← Mulloy Morrow · Blog

An essay · Mind / Brain / Body

The Body Beneath the Feeling

Where do emotions come from, and what lies beneath awareness? A guided tour of the rival answers — ending in a staged debate among five thinkers who would not agree about any of it.

01 — The question

You feel before you know why

Most of the time you are having feelings you did not choose and cannot fully explain. A mood arrives before its reason. A person unsettles you before you can say what's wrong. We like to picture a conscious self in the driver's seat, reaching down to pull emotional levers at will — but nearly every serious account of the mind in the last century says that picture is upside down.

Awareness is the thin lit layer. The machinery that produces it runs almost entirely in the dark. That much is near-consensus — you can see it in the fact that consciousness isn't even a simple on/off switch: in dreaming, the body lies still and unresponsive while a whole vivid world is experienced from the inside, which tells you experience comes in modes, not just amounts.

The hard disagreements begin one level down, and they organise this essay as a descent. First, what lies beneath awareness — a buried store of personal meaning, or impersonal machinery with no story to tell? Then, where emotions come from — are they evolved and innate, or assembled from learning and culture? And finally the place those questions physically meet: the body itself, sensed from within. Each layer sets up the next.

AWARENESS — the thin lit layer not on/off but graded; dreaming shows experience has modes WHAT LIES BENEATH? a store of buried meaning, or an impersonal prediction engine? WHERE DO EMOTIONS COME FROM? evolved & innate, or constructed from concepts and culture? THE BODY, SENSED FROM WITHIN interoception — where every theory above physically meets descending into the dark →
Fig. 1  The four layers this essay descends through. Each is more contested — and more physical — than the one above it.
02 — The map

Six positions on two axes

The arguments resolve onto two questions. Where do emotions come from — are they evolved and innate, or built from learning and culture? And what sits beneath awareness — a store of personal meaning, or impersonal regulatory process? Plot every thinker against those two axes and the alliances become visible.

Freud Barrett Panksepp Solms Seth de Botton
Innate · evolved circuits Constructed · learned & cultural Process · impersonal mechanism Meaning · personal content evolved machinery drives & conflicts prediction & categorising narrative self Freud drives + repressed content Panksepp evolved affect systems Solms conscious id · free energy Seth interoceptive inference Barrett constructed emotion de Botton emotional education · culture & belief innate-affect alliance meaning, two routes
Fig. 2  The upper half shares a belief in innate affect (Panksepp, Solms, the Freudian substrate); the lower-left is the mechanistic constructionist corner (Barrett, with Seth as the formal go-between). De Botton anchors the fourth, long-empty corner — emotion as constructed, like Barrett, yet saturated with personal meaning, like Freud. He sits diagonally opposite Freud: both treat feeling as deeply personal, and split on whether it wells up from innate drives or is assembled from culture.
03 — The core dispute

Two unconsciouses

The sharpest fault line runs between Freud and Barrett — and it is not about whether most of the mind runs out of sight (both agree it does) but about what is down there. Freud's unconscious is an archive: specific wishes and memories, actively pushed under by repression and straining to return. Barrett's is a rendering engine: no stored secrets, only a model endlessly predicting the body and the world.

FREUD — the archive awareness (thin layer) repressed wish childhood memory forbidden desire unresolved conflict repression pushes down ↓ ↑ returns as a symptom BARRETT — the engine constructed experience generative model (predicts the body) predictions ↓ error ↑ body signal (the only "data") nothing is stored — it is computed
Fig. 3  The same dark region, two readings. Freud: a locked basement of motivated content. Barrett: a live process with no contents to recover. The clinical word that divides them is motivated — is the hiding purposeful, or just out of view?

Cognitive science already split these two: Freud's dynamic (repressed) unconscious versus the cognitive unconscious — the automatic processing that never surfaces simply because it needn't. Barrett's view is a predictive upgrade of the latter.

04 — The meeting point

Interoception: the body as the source of feeling

Interoception is the brain's sense of the body's own interior — heartbeat, breath, hunger, the churn in the gut, the tightness in the chest. It is mostly unconscious, surfacing selectively as thirst, nausea, or the flutter before you speak. It revives a very old idea, from William James: you do not tremble because you are afraid; the trembling, read by the brain, is a large part of the fear.

The modern turn flips the arrows. Older models had signals rising bottom-up to be read. The predictive account — interoceptive inference, formalised by Anil Seth and Karl Friston — says the brain mostly predicts the body's state from the top down, and the rising signal serves to confirm or correct the guess. The hub where this is integrated is the insula. This single loop is where Barrett's "affect," Solms's "felt uncertainty," and the plain "gut feeling" turn out to be the same machinery.

INSULA compares prediction vs. signal THE BODY heart · lungs · gut · viscera prediction ↓ "this is what the body should be" signal ↑ vagus & spinal prediction error raw affect = valence (good–bad) × arousal (calm–wired) + learned concepts "fear" · "dread" · "excitement" = a named emotion allostasis the "body budget"
Fig. 4  The interoceptive loop. The brain predicts the body's state; the rising signal corrects the prediction; the residual is felt as raw affect. Lay learned concepts over that affect and you get a named emotion. Distort the precision of any arm of this loop and you get a working model of anxiety, depression, or disordered eating.

Three measures, often confused, matter here: accuracy (can you really detect your heartbeat?), sensibility (how attuned you believe you are), and awareness (whether the two match). They dissociate, and the gaps track mental-health risk. Sharpening interoception — through practices such as mindfulness — is one of the few places this science touches daily life, and the bridge back to the idea that emotional skill can be taught rather than merely inherited.

05 — The room

Who's at the table — and why

A debate is only fair if no position wins by absence of an opponent. These six are chosen so the strongest case on each side is actually present. The structure is a set of alliances: Solms & Panksepp defend innate affect and a partly-vindicated Freud; Barrett holds the constructionist line, with Seth owning the shared mathematics; Freud is the ghost they all measure against; and de Botton opens the fourth front — agreeing with Barrett that emotion is constructed, but insisting it is built from meaning, not mechanism, which is the one corner the others leave empty.

F

Sigmund Freud

The archive · 1856–1939

The unconscious as buried, motivated meaning; drives; the conscience as superego. The historical anchor everyone else argues with.

B

Lisa Feldman Barrett

Constructed emotion · b. 1963

Emotions assembled in the moment from bodily affect + concepts. No innate emotion circuits, no archive — a rendering engine.

P

Jaak Panksepp

Affective neuroscience · 1943–2017

The fair counterweight: evolved, innate emotional systems demonstrated across species. The reason the room isn't stacked toward construction.

S

Mark Solms

Neuropsychoanalysis · b. 1961

Uses Friston's free-energy maths to revive Freud — the "conscious id," affect as the ground of consciousness. Bridge and provocateur.

A

Anil Seth

Consciousness science · b. 1972

Co-author of interoceptive inference; the formal middle. Speaks to dreaming, selfhood and the "what is unconscious?" question without committing to either camp.

d

Alain de Botton

Applied philosophy · b. 1969

The meaning-constructionist. Agrees feeling is learned, not innate — but built from culture, expectation, and self-narrative, and answerable to wisdom, not just mechanism.

06 — The debate

A staged debate

SCENE — A dim trail at dusk. A coiled shape lies across the path. The walker's heart slams, breath catches, the body recoils — and half a second later: it is a coiled length of rope. On the table, one shared premise: the brain is a prediction machine that regulates the body. Everyone agrees on that sentence. On what just happened, they agree on almost nothing.
Freud

The rope does not interest me. The snake does. Of all the things a dim coil on the ground might have been, your mind reached for the one charged with dread. Why that image, and so fast? The snake was in you before it was on the path. Perception handed back what you were primed — and perhaps, in older and more personal ways, disposed — to fear.

Barrett

Just so — except what was "in you" is a concept, not a buried symbol. Poor light, sparse data; your brain ran its best guess, and "snake" carried the highest stakes, so it won. The percept and the fear were built together, before any snake was confirmed. Better light arrived, the prediction updated to "rope." Nothing was uncovered. A guess was corrected.

Panksepp

And notice the order of events. The body was already moving — heart, legs, a jump back — before "snake," let alone "rope," reached awareness. That is an evolved alarm system firing on the prediction, a circuit that would rather flinch at a hundred ropes than miss one snake. The affect led; the concept caught up. The feeling did not wait on the word.

Barrett

The alarm fired, yes — but the experience of snake-fear still needed the category. Someone raised where nothing slithers startles at the shape too, yet does not live it as "snake." The circuit gives you the lurch. The concept gives you the snake.

Seth

This is the cleanest case there is, so let me say the strong version. You did not see the path; you saw your brain's best guess of it, caught for an instant out of step with the world. Perception is a controlled hallucination — built top-down from prediction, reined in by sensory error. And the proportions are humbling: into the brain's visual relay, only about a tenth of the signal comes from the eyes. The rest is the brain's own model. The "rope" correction is that thin incoming stream doing its work in real time.

Solms

Strip it to the physics and the camps converge. A path predicted safe, then a sudden shape: an enormous prediction error, a spike of felt uncertainty. That spike is the fear — and it registers first in the deep, body-regulating brain, before cortex names a thing. The misread outside and the alarm inside are one loop. The body reacts to the guess, not to the world. Freud's drives were a clumsy early sketch of exactly this regulator.

Barrett

On the loop, agreed — and Mark, we run on the same equation. We part on what the guess contains. You hear drive, conflict, personal meaning. I hear stakes-weighted statistics with no story to tell until a concept supplies one.

de Botton

"No story to tell" — there, Lisa, is where I leave you, though we agree the feeling is made and not buried. The rope is the trivial case. The fears that actually run a life — that I am unlovable, that I have wasted my best years, that I am falling behind people I did not choose to race — are constructed too, exactly as you insist. But they are made of almost nothing but story: ideas absorbed from a culture, expectations no one set out deliberately to hold, comparisons we were schooled into. Strip the narrative from those and there is no residue of "stakes-weighted statistics" underneath. The meaning is not a label fixed onto a body-state. For the feelings that matter most, the meaning is the emotion.

Barrett

I'd still say the story rides on an affective current — the body sinks before "failure" names the sinking. But I grant you the interesting half: the richer and more cultural the concept, the more the construction floats free of the flesh. On the rope, biology does most of the work. On a wasted decade, your culture does.

Freud

That the guess reached for a snake and not a rope is the meaning. The content of our errors is a kind of confession.

Seth

Or simply a prior. Snakes have killed more of our ancestors than ropes have; the bias is paid for in evolutionary time and needs no biography. This is the crux, and worth saying plainly: the maths is neutral. It can tell us the brain predicts and corrects — it is far quieter on whether the prediction is a meaning or a mechanism. Some of this disagreement is interpretation wearing the costume of evidence.

Panksepp

Both can hold. The prior is evolutionary; the particular dread may be personal. The tissue carries the first — your buried wishes I leave to the clinic, Sigmund — but your older claim, that the feeling animal precedes the reasoning one, the neuroscience has borne out.

Barrett

And the leverage is in the part that is learned. Widen the light, sharpen the concepts, and the same shape stops being a snake on sight. We are not only the prediction — we can school it. That is testable, and it is hopeful.

de Botton

There, at last, we shake hands — the feeling is teachable. You would retrain it with finer concepts and graded exposure; I would do it with Seneca, with a novel, with a truer picture of what a life actually owes a person. Same wager, Lisa. We differ only on the curriculum — and mine has a four-thousand-year head start.

Seth

Which returns us to the path. Notice no one here thinks the snake was simply there, waiting to be seen. We split on what produced it — a primed symbol, an evolved alarm, a stakes-weighted prior, a concept reaching past its data, a story we were handed and never examined — and on whether the correction reveals a hidden basement or just a better-lit trail. Five readings, one rope.

de Botton

And a consoling reading to end on. Most of what we dread is the rope that looked like a snake — the unanswered message, the colleague's glance, the silence after an interview — each read in poor light by a frightened animal. Whatever the machinery beneath it, Seneca drew the lesson worth keeping two thousand years ago: we suffer more often in imagination than in reality. You do not need the fMRI to start using that tonight.

Where it actually lands

No one wins, and that is the honest result. There is real consensus that the brain predicts and corrects — that we meet the world through a best guess, with the senses feeding back only enough to keep the guess honest. The disagreement has migrated upward: from "is perception a construction?" (settled: largely yes) to "what is the construction made of?" Mechanism, says Barrett; buried meaning, says Freud; inherited story, says de Botton — three answers that no longer divide neatly into innate versus learned. Seth's caution is the one to leave with: the same equation underwrites all of them, so be suspicious when interpretation is dressed as proof.

07 — The state of the evidence

What the science currently supports

Are we conscious while dreaming?

Yes — a real but stripped-down mode of consciousness, with self-monitoring dialled down. Lucid dreaming is the hybrid case. Well supported.

Does emotion have fixed biological "fingerprints"?

Largely no — robust meta-analytic support for Barrett here. But "no clean fingerprint" does not by itself prove "fully constructed, not innate." That leap is still contested.

Are there innate affective systems?

Strong cross-species evidence (Panksepp) for evolved subcortical systems. How much of human emotional experience they fix is the open question.

Is the Freudian repressed unconscious real?

The literal recoverable-trauma archive: weak and embattled. The broader claim that motives operate outside awareness: comfortably mainstream.

Is interoception foundational to emotion?

Strong and growing consensus, with accumulating clinical links to anxiety, depression and eating disorders. The hottest area of the four.

Does the free energy principle settle it?

No. It is powerful but very general — flexible enough to fit Freud and anti-Freud, which is either its depth or its weakness, depending on temperament.

08 — The frontier

Where the questions are heading

The disagreements in this essay are not closing. They are sharpening — turning from clashes of worldview into specific, testable edges. Five of them are where the field is pushing now, and each inherits a tension from the pages above.

SETTLED CORE the brain predicts & regulates the body 1 · Interoception as medicine retraining the body-signal loop to ease anxiety, depression, trauma 2 · Does consciousness begin in the body? brainstem affect vs. cortical theories of awareness 3 · Can the free-energy principle be falsified? a unifying law, or a frame flexible enough to fit anything? 4 · Where the self comes from the located “I”, woven from interoceptive signal 5 · Does artificial emotion need a body? can a mind feel with no interior to sense and defend?
Fig. 5  A settled core — the brain predicts and regulates the body — branching into the five open questions now carrying the field forward.

1 · Interoception as medicine

The most immediate edge is clinical. If emotion is built on the brain's reading of the body, then many disorders of feeling may be disorders of that reading — specifically of its precision, how heavily the brain weights a bodily signal against its own prediction. Anxiety increasingly looks like a system that trusts a racing heart too much and spirals; depression, eating disorders, and trauma are being re-described in the same currency. The open question is whether the loop can be deliberately retrained — through interoceptive exposure, breathwork, biofeedback, and contemplative practice — and whether that yields durable change or only temporary relief. This is where the abstract science has to earn its keep.

2 · Does consciousness begin in the body?

One camp wagers that feeling is not a late cortical embellishment on thought but the oldest and most basic form of consciousness, rooted in the brainstem's monitoring of the body's survival. If that is right, the search for the seat of awareness has been looking too high in the brain. Cortical theories — global workspace, higher-order thought — locate experience instead in elaborate cortical processing. Resolving this decides something concrete and strange: whether a creature could be conscious with almost no cortex at all, on the strength of its feeling brain alone.

3 · Can the free-energy principle be falsified?

This is the unease beneath the whole debate. The same equation underwrites the constructionist case and its neo-Freudian rival — a flexibility that is either the signature of a deep unifying law or the warning sign of a framework that can be bent to fit anything and therefore forbids nothing. The frontier here is not empirical but methodological: can the principle be sharpened into discriminating, falsifiable predictions, or does it remain a beautiful lens rather than a theory? Until it can rule something out, it risks explaining everything and nothing in the same breath.

4 · Where the self comes from

The steady hum of interoception may be the thread from which the sense of self is woven — the constant background feeling of being a living body that makes “I” more than an idea. The evidence comes mostly from its breakdowns: depersonalisation, where people feel unreal or detached from their own bodies; body-ownership illusions that can relocate the felt boundary of the self; even hints that the mind's eye is anchored in bodily signal. The question underneath is whether selfhood is fundamentally a story the brain narrates, or a sensation it never stops having.

5 · Does artificial emotion need a body?

And the question that turns the whole inquiry forward. If feeling is the brain's account of a body it must keep alive, then a system with no body to lose, no balance to defend, no interior to sense may be unable to feel — or perhaps even to understand — in any sense we would recognise, however fluent it becomes. Or the body may prove incidental: one substrate among many for the same predictive principles, reproducible in silicon. This is no longer philosophy alone. It shapes what we build, and what we should expect from, machines that increasingly speak the language of emotion without the viscera that, in us, give rise to it.

The shape of it

Notice what the five have in common. The twentieth century asked what is hidden in the mind? The frontier asks what is the body doing in there — and whether a mind without one could feel at all. The argument has not been settled so much as relocated: out of the basement of repressed meaning, and down into the living tissue that does the predicting.