9 Chapter 4.2 — Pain: The Sense That Tells You Something Is Wrong

Two Pains, the Limbic Detour, and Sensation as Valuation

9.1 Why pain gets its own chapter — and why it sits here

Before we begin, a word about where this chapter lives, because the placement was a real decision and the reasons are themselves instructive.

You might reasonably expect pain to be a section inside the somatosensation chapter. It is, after all, a body sense: its receptors sit in the skin, its afferents enter through the dorsal roots with their cell bodies in the dorsal root ganglia, and one of its two ascending pathways — the anterolateral system — is one we already built in Chapter 4.1. So why pull it out? Two reasons. The first is simply weight: pain carries too much of its own conceptual machinery — a two-speed peripheral system, a detour through the emotional brain, a built-in opioid brake, a tangle of genuinely unsolved clinical puzzles — to fold into a chapter that is already doing the work of touch and proprioception. It would swamp the boat.

The second reason is the more interesting one, and it is also why pain could not simply live back in Unit II with the rest of interoception. Pain depends on the spinal-cord pathways we had to lay out first. You cannot understand why a cord injury dissociates pain from touch (Brown-Séquard syndrome), or why “first pain” and “second pain” arrive at different times and feel utterly different, without the anterolateral-versus-dorsal-column anatomy from the previous chapter in hand. We wanted to keep the spinal cord together — to build it once, properly — and pain is what you can finally explain once it is built. So pain comes after somatosensation, riding hardware now in place, but before we leave the body for the distance senses.

There is a deeper payoff to giving pain its own chapter, and it is worth stating at the outset so you can watch for it. Every sensory system we have discussed, and every one still to come, is in the business of modeling the world — building a re-projected representation the animal can act on. Pain is different. Pain does not primarily tell you what is out there; it tells you that something is bad, and commands you to do something about it. It is the most nakedly evaluative of the senses — sensation fused with valuation and a motivational push. That makes it the natural hinge between this unit, on sensing, and the units to come, on acting and on value. Pain is where the sensory side of the brain hands off to the motivated side.

A note on words: nociception is not pain

It will help to separate two things the everyday word “pain” runs together. Nociception is the neural process of detecting and signalling actually or potentially damaging stimuli — the receptors, the firing, the ascending traffic. Pain is the conscious, unpleasant experience. The two usually go together but can come apart in both directions: nociception without pain (you can be injured under anaesthesia, signals ascending, no experience), and — more tellingly — pain without nociception (the phantom-limb ache from a limb that no longer exists, or the allodynia we will meet, where a gentle touch is felt as pain though nothing damaging is happening). Keeping the two words apart is not pedantry; it is the whole reason this chapter ends where it does, because the gap between them is exactly where pain stops being a detector and becomes an evaluation.

9.2 Two pains

We have already met the body’s nociceptors in passing: the free nerve endings, the simplest receptors of all — bare axon terminals in the skin, with no specialized corpuscle around them. When tissue is damaged, the injury itself does the activating. A splinter breaking the skin releases a chemical broth — ATP, protons that acidify the area, prostaglandins, bradykinin and histamine from the immune response — and these molecules open channels on the free nerve ending, among them the TRPV1 channel we met in the last chapter (the same one capsaicin and noxious heat open). So the receptor end of pain is, fittingly, the least elaborate apparatus in the body, doing the most consequential job: announcing that the tissue is under threat [@basbaum2009cellular].

Now the part that organizes the whole chapter. Pain is not one sensation but two, carried on two different fibers, arriving at two different times, feeling like two different things — and, as we will see, heading to two different parts of the brain.

First pain is sharp, bright, and precisely located — the immediate sting of the pinprick. It rides the Aδ fibers: thinly myelinated, and therefore relatively fast.
Second pain is dull, aching, burning, and poorly localized — the throb that wells up a moment later and lingers. It rides the C fibers: unmyelinated, the slowest wires in the body.

You can feel the split yourself. Stub your toe hard and there is a clean instant of sharp — then a pause — then the deep sickening ache rolls in and stays. That sequence is not psychological; it is two populations of axon conducting at different speeds, the fast Aδ message arriving well ahead of the slow C message. The gap between them is, quite literally, the difference between a myelinated and an unmyelinated axon.

Figure 4.2.1. The two-pain sequence: a sharp, well-localized “first pain” on faster myelinated Aδ fibers, followed after a perceptible gap by a dull, aching, poorly localized “second pain” on slow unmyelinated C fibers. [Schematic to source — ideally a timeline showing the two arrivals.]

Here is a question worth sitting with for a moment — your instructor finds it genuinely strange. Why would the body put its warning sense on slow wires? You would think pain would be the thing you most want to know instantly. We answered half of this in the last chapter: the truly time-critical response to a noxious stimulus — yanking your hand off the stove — is handled below the brain by spinal reflexes, fast enough that your hand is already moving before the pain signal reaches cortex. So the conscious experience of pain can afford a delay. But the two-speed design adds something more. The fast first pain does the spatial job — where is the damage, so I can attend to it — while the slow second pain does the motivational job — the lingering wretchedness that makes you protect the injury, rest it, and remember not to do that again. The two pains are not redundant. They are a division of labor between localizing a threat and caring about it. Hold onto that division, because the brain honors it all the way up.

9.3 Two destinations: the localizing pain and the suffering pain

In Chapter 4.1 we followed the anterolateral system from the dorsal horn, across the midline, up to the thalamus. Now we can be more precise, because the two pains diverge once they are inside the brain — and the divergence is the heart of this chapter.

First pain takes the familiar route. Fast Aδ signals ascend the anterolateral system to the ventral posterior lateral nucleus (VPL) of the thalamus — the same body-sense relay from the last chapter — and from there to primary somatosensory cortex (S1), the homunculus. This is exactly the machinery touch uses, and it is why first pain inherits touch’s best feature: spatial precision. Routed onto the body map, first pain can be pinpointed — that spot, on that finger — for the same reason a touch can. First pain is, in a sense, just another input to the world-modeling system, telling you where.

Second pain takes a detour into the emotional brain. The slow C-fiber signal projects, by way of medial thalamic nuclei, not to the body map but to two regions we flagged in Chapter 4.1 as the cortex of feeling and interoception: the insula and the anterior cingulate cortex (ACC). These are limbic structures — Broca’s old “rim” of the brain — and they are emphatically not in the business of telling you where on the body something is. They are in the business of how bad it is and how it makes you feel. Second pain, routed here, is the suffering — the affective, motivational, unpleasant heart of pain, the part that makes pain matter rather than merely register.

Figure 4.2.2. The divergence of the two pains within the brain: fast first pain via VPL to S1 (the sensory-discriminative, “where” pathway, often called the lateral pain system), and slow second pain via medial thalamus to the insula and anterior cingulate (the affective-motivational, “how bad” pathway, the medial pain system). [Schematic to source.]

This double destination is one of the most important ideas in this unit, so let me state its consequence plainly: the sensory and the affective dimensions of pain are carried by partly separate machinery, and can be pulled apart. That is not just a tidy anatomical story; it is demonstrable. In a striking experiment, Rainville and colleagues used hypnotic suggestion to change how unpleasant a painful stimulus felt while holding its intensity fixed — and watched activity in the anterior cingulate track the unpleasantness, while activity in S1 tracked the raw intensity, the two moving independently [@rainville1997pain]. The “where” and the “how bad” really are dissociable, and the cingulate is doing the suffering.

The clinical shadow of this is haunting. There was, in the era of psychosurgery, an operation — cingulotomy — performed on patients with intractable pain, in which the anterior cingulate was lesioned. The reported result was not that patients stopped feeling the pain. They reported that the pain was still there — they could still locate it — but that it no longer bothered them. The sensation survived; the suffering was cut away. Whatever one thinks of the procedure (and it raises every ethical flag), the phenomenology is the cleanest possible demonstration of this chapter’s thesis: the badness of pain is a separable thing, built in the cingulate, layered onto a sensory signal that can persist without it. Pain is sensation plus an evaluation, and here is the evaluation coming apart from the sensation under the knife.

Note to author — replication check

Two claims in this section are doing a lot of work and both warrant your scrutiny on replication grounds.

The Rainville (1997) hypnosis study — ACC tracks unpleasantness while S1 tracks intensity — is a single elegant result from one group. The broader claim it supports (sensory/affective dissociation) has convergent support from other directions, but I could not confirm that this specific hypnosis-dissociation result has been independently reproduced. I’ve presented it as “demonstrable”; you may want to soften to “one influential study suggested,” or cite a converging line instead of leaning on this one.

The cingulotomy phenomenology — “pain still there but no longer bothers me” — is a frequently-retold characterization from older, uncontrolled clinical case series, exactly the genre you’ve grown wary of. The dissociation it illustrates is real and supported elsewhere, but the vivid patient quote is closer to lore than to a controlled finding. Consider keeping it explicitly as a clinical impression (“patients were reported to say…”) rather than as evidence, or dropping it if it’s carrying more weight than it can bear. The chapter’s thesis survives without either of these; they’re rhetorically powerful but evidentially soft.

9.5 The brain pushes back: descending control and the built-in opioid

So far the story runs upward — receptor to cord to brain. But one of the most important facts about pain is that the traffic runs downward too, and the brain can turn its own pain signal down.

The same limbic structures that receive second pain — the anterior cingulate and insula, by way of the midbrain and brainstem — send projections all the way back down to the dorsal horn of the spinal cord, where they can gate the incoming nociceptive signal before it ever ascends. This is descending modulation, and at the bottom of it sits something remarkable: the synapse onto the ascending pain neuron can be inhibited by enkephalin, an endogenous opioid — an opioid the body makes itself. We have, wired into the cord, a built-in morphine system. (This is also, not coincidentally, where opiate drugs act, hijacking a modulatory circuit that was already there.)

You have felt this system work. The soldier who does not notice a serious wound until the battle is over; the athlete who finishes the play on a broken bone; the way fear or focus or even a placebo can blunt pain — these are descending control turning down the gain on nociception when the brain “judges,” for better or worse, that there are more urgent priorities than attending to the injury. Pain, in other words, is not a faithful meter of tissue damage. It is a regulated signal, adjustable from above according to context. This was the deep insight behind Melzack and Wall’s gate-control theory in the 1960s — that the dorsal horn is a gate where other inputs and descending signals modulate whether pain gets through, rather than a simple relay [@melzack1965gate]. The details of their original circuit have been revised many times since, but the core idea — that pain is gated and modulated, not merely transmitted — has only been strengthened, and it is the conceptual root of the descending-opioid story.

This regulated quality is the final piece of the chapter’s thesis. A pure detector reports what is there. Pain is not a pure detector; it is a valuation the brain computes and continually adjusts — which is exactly why it can be turned down when the valuation changes, and, as we are about to see, why it can go wrong.

9.6 When pain comes apart from injury

If pain were simply a readout of tissue damage, the following phenomena could not happen. That they happen — and that they are clinically real, not curiosities — is the strongest evidence that pain is a constructed, evaluative experience rather than a meter.

9.6.1 Referred pain

Pain is often felt somewhere other than where the damage is. The classic case is the heart: a heart attack is frequently felt not in the chest but in the left arm, or the jaw. This is referred pain, and there are whole body maps of where visceral problems tend to be felt on the surface.

The textbook explanation is convergence: visceral and somatic pain afferents synapse onto some of the same second-order neurons in the dorsal horn, and because the brain has vastly more experience interpreting pain from the skin than from the heart, it misattributes the visceral signal to the body surface that usually drives that neuron. Your instructor will register a dissenting note here, in keeping with how we do things: I find the convergence story weaker than it is usually presented. Convergence is everywhere in the nervous system, and we are not generally confused about which input we are receiving — so convergence alone cannot be the whole answer to why this misattribution is so systematic [@murray2009referred]. I do not have a better mechanism to offer you, and I would rather say so than pretend the standard account closes the case. What referred pain does establish, beyond dispute, is the chapter’s theme yet again: what matters is where the brain decides the pain is coming from, not where it actually originates — the same lesson as the rubber-hand illusion in the last chapter, now in a clinical key.

9.6.2 Allodynia

After certain injuries, tumors, or neuropathies, a sensation that is normally entirely innocuous — the brush of clothing, the stroke of a comb through hair — begins to be experienced as pain. This is allodynia: pain produced by a non-painful stimulus. Here the gap between nociception and pain is total — there is no noxious stimulus, no tissue damage from the gentle touch, and yet the experience is genuinely, sometimes severely, painful. The wiring has been altered, by mechanisms still not well understood, so that the pain-construction machinery is triggered by the wrong input. Allodynia is pain without nociception, laid bare — and a reminder that the experience can be manufactured by the system even when the body is, at that spot, perfectly fine.

9.6.3 Congenital insensitivity, and a clue in the nose

And the reverse: a small number of people are born unable to feel pain at all. The most-studied form is caused by loss-of-function mutations in a single gene, SCN9A, which encodes a voltage-gated sodium channel, Na_v1.7, that nociceptors depend on to fire [@cox2006scn9a]. Knock it out, and the pain signal cannot get off the ground.

You might think this a gift — who would not want to be free of pain? It is the opposite. People with congenital insensitivity to pain accumulate unnoticed injuries: they bite through their tongues, break bones and keep walking, scald themselves without flinching, and develop joint destruction from damage they never feel and so never protect. Their life expectancy is shortened, because pain turns out to be a guardian we cannot do without — the warning that sends us to the doctor, the lesson that teaches us not to repeat the harm. The condition is the clinical proof that pain, for all its misery, is adaptive: a body that cannot suffer cannot protect itself.

There is a lovely diagnostic detail here that your instructor delights in, because it shows how a careful clinician reads the body. People with this SCN9A form of congenital pain-insensitivity are also frequently anosmic — they cannot smell [@weiss2011scn9a]. Why on earth should painlessness travel with a missing sense of smell? Because the very same channel, Na_v1.7, is also the dominant sodium channel in the olfactory sensory neurons, which need it to transmit odor signals to the brain. One broken gene, two seemingly unrelated deficits — and the “nose problem” is a clue pointing straight at the gene. (There is even a final twist that ties back to our descending-opioid story: simply blocking Na_v1.7 with a drug does not reproduce the total painlessness of being born without it, and part of the reason appears to be that the lifelong absence of the channel ramps up the body’s own opioid system. Pain-free people are, in a sense, permanently and internally medicated [@minett2015nav17].)

9.7 “I feel your pain”: a live controversy worth getting right

I want to end the substance of this chapter on a genuinely unsettled question, because it is both fascinating and a good test of how to hold evidence carefully — and because the field has moved in the last few years in a way that should change how this is taught.

Start with the observation, which is robust. If you watch another person in pain — a clip of someone slamming their hand in a door, a photograph of an injury — you activate some of the same regions that light up when you are in pain, particularly the anterior cingulate and insula, the affective second-pain regions. “I feel your pain,” it seemed, is close to literally true: empathy for pain appeared to run on the brain’s own pain machinery, sharing the affective system but not the sensory-discriminative one (you do not feel it located on your body). This was an influential and attractive idea, and as recently as a decade ago it was often taught as more or less established.

Here is where it gets interesting, and where I have to update the story. Two developments have complicated it, and neither is a footnote.

First, the regions in question — the cingulate and insula — turn out not to be specific to pain at all. The anterior cingulate in particular activates for a startling range of things: physical pain, yes, but also breathlessness, simple finger movements, emotionally arousing images of all kinds. A strong reading of the evidence is that what these regions encode is not “pain” but salience — the detection of anything important and demanding of response — of which pain is one powerful instance [@iannetti2013salience]. If that is right, then the overlap between feeling your own pain and watching another’s might reflect a shared salience/arousal response rather than a shared pain representation specifically. The inference “same region active, therefore same experience” — what is sometimes called reverse inference — is exactly the kind of move this should make you wary of.

Second, the field developed a sharper tool that cuts the question cleanly. Rather than asking whether a coarse region is active, Tor Wager and colleagues built a distributed, multivariate “neurologic pain signature” (NPS) — a fine-grained pattern across many voxels that reliably tracks the intensity of first-person physical pain [@wager2013nps]. Then they asked the decisive question: does that signature respond when you watch someone else in pain? The answer was no — the physical-pain signature did not register vicarious pain, even while the coarse regions did [@krishnan2016vps]. Empathy for pain, it turned out, has its own distinct multivariate signature, dissociable from the signature of felt pain. So at the fine grain, watching another’s pain and feeling your own are different patterns that happen to share some coarse real estate.

So where does this leave “I feel your pain”? Honestly, unresolved — but unresolved in a precise and productive way, which is the best kind. The two experiences overlap at the level of coarse regions and shared salience/arousal, and diverge at the level of fine-grained representation. Whether you call that “shared pain machinery” or “distinct processes sharing a salience hub” is partly a question about what grain of analysis you think is the real one — and that is a live argument among careful people, not a gap to be papered over. I will not hand you a verdict. I will note that this is a textbook example of a story that looked settled, got a better measuring instrument, and turned out to be subtler than the first pass suggested — and that learning to expect that, and to hold the open version without distress, is a large part of becoming a scientist. (There is even a reported phenomenon of empathic allodynia, in which a person feels genuine physical pain when witnessing another’s — not mere sympathy but actual pain — which sits at the far, strange end of this question and reminds us how leaky the boundary between sensing your own body and reading another’s can be.)

Note to author — this whole section may be over-built

This is the section I’d most flag against your stated standard, and I want to be direct about why. I framed it as a fascinating live debate with balanced sides — but on reflection that framing may itself overstate the standing of the original claim, which is exactly the failure mode you’re guarding against.

The original “shared pain matrix / empathy runs on your own pain machinery” idea is, structurally, a piece of the early social-neuroscience bubble: single-framework, attractive, heavily retold, and substantially walked back by better methods. So the honest treatment may not be a two-sider but something shorter and more pointed: “An influential early-2000s claim held that empathy for pain uses the brain’s own pain system. Better-controlled work has largely undercut it — those regions encode salience, not pain specifically, and fine-grained pain-pattern analysis finds that watching another’s pain does NOT engage the first-person pain representation. The clean version of ‘I feel your pain’ did not survive.” That’s less airtime, and it teaches your epistemic lesson more honestly than dressing a mostly-failed claim as a thrilling open question.

The salience critique (Iannetti/Mouraux and colleagues) and the multivariate-signature work (Wager group) are reasonably independent programs and do clear the replication bar — so the current picture is defensible. It’s the original claim I gave too much romance to.

Separately and more urgently: the “empathic allodynia” parenthetical should probably go. It’s a genuinely fringe, essentially single-report phenomenon — the very definition of the surprising one-off you no longer trust. I included it because it’s a great story, which is precisely the reason to be suspicious of myself here. My recommendation is to cut it; if you keep it, mark it explicitly as an isolated case report.

9.8 Looking ahead: from sensing to caring

Pain has carried us across a threshold. Every other sense in this unit builds a model of the world; pain builds an evaluation of what is happening to the body and a push to do something about it. We watched its two streams divide — a fast, precise “where” routed onto the body map, and a slow, aching “how bad” routed into the emotional brain — and we watched the affective stream behave like what it is: not a detector but a valuation, dissociable from sensation (the cingulate experiments, the cingulotomy patients), adjustable from above (descending opioid control), and capable of coming apart from injury entirely in both directions (allodynia, congenital insensitivity). Along the way pain kept pointing past itself: toward emotion, toward the interoceptive self of Unit II, toward the motivated, value-guided behavior that the later units are about.

That is the right note to leave the body senses on. We have spent two chapters at the reactive, here-and-now end of sensing — the body in contact with the world, and the body’s own alarm. Now we turn outward and forward, to the senses that reach across distance and so across time: vision and hearing, the senses that let the brain stop reacting and start predicting. We leave the skin behind, but we take with us pain’s central lesson — that sensation and value are not separate stages but two aspects of one system — because it is a lesson the rest of the brain will keep teaching.

What we’re sure of, and what we’re not

Reasonably settled:

Nociception (the detection/signalling of damaging stimuli) and pain (the unpleasant experience) are distinct and can come apart in both directions.
Nociceptors are free nerve endings; tissue damage activates them via a chemical broth including TRPV1-opening signals.
Pain comes in two streams: fast, sharp, well-localized first pain (Aδ) and slow, dull, aching second pain (C fibers).
The two diverge centrally: first pain via VPL → S1 (sensory-discriminative, “where”); second pain via medial thalamus → insula and anterior cingulate (affective-motivational, “how bad”). The two dimensions are dissociable (Rainville hypnosis study; cingulotomy phenomenology).
Descending modulation lets the brain gate pain at the dorsal horn, partly via endogenous opioids (enkephalin); gate-control theory was the conceptual origin.
Congenital insensitivity to pain (SCN9A/Na_v1.7 loss of function) shortens life through unprotected injury — and travels with anosmia because the same channel serves olfactory neurons.

Genuinely unsettled, and presented as such:

The mechanism of referred pain. The standard convergence-projection account is widely taught but, in your instructor’s view, weaker than advertised; convergence is ubiquitous yet misattribution is not. No fully satisfying mechanism.
The mechanism of allodynia. Real and sometimes severe, but the wiring changes that make innocuous touch painful are not well understood.
Whether “empathy for pain” shares pain machinery. Coarse regions (ACC, insula) overlap, but these encode salience broadly, not pain specifically, and the fine-grained physical-pain signature does not respond to others’ pain (a distinct vicarious signature does). Overlap at one grain, divergence at another — genuinely open, and a caution against reverse inference.

And, as always: the core anatomy and phenomenology here are solid. Two pains, two pathways, two dimensions; a gated, opioid-modulated signal; the adaptive necessity of pain shown by those who lack it — you can build on all of it. What remains open is mostly mechanism (referred pain, allodynia) and interpretation at the right grain (the empathy question) — which is to say, exactly the frontier worth your curiosity.