Method at a glance
- Modality: thermosensory / nociceptive (cutaneous)
- Measures: A demonstration/paradigm probing central integration of thermosensory and nociceptive channels — how the brain computes thermal pain from the balance of cooling and burning signals.
- Procedure: Interlaced innocuous warm (~40°C) and cool (~20°C) bars are presented to the skin simultaneously; neither bar alone is painful, but together they evoke a burning, ice-like pain sensation. Variants: pouring warm water on cold-numbed feet, or cooling the skin after blocking A-fibre (cool-receptor) conduction.
- Also called: thermal grill, Thunberg’s grill, cold-inhibits-pain demonstration
Strengths
- Cleanly dissociates the sensory channels: reveals that thermal pain is centrally computed from the relative balance of thermoreceptive vs polymodal-nociceptive activity, not read off a single ‘pain’ receptor.
- Functional imaging shows the grill reproduces the cortical activation of genuine noxious cold, and unmasks anterior cingulate activation — localizing the affect (‘urgency’) of thermal pain.
Limitations
- A perceptual demonstration rather than a measurement instrument; not used to quantify individual differences.
- Specific to thermosensory/nociceptive integration; its interpretation is tied to Craig’s labelled-line model of lamina I.
Connections
- Key studies: Craig (2002)
Thunberg’s (1896) demonstration, given a modern interoceptive interpretation in Craig (2002) (Box 1). Interlaced innocuous warm and cool bars — neither painful alone — together produce a burning, ice-like pain. See A. D. (Bud) Craig.
What it shows
The illusion reveals two things about the Lamina I spinothalamocortical pathway:
- “Cold inhibits pain.” Below ~24°C, a polymodal C-nociceptive channel is increasingly active but is normally masked centrally by activity in the specific A-fibre cooling channel. The grill (or an A-fibre block) reduces cooling-channel activity, disinhibiting the nociceptive channel → burning pain at merely cool temperatures.
- Thermal pain is a central comparison. The perception depends on the relative balance of two thermosensory/nociceptive labelled lines, computed in the forebrain — not on a dedicated “pain receptor.”
Why it matters for the wiki
The grill is Craig’s evidence that pain and temperature are homeostatic interoceptive channels (see Homeostasis (and feelings as homeostatic sensations)) that signal thermoregulatory distress, and that the ACC carries their affect: functional imaging finds the grill selectively unmasks ACC activation, matching the perception of thermoregulatory urgency. It anchors Craig’s broader argument that pain “is not a binary modality” but the extreme of a homeostatic afferent continuum.
It is also the field’s canonical Interoceptive illusion — a stable misperception of an internal state produced by a resolvable multisensory (here thermosensory/nociceptive) conflict. It differs mechanistically from the central, prior-driven illusions (false-feedback effects, Iodice’s illusion of effort): the grill’s conflict is resolved peripherally/spinally by labelled-line disinhibition, not by a top-down prior. The two ends of the interoceptive-illusion family thus bracket where in the hierarchy an internal percept can come apart from its cause.