Kinesthetic Intelligence Before Language
Before the first word, there was movement. Sheets-Johnstone's framework establishes that by the time an infant produces her first word — typically around twelve months — she has already accomplished cognitive feats of staggering complexity entirely through her body's engagement with the world. She has learned object permanence, causal reasoning, spatial awareness, the properties of materials, the relationship between effort and weight. She has learned all of this without instruction, without language, without representation in any form a computational model would recognize. She has learned it through kinesthetic intelligence. And when language arrives, it does not replace this foundation — it builds on top of it.
In the AI Story
The developmental evidence is extensive and converging. Infants as young as three months show 'violated expectation' responses — they look longer at physically impossible events (an object appearing to pass through a solid barrier) than at ordinary ones. This is not visual intelligence alone; the infant's expectation of solidity is built on her kinesthetic experience of solid objects, the resistance of the crib rail, the firmness of the floor. By six months, infants adjust the force of their reaching to match the apparent weight of objects. This calibration requires the body to have learned, through hundreds of reaches, the relationship between visual appearance and kinesthetic effort. The layers are invisible, stored in the body's patterns of muscular activation and proprioceptive feedback.
Sheets-Johnstone's departure from Piaget is crucial. Piaget recognized the sensorimotor stage as the foundation of cognitive development but treated it as a stage to be transcended — scaffolding that could be removed once higher cognition was built. Sheets-Johnstone argues the scaffolding is load-bearing. Remove it, and the building collapses — not immediately, not visibly, but through progressive loss of structural integrity that manifests as thinning of cognitive capacity. The kinesthetic intelligence of the sensorimotor period is retained beneath language, providing the experiential substrate that gives linguistic concepts their meaning.
For the AI discussion, the implications reshape what language models actually are. If language is a secondary cognitive system built on a primary kinesthetic one, then a technology that operates entirely within language is operating entirely within the secondary system. The model has learned statistical patterns in language produced by bodies that were moving, feeling, kinesthetically engaged with the world. The patterns are real. The kinesthetic substrate that originally generated them is absent. Humans who receive AI outputs are receiving de-animated language and attempting to re-animate it through their own kinesthetic engagement. The re-animation succeeds when the receiver's body is kinesthetically alive. It fails when the receiver's body has been reduced to fingertips on keys.
Origin
Developed in The Primacy of Movement and earlier works, drawing on extensive developmental research and Sheets-Johnstone's critical engagement with the Piagetian tradition.
Key Ideas
Pre-linguistic cognitive architecture. Infants build elaborate understandings of the world before language arrives — through movement, not through instruction.
Piaget's error corrected. The sensorimotor stage is not transcended by later cognition; it is retained beneath it as foundation.
Kinesthetic residue in language. Conceptual vocabulary — grasp, weigh, follow, hold — carries traces of the bodily experiences that originally grounded it.
LLMs as secondary-system processing. Language models operate in the secondary system of language, parasitic on the primary kinesthetic system they cannot access.
Re-animation requires animate receivers. De-animated language re-enters meaning through the bodies of its human readers; if those bodies are kinesthetically depleted, re-animation fails.
Appears in the Orange Pill Cycle
Further reading
- Sheets-Johnstone, Maxine. The Primacy of Movement, Chapters 4–6 (2011).
- Piaget, Jean. The Origins of Intelligence in Children (1952).
- Gopnik, Alison, et al. The Scientist in the Crib (William Morrow, 1999).
- Stern, Daniel. The Interpersonal World of the Infant (Basic Books, 1985).