Why does proprioception get so much say in what feels like 'my' body?
Body ownership experiments work by creating a conflict. You see a hand or body that is not exactly where your own unseen limb is, and researchers ask how much you treat the visible one as yours. The brain's job is to decide whether the visual and proprioceptive streams describe the same object. Proprioception is the sense of where your limbs are in space, generated from muscle and joint signals rather than from vision.
A virtual-reality reaching task made this quantitative. Participants reached while a virtual hand was displaced from their real hand, and the reaching errors they made served as an implicit measure of how much they had taken the virtual hand as their own. A Bayesian model — a formal account of how a system should combine noisy signals in proportion to how reliable each one is — predicted those errors well, and the model's estimate of each person's proprioceptive precision matched their independently measured proprioceptive precision [2]. In plain terms, people whose position sense was sharper leaned less on the misleading visual hand. Reaching errors also tracked participants' explicit ownership ratings, both across individuals and trial by trial [2].
The same logic appears when proprioception is damaged. In 20 stroke patients with sensory deficits in one upper limb, compared with 35 age-matched controls, patients were more likely to feel ownership over an incongruent virtual hand and to fold it into their movements. The model attributed this to a reduced ability to detect visuo-proprioceptive mismatch, scaling with the size of each patient's proprioceptive deficit [5]. That is a clean demonstration that the weighting is not fixed: weaken one input and the other gains influence.
What happens when the signals disagree, and does age change the balance?
Introduce a delay between what someone sees and what they feel, and ownership fades. In a mixed-reality setup spanning ages 7 to 80, participants saw their own body stroked or moved with varying visual delays; across the whole lifespan, the sense of ownership dropped as the temporal mismatch grew, in both the tactile and the motor version of the task [3]. Interestingly, older participants reported a stronger sense of ownership overall, independent of delay and modality, while sensitivity to delay was not consistently related to age, and neither visual dependence nor proprioceptive accuracy varied with age [3]. The authors read this pattern as consistent with a greater weighting of prior expectations and a lesser weighting of incoming sensory signals in the momentary sense of bodily self as people age — a proposal, not a settled mechanism.
Proprioceptive input can also be pushed around directly. Vibrating leg muscles, which perturbs the proprioceptive signal, modulates bodily self-consciousness and self-identification in experimental settings, with congruent proprioceptive signals supporting illusory ownership and incongruent ones reducing it [4]. The takeaway across these paradigms is consistent: ownership is a negotiation among vision, touch, and position sense, and shifting the reliability or content of any one of them shifts the outcome.
Do heartbeat signals feed into body ownership at all?
The cardiac side of the question is much less settled, and the available evidence is largely negative for a link to body awareness. In a study of 105 undergraduates, participants completed a mental heartbeat tracking task as a measure of cardiac accuracy, plus two proprioceptive tasks — discriminating the tension of arm flexor muscles and reproducing the angle of the elbow joint. Apart from a weak association between cardiac accuracy and confidence in that same task, the interoceptive measures showed no associations with each other, within or across the cardiac and proprioceptive modalities, and none with positive or negative affect or perceived body competence [1].
Two cautions belong right next to that result. First, the study measured body awareness, perceived body competence, and affect — not body ownership over an external or virtual limb — so it constrains the cardiac-ownership link indirectly rather than testing it head-on. Second, the samples in the ownership experiments described above did not include cardiac measures at all [2][3][5], and the vibration study likewise addressed proprioceptive, visual, and tactile integration without cardiac signals [4]. So the honest summary is that proprioception has a demonstrated, quantifiable role in self-attribution, while a comparable role for cardiac signals remains untested in these paradigms rather than disproven.
Why the distinction matters for anyone reading claims about the bodily self
It is tempting to treat 'the bodily self' as one integrated system fed equally by heart, muscles, and eyes. These studies point to a more lopsided picture. The ownership experiments converge on a mechanism in which the brain weighs visual and proprioceptive estimates by their reliability, and the consequences are measurable in behavior — reaching errors, ownership ratings, and how stroke changes both [2][5]. Age appears to tilt the same weighting toward prior expectations [3], and direct perturbation of proprioception shifts self-identification [4].
Cardiac accuracy, in the one dataset that put it alongside proprioceptive accuracy, behaved as a separate faculty: it did not predict proprioceptive performance, body awareness, or body competence [1]. That does not make heartbeat signals irrelevant to selfhood in general; it means that in the paradigms used to study body ownership, nobody has yet shown them doing the work that proprioception demonstrably does.
Sources used in this answer
Cardiac and Proprioceptive Accuracy Are Not Related to Body Awareness, Perceived Body Competence, and Affect
In 105 undergraduates, cardiac accuracy (heartbeat tracking) and proprioceptive accuracy (weight discrimination, joint position reproduction) were largely unrelated to each other and showed no association with body awareness, perceived body competence, or positive and negative affect, apart from a weak link between cardiac accuracy and confidence in that task.
The self and the Bayesian brain: Testing probabilistic models of body ownership through a self-localization task
In a virtual-reality reaching task, a Bayesian model of hand ownership based on visuo-proprioceptive congruence predicted reaching errors, individually measured proprioceptive precision correlated with the model's fitted values, and reaching errors tracked explicit ownership ratings at the individual and trial level.
Keeping up with ourselves: Multimodal processes underlying body ownership across the lifespan
Across participants aged 7 to 80 in a mixed-reality paradigm, body ownership decreased with increasing temporal visuo-tactile and visuo-motor mismatch, and older age was associated with a stronger sense of ownership independent of delay and modality, while delay sensitivity, visual dependence, and proprioceptive accuracy showed no consistent age effects.
Leg muscle vibration modulates bodily self-consciousness: integration of proprioceptive, visual, and tactile signals
Leg muscle vibration, which alters proprioceptive input, modulates bodily self-consciousness and self-identification, with congruent proprioceptive signals supporting and incongruent ones reducing illusory ownership in visuo-tactile paradigms.
Body ownership alterations in stroke emerge from reduced proprioceptive precision and damage to the frontoparietal network
Stroke patients with upper-limb sensory deficits were more likely than age-matched controls to feel ownership over an incongruent virtual hand and integrate it into reaching, an effect modeled as reduced detection of visuo-proprioceptive incongruence proportional to proprioceptive deficit, with additional ownership alterations linked to frontoparietal damage.