prizes/medicine/1914/barany/lecture/
- Brandt, T., Schautzer, F., Hamilton, D. A., Brüning, R., Markowitsch, H. J., Kalla, R., Darlington, C., Smith, P., & Strupp, M. (2005). Vestibular loss causes hippocampal atrophy and impaired spatial memory in humans. Brain, 128(11), 2732–2741. https://doi.org/10.1093/brain/awh617
- Bronstein, A. M. (1995). Visual vertigo syndrome: Clinical and posturography findings. Journal of Neurology, Neurosurgery & Psychiatry, 59(5), 472–476. https://doi.org/10.1136/jnnp.59.5.472
- Cullen, K. E., Brooks, J. X., & Sadeghi, S. G. (2009). How actions alter sensory processing: Reafference in the vestibular system. Annals of the New York Academy of Sciences, 1164, 29–36. https://doi.org/10.1111/j.1749-6632.2009.03866.x
- Cullen, K. E., Brooks, J. X., Jamali, M., Carriot, J., & Massot, C. (2011). Internal models of self-motion: Computations that suppress vestibular reafference in early vestibular processing. Experimental Brain Research, 210(3–4), 377–388. https://doi.org/10.1007/s00221-011-2555-9
- Cullen, K. E. (2012). The vestibular system: Multimodal integration and encoding of self-motion for motor control. Trends in Neurosciences, 35(3), 185–196. https://doi.org/10.1016/j.tins.2011.12.001
- Cullen, K. E. (2019). Vestibular processing during natural self-motion: Implications for perception and action. Nature Reviews Neuroscience, 20(6), 346–363. https://doi.org/10.1038/s41583-019-0153-1
- Day, B. L., & Fitzpatrick, R. C. (2005). Virtual head rotation reveals a process of route reconstruction from human vestibular signals. The Journal of Physiology, 567(2), 591–597. https://doi.org/10.1113/jphysiol.2005.092544
- Demer, J. L. (1995). Evaluation of vestibular and visual oculomotor function. Otolaryngology–Head and Neck Surgery, 112(1), 16–35. https://pubmed.ncbi.nlm.nih.gov/7816451/
- Fetsch, C. R., Wang, S., Gu, Y., DeAngelis, G. C., & Angelaki, D. E. (2007). Spatial reference frames of visual, vestibular, and multimodal heading signals in the dorsal subdivision of the medial superior temporal area. Journal of Neuroscience, 27(3), 700–712. https://doi.org/10.1523/JNEUROSCI.3553-06.2007
- Fetsch, C. R., Turner, A. H., DeAngelis, G. C., & Angelaki, D. E. (2009). Dynamic reweighting of visual and vestibular cues during self-motion perception. Journal of Neuroscience, 29(49), 15601–15612. https://doi.org/10.1523/JNEUROSCI.2574-09.2009
- Fetsch, C. R., DeAngelis, G. C., & Angelaki, D. E. (2010). Visual–vestibular cue integration for heading perception: Applications of optimal cue integration theory. European Journal of Neuroscience, 31(10), 1721–1729. https://doi.org/10.1111/j.1460-9568.2010.07207.x
- Fetsch, C. R., Pouget, A., DeAngelis, G. C., & Angelaki, D. E. (2012). Neural correlates of reliability-based cue weighting during multisensory integration. Nature Neuroscience, 15(1), 146–154. https://doi.org/10.1038/nn.2983
- Fitzpatrick, R. C., & Day, B. L. (2004). Probing the human vestibular system with galvanic stimulation. Journal of Applied Physiology, 96(6), 2301–2316. https://doi.org/10.1152/japplphysiol.00008.2004
- Gu, Y., Angelaki, D. E., & DeAngelis, G. C. (2008). Neural correlates of multisensory cue integration in macaque MSTd. Nature Neuroscience, 11(10), 1201–1210. https://doi.org/10.1038/nn.2191
- Karmali, F., Lim, K., & Merfeld, D. M. (2014). Visual and vestibular perceptual thresholds each demonstrate better precision at specific frequencies and also exhibit optimal integration. Journal of Neurophysiology, 111(12), 2393–2403. https://doi.org/10.1152/jn.00332.2013
- Laurens, J., & Droulez, J. (2007). Bayesian processing of vestibular information. Biological Cybernetics, 96(4), 389–404. https://doi.org/10.1007/s00422-006-0133-1
- Laurens, J., & Angelaki, D. E. (2011). The functional significance of velocity storage and its dependence on gravity. Experimental Brain Research, 210(3–4), 407–422. https://doi.org/10.1007/s00221-011-2568-4
- Lopez, C., & Blanke, O. (2011). The thalamocortical vestibular system in animals and humans. Brain Research Reviews, 67(1–2), 119–146. https://doi.org/10.1016/j.brainresrev.2010.12.002
- Mardirossian, V., Karmali, F., & Merfeld, D. M. (2014). Thresholds for human perception of roll tilt motion: Patterns of variability based on visual, vestibular, and mixed cues. Otology & Neurotology, 35(5), 857–860. https://doi.org/10.1097/MAO.0000000000000346
- Mittelstaedt, H. (1983). A new solution to the problem of the subjective vertical. Naturwissenschaften, 70(6), 272–281. https://doi.org/10.1007/BF00404833
- Mittelstaedt, H., & Glasauer, S. (1993). Illusions of verticality in weightlessness. Clinical Investigator, 71(9), 732–739. https://doi.org/10.1007/BF00209728
- Morgan, M. L., DeAngelis, G. C., & Angelaki, D. E. (2008). Multisensory integration in macaque visual cortex depends on cue reliability. Neuron, 59(4), 662–673. https://doi.org/10.1016/j.neuron.2008.06.024
- Oman, C. M. (1982). A heuristic mathematical model for the dynamics of sensory conflict and motion sickness. Acta Oto-Laryngologica Supplementum, 392, 1–44. https://pubmed.ncbi.nlm.nih.gov/6303041/
- Probst, T., Straube, A., & Bles, W. (1985). Differential effects of ambivalent visual-vestibular-somatosensory stimulation on the perception of self-motion. Behavioural Brain Research, 16(1), 71–79. https://doi.org/10.1016/0166-4328(85)90083-X
- Riecke, B. E., Murovec, B., Campos, J. L., & Keshavarz, B. (2023). Beyond the eye: Multisensory contributions to the sensation of illusory self-motion (vection). Multisensory Research, 36(8), 827–864. https://doi.org/10.1163/22134808-bja10112
- Staab, J. P., Eckhardt-Henn, A., Horii, A., Jacob, R., Strupp, M., Brandt, T., & Bronstein, A. (2017). Diagnostic criteria for persistent postural-perceptual dizziness (PPPD): Consensus document of the committee for the classification of vestibular disorders of the Bárány Society. Journal of Vestibular Research, 27(4), 191–208. https://doi.org/10.3233/VES-170622
- Wardman, D. L., Day, B. L., & Fitzpatrick, R. C. (2003). Position and velocity responses to galvanic vestibular stimulation in human subjects during standing. The Journal of Physiology, 547(1), 293–299. https://doi.org/10.1113/jphysiol.2002.030767
- Wiest, G. (2015). The origins of vestibular science. Annals of the New York Academy of Sciences, 1343, 1–9. https://doi.org/10.1111/nyas.12706