How do you feel? Interoception — the sense of the physiological condition of the body

A. D. (Bud) Craig was a neuroanatomist at the Atkinson Pain Research Laboratory at the Barrow Neurological Institute in Phoenix; his career-long programme was mapping the spinothalamic tract in primates and naming what the body senses itself with. The 2002 review in Nature Reviews Neuroscience synthesised two decades of his lamina-I tract-tracing work into the modern construct of interoception — a reframing of the homeostatic afferent system as a distinct sensory modality with its own thalamocortical pathway terminating in the right anterior insula. Craig died in 2023; obituaries from Barrow named this paper as the definitive statement of the position he built his career on.

Nature Reviews Neuroscience is one of the highest-impact review venues in the field, with editorial filtering at the level the field expects. The piece is a synthesising review, not a primary report — Craig is reorganising his own earlier lamina-I tract-tracing primary work and adjacent literature into a single argument. The argument's two big moves are first, that interoception is not a special case of somatosensation but a separate modality with its own anatomical substrate (the dorsal posterior insula receives it; the right anterior insula re-represents it as feeling), and second, that the lateralisation to the right hemisphere connects the construct to subjective emotional experience in a way other interoceptive frames had not.

The paper sits as the foundational secondary citation point for the modern interoception literature. It is cited by nearly every post-2002 paper on the topic, and the right-anterior-insula hypothesis it advances has become the dominant frame for two decades of fMRI work on emotional awareness, addiction, and psychiatric interoceptive dysfunction. Where it has been contested — Lisa Feldman Barrett's constructed-emotion school reads the insula as multimodal rather than interoception-specific — the disagreement is with Craig's particular model of what the insula does, not with the broader project of operationalising interoception as a measurable construct.

The stake is professional and intellectual. Craig was advancing a model he had built across his entire career; the paper is the synthesis that secured his standing as the founder of the modern interoception field. The piece's life as a citation generator is enormous and worth flagging: this is the paper that gets cited as if it were a primary report, when much of what it claims is a reframing of his own earlier primary work. Read it for the argument; chase the primary lamina-I tract-tracing literature for the evidence.