The body ownership destabilization hypothesis: a cognitive-resource account of visually induced motion sickness in virtual reality

1 Introduction

Although virtual reality (VR) technology has advanced rapidly in recent years, visually induced motion sickness (VIMS) remains a critical barrier to deeper user engagement and the broader adoption of VR products. VIMS presents as a constellation of symptoms, including dizziness, nausea, and disorientation experienced during VR exposure (Reason and Brand, 1975).

The mechanisms underlying VIMS are multidimensional in nature. Beyond technical constraints such as display latency (Palmisano et al., 2020), the literature has proposed a range of explanatory models (Stanney et al., 2020). Among these, sensory conflict theory remains one of the most influential: when the visual system receives motion information while the vestibular system registers the body as stationary, the resulting signal mismatch is thought to trigger motion sickness symptoms (Reason and Brand, 1975; Reason, 1978). While this theory captures an important physiological basis of VIMS, its explanatory power has clear limitations. First, it treats the body as a passive receptacle for sensory stimuli and thus does not fully account for how users, as cognitive agents, actively engage with virtual environments. Second, even under identical visual stimuli, different interaction modes can produce markedly different levels of VIMS (Nam et al., 2022), suggesting that the emergence of VIMS likely involves cognitive regulatory mechanisms beyond the visual–vestibular conflict itself.

Beyond sensory conflict theory, two complementary models hold particular relevance to our cognitive perspective: the well-established postural instability theory (Riccio and Stoffregen, 1991; see also Bailey et al., 2022) and the more recent DVP hypothesis (Palmisano et al., 2020), which extend the analysis at the behavioral and technical levels, respectively. The postural instability theory holds that VIMS arises when individuals struggle to maintain effective postural control in dynamic environments (Riccio and Stoffregen, 1991). At the HMD-specific technical level, Palmisano et al. (2020) proposed the DVP hypothesis, identifying time-varying differences between virtual and physical head pose (DVP) as a major trigger for VIMS in head-mounted displays (HMDs). While these two models broaden our understanding of VIMS in valuable ways, they share a common limitation: neither sufficiently addresses the user’s active role as a cognitive agent in shaping the VR experience. The postural instability theory posits postural instability as a necessary precursor to VIMS; its cognitive antecedents, however, fall outside the theory’s scope. The DVP hypothesis offers a precise, measurable index of virtual–physical head-pose discrepancy, but its level of analysis remains anchored in physical signal mismatch and does not address the cognitive process by which users actively integrate multisensory information. In short, although existing theories have accumulated substantial evidence at the physiological and behavioral levels, they still lack a systematic account of how the cognitive system regulates and integrates multisensory information, and under what conditions this process breaks down.

To date, theoretical and empirical work that directly links the destabilization of the sense of body ownership to VIMS remains limited. To address this gap, drawing on the framework of embodied cognition, we propose a complementary hypothesis—the Body Ownership Destabilization Hypothesis (BODH)—which posits that body-ownership stability is a key cognitive factor contributing to VIMS susceptibility and symptom severity. Although body ownership destabilization can in principle arise through multiple sensory routes, under current mainstream VR conditions characterized by limited haptic feedback, attenuated proprioceptive correspondence, and visual–vestibular mismatch, users often sustain body ownership through a visually weighted form of cross-modal integration (Rock and Victor, 1964), making the visual channel a major route through which users sustain a sense of belonging to their virtual body (Kilteni et al., 2012; Slater et al., 2010). Under these conditions, the user’s cognitive system is expected to coordinate two interrelated cognitive tasks: maintaining the sense of body ownership through visually weighted multisensory integration, and regulating the inherent visual–vestibular mismatch in VR (Reason and Brand, 1975; Reason, 1978). Both tasks draw on limited cognitive-resource capacity (Kahneman, 1973); as the combined cognitive-resource demand imposed by these tasks exceeds the user’s available regulatory capacity, body ownership is expected to become increasingly vulnerable to destabilization, thereby increasing users’ susceptibility to VIMS.

The aim of this paper is to develop a theoretical framework for explaining VIMS from the perspective of cognitive resources, thereby complementing sensory conflict theory along the cognitive dimension. We examine how cognitive-resource demand may arise in parallel through three concurrent sources—sensory misalignment, visual-environmental load, and predictive instability—and how these sources may jointly make body ownership more vulnerable to destabilization, thereby increasing susceptibility to VIMS. We then derive design implications from this framework for VR visual design.

The remainder of this paper is organized as follows. Section 2 introduces embodied cognition as the theoretical premise. Section 3 examines three concurrent sources of cognitive-resource demand that may jointly increase the vulnerability of body ownership to destabilization. Section 4 derives design implications from the framework. Section 5 concludes.

2 Embodied cognition as a theoretical premise for understanding VR experience

2.1 Embodied cognition

Embodied cognition theory challenges the traditional mind-body dualism, holding that human cognition is not an abstract computation sealed within the brain, but a dynamic process jointly enacted by the brain, the body, and the environment (Varela et al., 1991; Gallagher, 2005). This perspective offers a crucial entry point for understanding VR experience. Compared with the more spectator-oriented mode of traditional two-dimensional media, the core of VR interaction lies in bodily involvement: through head rotations, spatial displacement, and similar movements, users employ their physical body as a reference frame through which they perceive digital space.

Notably, the user does not enter the virtual environment as a blank slate, but brings with them a body schema shaped by prior real-world sensorimotor experience. The body schema can be understood as a pre-reflective bodily organization through which posture, movement, and action possibilities are coordinated (Merleau-Ponty, 2013). From a phenomenological perspective, virtual space does not become experientially meaningful merely through visual presentation; only when the user takes it up through intentional activity and brings it into alignment with their existing body schema does the virtual environment become a perceptual field they can genuinely inhabit. The core question of VR experience, therefore, is not only the fidelity of visual representation, but also whether the user’s physical body can establish a stable embodied coupling with the virtual environment. Embodied cognition thus provides a theoretical premise for understanding VIMS: motion sickness is not simply a physiological response to “what is seen,” but can also be understood as involving a disruption of sensorimotor integration during virtual experience, in which bodily action, sensory feedback, and environmental structure fail to remain coherently aligned.

2.2 The sense of body ownership and its vulnerability in VR

In the physical world, embodied perception is grounded in the close coherence of multisensory signals—visual, tactile, proprioceptive, and vestibular. It is this multisensory coherence that supports the emergence of the sense of body ownership: the subjective conviction that ‘this body belongs to me’ (Gallagher, 2005). Within the context of VR, Kilteni et al. (2012) further decompose the sense of embodiment into three interrelated subcomponents—sense of body ownership, sense of agency, and sense of self-location. Among these, the present paper focuses on body ownership, which provides the most direct anchor for analyzing the destabilization of embodied experience under conditions of limited or mismatched multisensory cues in VR. Empirical research has demonstrated that the sense of body ownership is highly malleable. The classic rubber hand illusion (Botvinick and Cohen, 1998) showed that synchronous visuo-tactile stimulation alone can lead individuals to perceive a rubber hand as part of their own body. Maselli and Slater (2013) extended this principle to full-body illusions, showing that under multisensory synchrony, users can develop a sense of ownership over an entire virtual body. Taken together, these findings converge on the spatiotemporal congruence principle of multisensory integration (Stein and Meredith, 1993): the more closely sensory signals align in time and space, the more readily the sense of body ownership can be established.

In current VR environments, however, limited haptic feedback, attenuated proprioceptive correspondence, and mismatch between visual motion and vestibular signals can weaken the multisensory synchrony on which body ownership normally depends. To compensate for this limited multisensory support, the cognitive system often redistributes sensory weighting: through the mechanism of visual capture, the brain actively recalibrates proprioception and, when faced with cross-modal conflict, tends to prioritize visual input (Rock and Victor, 1964; Slater et al., 2010). As a result, VR users come to rely heavily on the visual channel to sustain the perception of their own body. Yet a sense of body ownership maintained primarily through visual information under conditions of reduced multisensory congruence may be less stable than one grounded in coherent multisensory integration; it is also more vulnerable to external perturbations. We propose that this fragility may, in turn, amplify users’ sensitivity to the inherent visual–vestibular mismatch in VR. Within the embodiment framework of Kilteni et al. (2012), body ownership constitutes a core embodied component of presence; the stability of body ownership is therefore expected to bear directly on users’ overall VR experience. Indirect support comes from the systematic review by Weech et al. (2019), which reports a negative correlation between presence and cybersickness (the HMD-specific subset of VIMS), with subsequent empirical work providing further support (Salimi and Ferguson-Pell, 2021). Following this logic, the destabilization of body ownership may help explain part of this negative relationship at the embodied level.

It should be noted that the destabilization of body ownership can be analyzed from multiple sensory and sensorimotor routes—including limited haptic feedback, attenuated proprioceptive correspondence, visual–vestibular mismatch, auditory spatial cues, and disruptions of motor control. Auditory spatial cues, in particular, have been associated with self-motion perception, presence, and immersion in VR (Mavridou et al., 2025; Riecke et al., 2009). Under the technical conditions of current consumer-grade HMD-based VR systems, however, haptic feedback remains limited, virtual–physical proprioceptive correspondence is often imperfect, and vestibular signals cannot be directly adjusted through interface design. User-level factors, including prior VR or gaming experience, baseline motion-sickness susceptibility, age, and gender, may further modulate VIMS susceptibility (Kourtesis et al., 2024; Llana et al., 2025). However, because these factors are not directly manipulable design variables, the present framework treats them as boundary conditions rather than as components of the proposed visual-design mechanism. For this reason, the present paper adopts the visual pathway as its primary analytical scope, because visual information remains one of the most tractable components for interface-level intervention. Accordingly, this paper focuses its analysis on the cognitive-resource demands associated with the visual pathway.

Taken together, the analysis above suggests that body ownership in consumer-grade HMD-based VR is often maintained under conditions in which sensory and sensorimotor cues are limited, mismatched, or only partially aligned, leading users to rely more heavily on visual information for embodied coupling. This condition serves as the point of departure for our analysis in Section 3, where we examine how cognitive-resource demand can arise from three concurrent sources and jointly increase the vulnerability of body ownership to destabilization. To contextualize this theoretical shift, Figure 1 contrasts the classical sensory-conflict account of VIMS with the embodied cognitive-resource-regulation framework proposed here.

3 Cognitive-resource sources underlying VIMS susceptibility

Building on the vulnerability of body ownership under the limited and mismatched multisensory conditions outlined above, this section examines how cognitive-resource demand may arise in parallel in VR through three concurrent sources: sensory misalignment, visual-environmental load, and predictive instability. In this paper, cognitive resources are used as a broad, capacity-limited construct encompassing perceptual processing capacity, attentional control, multisensory integration resources, and predictive/sensorimotor recalibration resources, rather than referring narrowly to working memory alone. Within the proposed additive-threshold model, these sources are treated as concurrent rather than sequential: they jointly draw on these overlapping cognitive resources, and their combined demand is expected to increase VIMS susceptibility once it exceeds the user’s available regulatory capacity. Under such conditions, body ownership is proposed to become more vulnerable to destabilization.

3.1 Material constraints and sensory misalignment

From a phenomenological perspective, space is not an external, fixed physical container, but a lived situation generated through the interaction between body and environment (Merleau-Ponty, 2013). In the everyday physical world, this body-environment interaction inherently relies on tangible and resistant material support. A stable spatial experience, in other words, is grounded in the presence of material entities. VR environments, by contrast, are intrinsically marked by material constraints: real physical resistance is largely unavailable, haptic feedback is often limited, and proprioceptive correspondence is often imperfect in virtual spaces, thereby weakening the natural coupling between body and world that embodied cognition presupposes and creating challenges for multisensory integration and embodied action. Recent empirical work has shown that limited or absent haptic feedback in VR can substantially alter users’ eye-hand coordination patterns and reduce their sense of embodiment (Lavoie et al., 2025). One important consequence of material constraints is a systematic mismatch among sensory and sensorimotor cues, aligning with the classical sensory conflict account (Reason and Brand, 1975). Such cross-modal sensory mismatch reflects the objective material constraints of VR and can increase susceptibility to VIMS.

This mismatch first manifests at the macro level of spatial orientation and self-motion perception. In the physical world, spatial orientation depends on the precise coordination between the visual and vestibular systems. Furthermore, physical self-motion is strictly bounded by gravity, inertia, and bodily effort. Material constraints in VR weaken these constraints, enabling less constrained motion experiences—such as high-speed locomotion or sudden viewpoint shifts—that may conflict with the bodily expectations built into the body schema (Merleau-Ponty, 2013). During such experiences, the visual system signals rapid movement, yet the physical body remains stationary. This specific visual–vestibular conflict constitutes a major source of sensory conflict associated with motion sickness (Reason and Brand, 1975). Importantly, this conflict can undermine the physical grounding of visually supported body ownership, thereby potentially weakening embodied perception. Furthermore, material constraints induce a sensory disconnect at the micro level of interaction. Specifically, it creates a structural gap between visually implied affordances (Gibson, 1979) and actual haptic feedback. When realistic visual cues in VR—such as a graspable object or a solid wall—strongly suggest a tangible or interactive affordance, the body schema (Merleau-Ponty, 2013) is activated in anticipation of corresponding physical resistance and haptic feedback. Yet because real physical feedback is limited or unavailable, the affordance promised by vision cannot be fully verified, the action feedback anticipated by the body schema cannot be realized, and the embodied perception-action loop is thereby disrupted. Faced with this cross-modal sensory disruption, the visual pathway is placed under increased demands for cognitive integration. It is expected to support the sense of body ownership while accommodating both sensory conflict and the affordance gap.

3.2 Visual-environmental load and visual redundancy

The material constraints of VR can create sensory misalignment at both the macro level of motion perception and the micro level of interaction, placing the visual pathway under increased demands of cognitive integration. Sensory misalignment alone, however, cannot fully account for another observed pattern: across both recent VR studies and earlier work on simulator-based environments, more visually realistic or visually complex scenes may sometimes be associated with more severe sickness symptoms (Davis et al., 2015; Pouke et al., 2018; see also Jaeger and Mourant, 2001; Liu and Uang, 2011; both as cited in Rebenitsch and Owen, 2016). This pattern suggests that visual-environmental complexity may be one factor contributing to VIMS severity under certain conditions. In this paper, this additional burden is conceptualized as visual-environmental load, referring to the perceptual and attentional demand generated by high-density, task-irrelevant, or redundant visual details in immersive VR environments. Here, visual-environmental load refers to an environment-driven source of demand, whereas cognitive load refers to the broader resource demand experienced by the user during VR exposure. Although such load may contribute to perceptual overload under certain conditions, it is used here as a design-level construct concerning the surrounding visual environment rather than as a direct extension of Lavie’s task-based perceptual load theory. The dominant design paradigm in current VR often pursues “physical fidelity”: through fine-grained visual details—realistic skin pores, fine-grained material textures, and precise lighting and shadow—it seeks to closely simulate the physical world. Underlying this approach is a widely accepted design intuition: the closer the visual representation is to reality, the better the immersive experience. The issue, however, lies not in the pursuit of realism as such, but in the dominant strategy by which this goal is implemented—stacking high-density visual details to approximate real-world vision—an approach that may impose additional visual-environmental load by increasing visual detail without providing equivalent material or sensorimotor correspondence. In environments where physical resistance and haptic feedback remain limited, these visible-but-untouchable realistic details may fail to complete the embodied perception–action loop and may instead become an additional source of visual-environmental load on the visual channel.

How does this load arise at the cognitive level? Lavie’s perceptual load theory (Lavie, 1995) provides an important point of contrast rather than a direct explanatory template for the present account. In Lavie’s task-based framework, high core-task load can reduce the processing of task-irrelevant distractors by exhausting early perceptual capacity. The condition considered here is different: in immersive VR, dense environmental detail is not simply presented as a discrete distractor outside a bounded focal task, but forms part of the continuously available surrounding visual field through which users maintain spatial orientation, action readiness, and embodied coupling. For this reason, the present paper uses visual-environmental load as a design-level, environment-driven construct, referring to the sustained perceptual and attentional demand generated by dense, salient, or task-irrelevant environmental detail in immersive VR. This construct is therefore not a direct extension of Lavie’s task-load framework, but a complementary account of how surrounding visual environments may constrain shared processing resources in VR. When this visual-environmental load remains high, the resources available for core task execution, multisensory integration, and the maintenance of body ownership may be correspondingly constrained. Within an attentional system that is capacity-limited and resource-shared (Kahneman, 1973; Lavie et al., 2004), this sustained demand imposes potential constraints on cognitive functions that depend on shared resources. Beyond this, while the perceptual system processes redundant information, a parallel competition for attention is also taking place. In realistic visual environments, redundant details that are visually striking yet entirely unrelated to the user’s core interactive task—such as high-contrast lighting transitions and vivid material textures—can act as salient distractors in VR (Yantis, 1993) that trigger bottom-up attentional capture (Theeuwes, 1992). To prevent attention from being misallocated, the brain may engage a signal suppression mechanism (Sawaki and Luck, 2010); the operation of this mechanism, however, itself consumes additional cognitive resources. In short, the brain is expected to simultaneously process redundant perceptual information and expend attentional resources to suppress the competition these details introduce. It should be noted that, unlike the bounded planar field of conventional 2D screens, VR provides a 360° immersive panoramic space, with visual stimuli surrounding the user from every direction. Empirical evidence also indicates that cognitive load is significantly higher under VR immersive conditions than under 2D screen conditions (Breves and Stein, 2023). We therefore suggest that, under 360° VR conditions, the dual consumption mechanism described above can persist throughout exposure rather than remaining confined to a single focal task.

What functions do these consumed cognitive resources normally support? According to the two-mode visual processing hypothesis (Leibowitz and Post, 1982), the human visual system can be divided into focal vision, which handles object recognition and detail processing, and ambient vision, which handles spatial orientation, postural balance, and the processing of optic flow. In VR interaction, ambient vision continuously provides the brain with baseline information about “where the body is located in space,” and constitutes an important processing basis through which the cognitive system regulates the sensory conflict described in Section 3.1. Schwartz et al. (2005) found that high attentional load at central fixation suppresses peripheral visual processing. This finding suggests that when visual attention is repeatedly concentrated on local details in VR, the peripheral processing that supports ambient vision may be reduced. In high-fidelity VR, dense local details may be distributed across multiple regions of the visual field, so repeated fixation on such details can impose elevated local load, further reducing the peripheral processing on which ambient vision relies for spatial orientation. We therefore infer that as the processing of redundant visual details and active attentional suppression continue to consume cognitive resources, the processing capacity of ambient vision may correspondingly decline. This redistribution of resource demand is proposed to increase reliance on focal vision—originally responsible for object recognition—for part of the spatial-orientation function. However, the processing characteristics of focal vision are such that it can only handle focused, non-global spatial information, and is unlikely to fully provide the continuous, holistic spatial localization that ambient vision affords. This compensatory processing, which lies outside its usual functional domain, may be inefficient and resource-demanding.

This process has important implications for embodied perception. We hypothesize that, in normal embodied perception, ambient vision provides the individual with stable spatial localization information, which constitutes part of the perceptual basis for maintaining the sense of body ownership. When the processing efficiency of ambient vision is reduced under conditions in which perceptual processing resources are occupied by dense visual details and attentional-control resources are used to suppress irrelevant attentional capture, and when focal vision’s compensation is insufficient to offset this reduction, the combined perceptual and attentional demand may exceed available cognitive capacity, making the brain less able to sustain a stable, macroscopic perception of the relation between self and space. The user may consequently lose part of the perceptual basis for sustaining spatial stability, while the cognitive system’s capacity to manage the underlying sensory conflict is correspondingly constrained. Within the BODH framework, this condition is expected to increase the likelihood of autonomic defensive responses such as nausea and dizziness (Reason and Brand, 1975; Treisman, 1977). The sense of body ownership is therefore expected to become more vulnerable to destabilization under conditions of elevated visual-environmental load. This process is also potentially linked to the postural instability theory (Riccio and Stoffregen, 1991), which holds that an individual’s failure to maintain effective postural control under certain conditions is a major contributing cause of motion sickness. The decline in spatial-orientation capacity brought about by the sustained consumption of perceptual and attentional-control resources, as proposed here, may represent one possible cognitive pathway leading to postural destabilization. This proposed pathway is further operationalized in the testable predictions outlined in Section 5.1.

In sum, alongside sensory misalignment, overly detail-dense realistic visual design can introduce large quantities of task-irrelevant or redundant visual information, simultaneously consuming cognitive resources at both the perceptual processing and attentional suppression levels, and reducing the processing capacity available to ambient vision for spatial orientation and balance maintenance. The compensatory engagement of focal vision, owing to limitations in its processing characteristics, is unlikely to fully replace ambient vision, and the user may lose part of the perceptual basis for sustaining a sense of spatial stability, leaving the sense of body ownership more vulnerable to destabilization. It should be noted that high visual-environmental load is not a necessary condition for VIMS—VIMS can still be observed in low-fidelity environments, although its severity is numerically lower (Pouke et al., 2018). However, in detail-dense VR environments, visual-environmental load can still act as an exacerbating factor for VIMS by placing sustained demand on the cognitive resources also needed for multisensory integration, spatial orientation, and predictive recalibration. The full pathway is summarized in Figure 2.

3.3 Predictive instability and embodied vulnerability

Within the BODH framework, predictive instability is conceptualized as a third concurrent source of cognitive-resource demand, operating alongside sensory misalignment and visual-environmental load. This section examines how disruptions in prediction-action-feedback matching may increase the resources required for predictive and sensorimotor recalibration, thereby making body ownership more vulnerable to destabilization.

Predictive coding theory in cognitive neuroscience (Rao and Ballard, 1999; Friston, 2010) holds that the brain does not passively receive external information, but instead actively and continuously predicts upcoming sensory input on the basis of prior experience. This framework has also been applied to VR-related presence research in recent years (Jicol et al., 2023). In typical physical environments, the brain generates predictions about incoming sensory feedback based on motor intentions; when predictions match actual feedback, a dynamic loop of prediction, action, and verification is formed, and the coupling between body and environment tends to remain stable. When discrepancies arise, prediction errors are generated, and the brain expends cognitive resources to recalibrate its predictive model and minimize them. In VR, however, this loop can encounter particular difficulties, especially when the correspondence between motor intention, action, and visual feedback is weakened. Externally controlled viewpoints, passive locomotion, or delayed feedback can disrupt the expected relation between intended movement and sensory consequence. Under such conditions, prediction errors may persist, and the resources required for predictive and sensorimotor recalibration can increase. Predictive processing is therefore better described as becoming unstable, rather than as having failed.

One important manifestation of predictive instability is a weakened sense of agency. In many VR applications in which the system externally controls the viewpoint—the typical virtual roller coaster being a clear example—users may be less able to actively verify visual changes through their own movements and are instead positioned as passive recipients of abrupt viewpoint shifts. This passive-viewing mode has been shown to be associated with more severe motion sickness (Nam et al., 2022). At such moments, although the brain attempts to match its existing body schema to rapidly changing visual scenes, it may struggle to attribute these visual changes to its own motor intentions. According to the comparator model within predictive coding, the sense of agency depends on a match between intended movement and actual sensory feedback, and agency interacts closely with, and reinforces, the sense of body ownership (Forster et al., 2022; Haggard, 2017). Consequently, prediction errors can increase and place greater demand on sensorimotor recalibration; the user’s sense of agency may be weakened, leaving the sense of body ownership less stable (Blanke and Metzinger, 2009). When the brain can no longer reliably integrate visual input into a coherent bodily experience through the predictive loop, the relation of belonging between the user and the virtual body may be weakened. Faced with this perceptual mismatch, the autonomic nervous system is hypothesized to generate defensive physiological responses, potentially contributing to symptoms such as nausea, dizziness, and cold sweats (Reason and Brand, 1975; Treisman, 1977).

This account suggests that VIMS susceptibility is not fully explicable by sensory conflict alone. Rather, within the proposed additive-threshold model, it may be understood as a systemic outcome of increased vulnerability in embodied coupling when concurrent cognitive-resource demands exceed the user’s available regulatory capacity. Sensory misalignment, visual-environmental load, and predictive instability jointly draw on the brain’s limited information-processing capacity. Importantly, the relative contribution of these sources is expected to vary across VR scenarios, with sensory misalignment or visual-environmental load contributing more strongly in some cases and predictive instability playing a more prominent role in others, even when visual complexity is relatively low. Within this process, the sense of body ownership is proposed to function as a particularly sensitive cognitive juncture because it relies on visually weighted cross-modal integration under VR conditions in which haptic and proprioceptive correspondence can be limited; its stability is hypothesized to modulate the user’s tolerance for sensory conflict.

4 Design implications: from “physical fidelity” to “perceptual plausibility”

The analysis above suggests that VIMS susceptibility can be understood as the systemic outcome of combined cognitive-resource demand across three concurrent sources—sensory misalignment, visual-environmental load, and predictive instability—which can increase the vulnerability of body ownership to destabilization when the user’s regulatory capacity is exceeded. Visual design can therefore be reoriented away from the pursuit of “physical fidelity” toward one of “perceptual plausibility,” an orientation that aims to align visual presentation with, rather than overload, users’ limited information-processing capacity. Perceptual plausibility, as understood here, does not pursue physical truthfulness through the maximization of visual detail; rather, it retains in the visual presentation only the minimal effective cues needed to sustain body ownership, while avoiding unnecessary visual-environmental load. As Heim (1993) has argued, the real does not refer only to entities with a physical-material basis but also encompasses experiences constructed through interaction with digital environments. Given that current consumer-grade HMD-based VR is marked by material constraints and limited multisensory correspondence, visual-design interventions need not prioritize physical realism, but should instead focus on sustaining the stability of embodied perception through visual configurations that align with cognitive and sensorimotor constraints. The earlier analysis indicates that reliance on the visual-dominant route is one important source of the fragility of body ownership; yet under the current technical conditions, in which VR multisensory feedback remains underdeveloped, the visual channel is also a particularly feasible and tractable avenue for design interventions. Maselli and Slater (2013) experimentally demonstrated that, under a first-person perspective and even in the absence of haptic synchrony, the plausibility of visual appearance alone can elicit a certain degree of body-ownership illusion. This provides a rationale for considering visual-channel design as one route for deriving the strategies proposed below.

First, where genuine haptic feedback is limited or unavailable, dynamic visual cues can serve as a substitute, supplying alternative signals of physical feedback. When the user’s virtual hand touches an object, local UI highlights, localized particle effects at the contact point, or elastic deformation animations can provide clear visual cues of resistance and collision. This approach is already widely adopted in current VR interaction design. As Merleau-Ponty (2013) observes, the tactile perception of someone using a cane extends to the tip of the cane, and the tool thereby becomes incorporated into the body schema. The design logic of visual compensation parallels this: by replacing missing haptic feedback with dynamic visual cues, the visual channel is extended into a substitute channel for perceiving physical properties. Research on “pseudo-haptics” in virtual environments has shown that, when corresponding haptic feedback is not provided, visual feedback coupled with users’ actions can simulate haptic sensations without requiring a dedicated haptic interface (Lécuyer, 2009). By providing substitute visual feedback, this form of visual compensation can help maintain the perception–action coupling on which the body schema depends, reduce the cognitive-resource demand involved in resolving visuo-tactile mismatches, and thereby help limit prediction error.

Second, “visual simplification”—which we frame here through the principle of liubai—can be adopted to address visual-environmental load. As noted above, realistic visual environments can introduce redundant visual detail, thereby increasing this load under current VR conditions. Design interventions therefore need not pursue physical fidelity to excess, but should instead practice deliberate visual subtraction. The Chinese aesthetic principle of liubai—understood as “intended blank” or “negative space” (Fong, 1984)—offers a productive conceptual framework. In traditional Chinese painting theory, liubai is not mere emptiness; rather, through the compositional principle of xushi (emptiness–substance) (Kuo, 2015), it places a finite quantity of brush-and-ink expression in dialogue with boundless space. The “substantive” (shi) conveys pictorial meaning, while the “empty” (xu) accommodates the viewer’s imagination; the two are mutually dependent and mutually defining (for a foundational Chinese-language treatment of these aesthetic categories, see Zong, 2015). What appears sparse on the surface is not experienced as lacking, because the viewer is invited to imaginatively complete the visual field. This principle suggests a conceptual parallel for VR design: by intentionally reducing task-irrelevant visual detail, the designer does not create a void but rather reduces unnecessary visual-environmental demand and directs visual attention toward core embodied interaction objects. In this sense, liubai provides a conceptual basis for developing visual-simplification strategies in VR. In recent years, design researchers have begun to explore how this kind of emptiness–substance spatial order can be reconstructed in VR environments (Chen et al., 2025); the present paper translates this aesthetic principle further into a visual-simplification strategy for VR design. Visual simplification, inspired by the principle of liubai, is not the pursuit of degraded visual quality. By stripping away redundant details in non-interactive regions, it can reduce unnecessary visual-environmental load while preserving the cues needed to sustain body ownership. Concretely, VR design can adopt strategic fidelity modulation: instead of relying on highly detailed material textures and photorealistic lighting, it can employ low-poly modeling and non-photorealistic rendering (NPR) as an intentional visual language, and reduce color contrast in non-interactive background zones. A recent VR experiment provides preliminary empirical support for this direction, showing that geometric scene simplification was associated with reduced VIMS symptoms and reduced neural activation suggestive of diminished cognitive involvement (Abdalhadi et al., 2026). As an illustrative design case, the VR game Beat Saber (Beat Games, 2019) exemplifies this visual-simplification logic through its low-poly, minimalist visual style, which reduces realistic detail in non-interactive regions and foregrounds the core task of rhythmic striking. Taken together, this preliminary evidence and illustrative example suggest that judiciously reducing visual information density is a plausible design direction for addressing visual-environmental load.

Finally, predictive alignment of visual feedback can be implemented as a strategy for limiting predictive instability. Design should strive to ensure that visual changes in VR remain spatiotemporally consistent with the user’s motor intentions and expected sensory consequences. In the temporal dimension, visual feedforward cues can be introduced before abrupt scene transitions or large-scale displacements occur. For example, in teleportation-based displacement, destination indicators or pointing rays can be presented to allow users to specify the target location in advance (Bozgeyikli et al., 2016). Such pre-indications may make the upcoming spatial transition more predictable and reduce the perceived abruptness of the subsequent passive visual change, thereby helping maintain predictive coherence. In addition, in scenarios requiring continuous motion, a “rest frame” (Prothero, 1998) that remains relatively stationary with respect to the user’s body—such as a virtual cockpit or helmet rim—can be introduced. This provides ambient vision with a stable spatial anchor. By aligning static peripheral visual cues with the vestibular sense of stillness, this design can help users interpret fast-changing central scenes as external environmental motion while maintaining a stable spatial frame. Half-Life: Alyx (Valve Corporation, 2020), for example, uses visual pre-cues before teleportation to make the upcoming spatial transition more visually predictable; the virtual cockpit rims in flight simulators, or “virtual noses” (Wienrich et al., 2018) inserted into first-person navigation games, follow a related design logic of visual anchoring. Cao et al. (2018) reported that such rest frames improve users’ real-time comfort during VR locomotion and allow them to remain in the experience for longer durations. We argue that such stable visual reference frames may support predictive stability by making visual motion more compatible with vestibular and proprioceptive expectations; within the BODH framework, this reduction in predictive instability is proposed to help preserve agency and body-ownership stability, thereby potentially reducing VIMS susceptibility.

In sum, strategies such as virtual cockpits, UI highlights, and field-of-view restriction are already used in VR design; yet many of these comfort-oriented measures have emerged primarily from engineering practice and, although individually supported by partial theoretical work, they still lack a coherent integrative framework at the cognitive level. This section takes cognitive-resource regulation as an integrative perspective, situating these scattered design practices within the orientation of “perceptual plausibility” and organizing them into three mechanism-linked strategies—visual compensation for addressing sensory misalignment, visual simplification for addressing visual-environmental load, and predictive alignment for limiting predictive instability. Through this integration, visual design can be understood as supporting body-ownership stability by reducing the cognitive-resource demands that destabilize embodied perception, thereby offering a more systematic theoretical basis and design-oriented guidance for VIMS mitigation in VR. Taken as a whole, these strategies should be understood as theoretical design implications derived from BODH, rather than as empirically established interventions whose effectiveness has already been fully validated. Figure 3 summarizes these three strategies and their proposed convergence toward body-ownership stability.

5 Discussion

5.1 Testable predictions and falsifiability

Because BODH is proposed as a theoretical framework rather than an empirically established causal model, its value depends on whether it can generate operationalized and falsifiable predictions. The core claim of BODH is not that body ownership destabilization replaces sensory conflict, postural instability, or predictive-processing accounts, but that it may function as a cognitive juncture through which sensory mismatch, visual-environmental load, and prediction error modulate the severity of VIMS. This claim differentiates BODH from generic presence–cybersickness accounts and from the rest-frame hypothesis: body ownership, rather than global presence or visual anchoring per se, is treated as the mediating juncture through which sensory mismatch, visual-environmental load, and predictive instability shape VIMS susceptibility. This positioning also allows BODH to accommodate heterogeneous findings in the literature (Weech et al., 2019; Rebenitsch and Owen, 2016): high presence or high visual realism need not uniformly reduce or increase VIMS, because their effects may depend on how sensory misalignment, visual-environmental load, predictive instability, and body-ownership support interact in a given VR scenario. On this basis, future empirical work could examine BODH by holding basic visuo-vestibular conflict constant while manipulating visual density, body-ownership support, or the predictability of visual motion. Body ownership may be assessed using validated embodiment or body-ownership questionnaires; VIMS may be measured using the Simulator Sickness Questionnaire (SSQ) or Cybersickness Questionnaire (CSQ); and cognitive load may be indexed through NASA-TLX, with pupillometry or EEG measures serving as optional convergent indicators. Table 1 summarizes several testable predictions and the observations that would count against the proposed framework.

TABLE 1. Testable predictions derived from BODH and observations that would challenge the framework.

The predictions are intended as operational routes for future empirical tests rather than as a single required experimental protocol.

5.2 Boundary conditions and future directions

Drawing on embodied cognition theory, this paper proposes the Body Ownership Destabilization Hypothesis (BODH), conceptualizing VIMS susceptibility as potentially arising from combined cognitive-resource demand across three concurrent sources—sensory misalignment, visual-environmental load, and predictive instability—which may increase the vulnerability of body ownership to destabilization. Beyond sensory conflict theory, this framework offers a complementary account of VIMS from the perspective of cognitive-resource regulation. Building on this theoretical account, the paper argues that visual design may benefit from a shift from “physical fidelity” to “perceptual plausibility,” with the aim of supporting body-ownership stability at the cognitive level through three strategies: visual compensation, visual simplification, and predictive alignment. As a theoretical study aimed at constructing an explanatory framework, the present work opens four directions for future research. First, the model focuses on the visually dominant nature of current consumer-grade HMD-based VR, where visual information remains a particularly tractable target for interface-level intervention. Future research could extend this scope by incorporating auditory spatial cues, haptic feedback, and virtual–physical proprioceptive correspondence, examining how multimodal coordination may support body-ownership stability and modulate VIMS susceptibility. Second, the proposed cognitive-resource framework and the visual-design strategies introduced in this paper open a clear space for empirical follow-up: future interdisciplinary work, drawing on behavioral and neurophysiological measures, could quantitatively test the proposed pathways and evaluate whether the associated design strategies alleviate VIMS and support body-ownership stability, while also examining whether changes in VIMS are associated with presence, task performance, and related cognitive or motor outcomes (Kourtesis et al., 2024; Llana et al., 2025). Third, among the three design strategies proposed here—visual compensation, visual simplification, and predictive alignment—visual simplification draws on the Chinese aesthetic principle of xushi (emptiness–substance), representing an initial attempt to translate Chinese aesthetic resources into a cognitive framework for VR design. Fourth, future empirical studies should incorporate individual-difference moderators, including prior VR experience, baseline motion-sickness susceptibility, age, and gender, when testing BODH. These variables may condition the strength of the proposed pathways and should therefore be treated as covariates or moderators rather than as components of the visual-design mechanism itself. Future research could build on this foundation by translating these strategies into concrete design prototypes and tools through design practice and empirically evaluating their effectiveness in real VR application scenarios. It could also further explore the deeper integration of Chinese aesthetic traditions with cognitive VR design, developing design guidelines tailored to different types of VR content. Collectively, these lines of inquiry will further test and deepen the BODH framework, contributing to a more complete theoretical and practical foundation for cognitive-resource-informed design approaches to VR-induced motion sickness.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. During the preparation of this work the author(s) used Claude (Anthropic) in order to improve language and readability. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: T. John Tharakan, Indian Space Research Organisation, India

Reviewed by: Marta Mendez, University of Oviedo, Spain

Dian Novian, Universitas Negeri Gorontalo, Indonesia