Kinesthesia is the body's own sense of its movement — the felt awareness of muscular effort, limb position, the unfolding of motion through time. Distinct from proprioception (the sense of static position), kinesthesia is the dynamic sense of moving through space, the experiential dimension of the body's own activity. For Sheets-Johnstone, kinesthesia is not a minor sensory channel but the primordial mode of cognitive engagement with the world — the sense through which the infant first learns what objects are, what space means, what it is to be a body among other bodies. Every subsequent cognitive capacity, however abstract, retains its kinesthetic origins. To grasp an idea is not merely to use a metaphor; it is to activate neural systems that overlap with those used in the physical act of grasping.
Kinesthesia operates continuously, mostly beneath awareness, providing the felt quality of embodied experience. When a cognitive event carries kinesthetic accompaniment — the release that follows solving a hard problem, the postural tension of struggle, the whole-body recognition that something has clicked into place — the event is more deeply encoded, more durable, more resistant to displacement by subsequent information. Sheets-Johnstone's framework identifies this kinesthetic accompaniment not as epiphenomenal but as constitutive: it is part of how cognition works, not decoration on top of cognition.
The cognitive vocabulary of human languages is thick with kinesthetic residue. To follow an argument, to weigh evidence, to hold a position, to stumble upon a discovery. Conceptual metaphor theorists like Mark Johnson and George Lakoff documented these patterns and argued they reveal the embodied structure of thought. Sheets-Johnstone pushes further: these are not metaphors in the decorative sense but structural features, traces of the kinesthetic origins of the concepts themselves. The word grasp applied to ideas is not a transfer from the physical to the abstract — the abstract concept was built on the physical experience to begin with.
For large language models, the word grasp is a token with learned statistical associations. The model can use it correctly, distinguish its uses, deploy it with rhetorical precision. What the model cannot do is carry the kinesthetic substrate beneath the word — the infant's hand closing around an object, the muscular effort of holding, the felt resistance of material that will not yield. The statistical association is there. The experiential grounding is not. Sheets-Johnstone's framework predicts that this absence, invisible in individual outputs, compounds across sustained human-AI collaboration as the human partner's own kinesthetic life thins to match the machine's disembodied processing.
The term comes from the Greek kinein (to move) and aisthesis (perception) — literally, the perception of movement. Sheets-Johnstone recovered and extended the concept from Edmund Husserl's phenomenology, where kinesthesia played a foundational role in the constitution of spatial experience, and from developmental research showing that kinesthetic sensitivity is present at birth and provides the primary mode through which infants encounter the world before language arrives.
Felt motion, not positioned stillness. Kinesthesia is the dynamic sense of moving, as opposed to proprioception's static sense of position — though the two work together continuously and are often treated as a single sense.
Cognitive infrastructure. Kinesthetic experience is not incidental to cognition but foundational: it provides the experiential substrate on which linguistic concepts are built and without which those concepts become thin.
Retention beneath language. When language arrives, it does not replace kinesthetic understanding; it overlays it, drawing on kinesthetic experience for its depth and meaning.
Encoding through accompaniment. Ideas that arrive with kinesthetic accompaniment — the bodily sense of discovery, the felt shift of recognition — are retained more deeply than ideas received as pure information.
Vulnerability to screen work. The narrow kinesthetic range of typing and screen interaction threatens to atrophy the broader kinesthetic intelligence on which deep cognition depends.
Critics from the computational theory of mind argue that kinesthetic accompaniment is real but inessential — an epiphenomenal byproduct of cognition rather than part of it. Sheets-Johnstone's response draws on developmental and neurological evidence: the neural systems for abstract reasoning overlap substantially with those for motor planning and spatial orientation, and patients with proprioceptive or motor deficits show corresponding cognitive deficits that cannot be explained on a purely computational account.