Higher-order theories of consciousness
An imported commitment rather than a finding, and the wiki should keep it labelled as such: it does load-bearing work in can-we-know-animal-feelings and where-are-feelings-constituted while being argued for in neither of the sources that rely on it.
The claim
A mental state manifests in consciousness when it is the target of another, higher-order representation of it. Two variants:
- HOP (Higher-Order Perception) — the higher-order state is a perception of the first-order state (Rosenthal 2005 is cited for the family; Armstrong for HOT).
- HOT (Higher-Order Thought) — the higher-order state is a thought about the first-order state.
Scarantino declares himself sympathetic to HO theories and neutral between HOP and HOT, and additionally neutral on whether the higher-order representation must be occurrent or merely dispositional (Carruthers 2000 for the discussion). Those neutralities matter: they mean the framework is not committed to a feeling requiring an active thought about the state, only that one be available.
What it buys an emotion theorist
The consequence Scarantino draws is structural and clean:
Conscious basic emotions involve two distinct mental states: a basic emotion, and a higher order mental state whose content is that one is having a basic emotion. A basic emotion which is not the target of a higher order mental state remains unconscious, but it continues to motivate behavior.
This is what makes unfelt emotions coherent rather than paradoxical, and it is the machinery underneath the whole what-should-emotion-terms-refer-to dispute. If feelings are a separate representational fact about a motive state, then the motive state and the feeling are two things — and the only question left is which one the word “emotion” names.
It also licenses attributing basic emotions to infants and non-human animals “even in the absence of agreed upon dependent measures for studying feelings in creatures without language” — the motive state is what is attributed, and the higher-order question is set aside rather than answered.
The anatomy, and why it matters here
Both principals put the higher-order representation in cortex, and this is where the wiki’s debates hang off the theory.
LeDoux (2012): feelings arise when a global organismic state is represented in a prefrontal/parietal cognitive workspace (Crick & Koch; Dehaene & Changeux; Baars; Shallice). Six ingredients integrate there. The insula plays no role. Dorsolateral prefrontal cortex is a key component — and is largely absent in other mammals, which is exactly what grounds the animal-feelings skepticism.
Scarantino (2018), endorsing LeDoux & Brown (2017): the higher-order representations required for consciousness are cortically implemented by general networks of cognition, “which involve regions in the prefrontal cortex, posterior parietal cortex, and insular cortex that underlie attention, working memory, and metacognition.”
A discrepancy worth flagging
The insula is in the 2017 list and absent from the 2012 account. The wiki records LeDoux (2012) as giving the insula no role in feelings — accurately; that is what the paper does. But his own co-authored 2017 higher-order theory, as reported by Scarantino, includes insular cortex among the general networks of cognition.
Three readings, none decidable from the commentary:
- A shift. LeDoux’s position moved between 2012 and 2017, and the wiki’s where-are-feelings-constituted framing is dated.
- No shift, different job. The insula appears as one workspace component supporting attention/working memory/metacognition — not as Craig’s medium of feeling. On this reading the Craig/LeDoux opposition survives intact, because the disagreement was never about whether the insula is involved but about whether the AIC’s integration of body state is the feeling.
- Scarantino’s gloss. The list may be his compression rather than LeDoux & Brown’s emphasis.
Reading (2) is the most likely and the least interesting; reading (1) would materially weaken can-we-know-animal-feelings, whose whole force comes from DLPFC being absent in most mammals — an insula-inclusive substrate is one other mammals have. Resolving this requires LeDoux & Brown (2017), which is not in raw/. Flagged for a human; recorded on where-are-feelings-constituted and scarantino-2018-basic-emotions as well.
Relation to the wiki’s other accounts of feeling
The HO structure is a two-stage account — a first-order state, plus something that represents it — and the wiki is now full of two-stage accounts descended from two-factor theory:
| Account | First-order state | Second stage |
|---|---|---|
| Schachter–Singer | undifferentiated arousal | cognitive labeling |
| Barrett / Lindquist | core affect | conceptualization + language |
| LeDoux | global organismic state | workspace representation, appraisal, labeling |
| Scarantino | activated basic emotion | higher-order representation |
The odd one out is Craig, and the contrast is the sharpest way to state what is at issue in where-are-feelings-constituted. Craig’s global emotional moment is one-stage: the AIC’s integration is the awareness, not raw material awaiting representation. There is no higher-order state in his model because there is nothing left for it to do.
So the wiki’s central anatomical spine (Craig, and by extension the interoception-as-medium-of-feeling reading) is not a higher-order theory, while both parties to the survival-circuits dispute are. That is a deeper divide than the lobe disagreement, and it is not currently recorded anywhere else: it is the difference between feeling is what a certain integration is like and feeling is what happens when a state gets represented again.