Multisensory integration
The formal definition the wiki uses is Quigley et al.’s (2021) glossary entry: sense data from two or more modalities combined “to form a product that is distinct from, and thus cannot be easily ‘deconstructed’ to reconstitute, the components from which it is created.”
Why this needed a page
The wiki has used the term since its first ingests, always in its traditional sense — integration across two exteroceptive modalities. The rubber-hand-illusion is vision plus touch. The classic self literature (Tsakiris 2017) is visuo-tactile.
Quigley et al. widen it deliberately, and say so: “we use the term multisensory integration in a broader way than the more traditional sense of integration across exteroceptive modalities.” Three cases, not one:
| type | example |
|---|---|
| exteroceptive × exteroceptive | the classical RHI; visuo-tactile self |
| interoceptive × interoceptive | glucose and proinflammatory cytokines jointly signalling energetic state; ghrelin and leptin as paired energy signals |
| interoceptive × exteroceptive | vision and ghrelin in feeding; bladder fullness and proximity to one’s front door; cardio-visual synchrony in the RHI |
The consequence is structural rather than empirical. If integration crosses the boundary, then interoception is not a channel that terminates somewhere and then influences cognition and emotion — it is a party to perception at the point where perception is assembled. The interoceptive signal and the exteroceptive signal are combined into something that is not decomposable back into either.
Where the wiki already had this without the name
Several pages turn out to be about interoceptive-exteroceptive integration:
- experience-of-body-ownership and rubber-hand-illusion: cardio-visual synchrony (Suzuki et al. 2013; Aspell et al. 2013) enhances ownership. That is a heartbeat integrated with a visual flash — an interoceptive×exteroceptive pair, and the wiki’s oldest instance.
- The precision competition. Quigley et al.’s framing of Tsakiris et al. (2011) — lower interoceptive-sensitivity → stronger RHI — is that “in the absence of accurate interoceptive representations, one’s model of self is dominated by exteroceptive input.” Integration here is a weighted competition, with the weights set by relative precision (Ainley et al. 2016). See predictive-coding.
- bodily-sensation-maps: a felt topography is arguably an integrated product of interoceptive state and an exteroceptively-anchored body schema; the wiki’s standing question about whether the maps report a body or a concept is partly a question about what got integrated.
- affective-touch: the sharpest case, because the two modalities arrive through the same skin. Hairy skin co-mingles myelinated (exteroceptive, discriminative) and unmyelinated CT (interoceptive, affective) afferents, so an ordinary caress is one stimulus delivering both — Quigley et al.’s stated reason that touch “blurs the classical distinction.”
Where it does new work
Energy regulation requires it. Interoceptive energy-status cues do not obligatorily trigger feeding; they must be integrated with prior learning and with exteroceptive information. Foraging is the clean case: the interoceptive deficit signal that initiates it is useless without visuospatial and olfactory information about where food is. The brain regions where both kinds of signal converge (orbitofrontal cortex, nucleus accumbens, lateral hypothalamus) are also reward/affect regions — so the same integration can influence several behavioural outcomes at once.
Memory is where it is best evidenced. The gut-hippocampal findings are an integration story: vagal energy-status signals prime the hippocampus to encode the visuospatial and contextual (exteroceptive) details of a meal. Interoceptive signal in, exteroceptive memory out, one structure.
Urgency, too. Bladder urgency integrates fullness with environmental cues — the front-door effect — and is amplified by stress, anxiety and depression.
The honest status
Programmatic. Quigley et al.’s own summary is that “fundamental questions remain regarding multimodal integration that involves interoceptive modalities” and that “future empirical work is required to better understand how and where multimodal interoceptive integration occurs.” The wiki should hold the widened sense as a useful reframing with illustrative support, not as a characterized mechanism: no locus is identified, no computation is specified, and the exteroceptive integration literature’s quantitative apparatus (optimal cue combination, reliability weighting) is not applied to any interoceptive case here.
The one place where a mechanism is specified is precision-weighting under predictive-coding — which is a proposal about how integration is governed, made by a different literature (seth-2013-interoceptive-inference, Ainley et al. 2016), and which Quigley et al. adopt for the self section without extending it to feeding, memory or micturition.