Cortical somatotopy (the homunculus)
Somatotopy is the principle that the body is mapped in order onto cortex: adjacent body parts project to adjacent cortical territory, producing the distorted little body — the homunculus — that Penfield and Boldrey (1937) drew across the primary motor and sensory strips. The canonical maps are M1 (motor) and S1 (somatosensory) on the lateral surface; the SMA carries another; and the lateral parietal lobe holds a whole-body somatosensory map (Sereno & Huang 2006).
Why an interoception wiki has this page
Not because somatotopy is interoception — it is its exteroceptive cousin — but because the word “body map” is dangerously overloaded in this wiki, and this page exists to hold three distinct things apart:
| “body map” sense | maps what | measured by | wiki page |
|---|---|---|---|
| cortical somatotopy | where on the body surface a touch/movement is (or which part is moving) | fMRI/stimulation topography | this page |
| felt bodily sensation map | where in the body a person reports feeling an emotion | self-report painting (embody) | Nummenmaa’s emBODY maps |
| interoceptive map | the state of internal organs (cardiac, visceral, homeostatic) | heartbeat tasks; insula gradient | interoception / insula |
These come apart. Somatotopy is a spatial layout of the skin and musculature; the felt emotion map is a phenomenological report that need not track any single cortical map; the interoceptive map is of the viscera and lives in the posterior→anterior insula gradient, which Craig explicitly separates from the parietal somatosensory cortices. Confusing them is the exact error the bodily-sensation-maps page warns about when it notes that “fingerprints” is used in two literatures. This page is the pin that keeps the categories from collapsing into each other.
The new precuneus homunculus (Zeharia et al. 2019)
Zeharia, Hofstetter et al. (2019) add a homunculus the field had forgotten. Penfield and Jasper (1954) predicted a “supplementary sensory” body map in the medial parietal wall (the precuneus), mirror to the SMA, but hedged it and it lapsed. Using phase-encoded fMRI over 20 moved body parts, Zeharia et al. recover it: an anterior-to-posterior, dorsal-to-ventral, toes-to-tongue gradient in the precuneus, mirror to the SMA with the M1 leg area between them.
Two features matter beyond the mapping itself:
- It sits in a “cognitive,” non-primary region. Somatotopy is not confined to sensorimotor strips; it turns up in the precuneus, an associative/self-processing hub. The same group had reported a homunculus in the insula (Zeharia et al. 2012) — a topographic body layout in the interoceptive cortex. Body-mapped organization in “higher” regions is the throughline, and the insula case is where somatotopy and interoception genuinely touch: the interoceptive cortex is itself topographically organized by body region.
- It connects to the extrastriate body area (EBA). Every body-part sector of the precuneus map is functionally coupled to the EBA — the region for seeing body parts. A motor/sensory body map wired to a visual body map: somatotopy as a cross-modal, not purely somatic, principle.
The insula homunculus, demonstrated for pain (Brooks et al. 2005)
The bullet above cites the Zeharia group’s 2012 “insula homunculus” as the case where somatotopy and interoception touch. That claim now has a first-hand, modality-specific source in the wiki. Brooks et al. (2005) — with Craig as co-author — mapped painful heat to the face, hand and foot in 14 people at high-resolution 3T, and found an orderly body map in the dorsal posterior insula: face anterior and lateral, foot medial in the circular sulcus, hand lateral (only that one region of five carried a stimulus-site-dependent map; SII did not). The coordinates match human depth-electrode stimulation and recording.
This is the strongest thing on this page for the “body-mapped organization in higher regions” throughline, because it is body-mapping in the interoceptive cortex for a lamina I modality — the VMpo→dpIns pathway laid out by body part, which is what Craig’s labelled-line account predicts. It keeps the disambiguation table honest in the other direction too: this is a genuine cortical somatotopy of a homeostatic afferent, distinct from both the felt-emotion map and the S1/precuneus surface maps, all three of which happen to be “body maps.”
Relation to felt emotion maps — a caution, not a bridge
It is tempting to link a cortical body map to the felt bodily map of emotion, but the wiki records no evidence that the emBODY topographies (nummenmaa-2014-bodily-maps, volynets-2020-cultural-universality) are read off any somatotopic cortical map. The felt maps are cross-culturally stable phenomenological reports whose physiological commitment the authors themselves disclaim. A somatotopic homunculus tells you where cortex represents the toe; it does not tell you why sadness is felt as heavy legs. Keeping this page and bodily-sensation-maps cross-linked but distinct is the point.