Munchrd?
Ever Wondered? · The Mind

Why do amputees feel a limb that's gone?

A missing hand can itch, move, feel warm — and clench into a fist so tight the vanished nails seem to dig in. So how do you feel agony in a hand that isn't there? And how did a two-pound mirror make that pain fade?

fact-checked
Munchrd illustration for: Why do amputees feel a limb that's gone?
✓ The short answer

Because your limb was never really in your limb. Your body is a map drawn inside your brain, and when a limb is amputated the map keeps its section alive — so you still feel the missing hand vividly, and sometimes painfully. The leading idea is that a mirror can ease the pain by giving that map fresh visual proof the phantom obeyed.

The 20-second version

  • Most amputees keep a vivid phantom limb — it can itch, feel hot or cold, seem to move, and for many it hurts.
  • Cut nerves in the stump play a part, but they can't explain a whole moving hand with individual fingers. The answer is largely in the brain's body-map.
  • That map — the somatosensory homunculus — lays out the body across the cortex. The hand's patch sits right next to the face's, which is why stroking some amputees' cheeks is felt on the missing fingers.
  • The leading idea for the pain is a "learned paralysis": the brain keeps ordering the phantom to move, gets no feedback that it did, and reads it as frozen in a painful clench.
  • V.S. Ramachandran's mirror box reflects the intact hand into the phantom's position; watching it "move" can ease the pain for some — though the evidence is genuinely mixed and the deeper cause is still debated.

When someone loses an arm or a leg, most of them don't feel it vanish. They feel it stay. The missing limb is still vividly there — it can itch, it can feel warm or cold, it seems to move. And for a great many people, it hurts. A phantom hand can clench into a fist so tightly that the fingernails feel like they're digging into the palm, and no matter how hard the person tries, they cannot get it to open. So here is the puzzle: how do you feel genuine agony in a hand that simply isn't there?

01 · The obvious guessIt's not just the stump

The first obvious guess is the stump. The nerves that were cut are still there, frayed and firing off signals, sometimes tangling into knots called neuromas. And that is a real part of the story. But it cannot explain a whole, detailed, moving hand — one with individual fingers you can feel curl and clench. Cutting a cable doesn’t tell you why the screen still shows a picture. For that, you have to go much further up. Right into the brain.

02 · The mapThere's a little you painted on your brain

Because your brain holds a complete map of your body, spread across its surface. A scaled-down you, painted onto the strip of cortex just behind the crown of your head — the somatosensory homunculus. We know it exists because in the mid-20th century the neurosurgeon Wilder Penfield, operating on conscious patients, touched tiny points on the cortex with an electrode and asked what they felt — and built the map point by point. Your hand has its own patch of territory. Your face has another. Your feet, your lips, all of it, laid out in order like districts on a map — and sized not by how big the body part is, but by how sensitive it is, which is why the hands and lips are enormous.

03 · The vacancyThe district goes dark — and the neighbours move in

Now, amputate the hand, and its district on that map suddenly goes quiet. No more signals coming in. But brain territory doesn’t just sit empty. To some degree the neighbouring regions, starved for space, can start to creep in and take it over — a striking example of the brain’s plasticity. And here is where it gets genuinely uncanny. On that map, the patch for your hand sits right next door to the patch for your face. So in many amputees, if you gently stroke their cheek, they feel it — clear as day — on the fingers of the hand that is no longer there. The face has quietly reached across into the empty hand. Ramachandran first described this in the 1990s, and brain imaging has since confirmed it.

04 · The painA cruel little feedback loop

So what about the pain? Here the science gets less settled, so it’s worth being honest about it. The best-known explanation, from the neuroscientist V.S. Ramachandran, is a cruel little feedback loop he called learned paralysis. The brain keeps firing off commands to move the phantom hand — but no signal ever comes back to say it moved. So the brain concludes the hand is stuck, jammed, frozen in that clench. And a limb frozen tight, forever, hurts. It’s an elegant idea, and it directly inspired the treatment we’ll come to. But it’s the leading idea, not the last word.

Here's where it gets good

One of the cheapest ideas in the history of medicine — a plain mirror — grew straight out of that theory. Stand it right, and the frozen fist can let go.

05 · The mirrorTwo pounds of glass against years of pain

That “learned paralysis” idea led Ramachandran, in the mid-1990s, to one of the most elegant experiments in neuroscience. A mirror. He stood a plain mirror upright beside the patient, angled so the reflection of their intact hand appeared exactly where the phantom one felt like it should be. The patient looks into the mirror and sees two whole hands again. Then they move the real one — and the brain, watching the reflection, finally sees the phantom hand do what it’s told. It opens. It unclenches. The frozen fist lets go. In Ramachandran’s early cases, patients who hadn’t been able to “move” their phantom for years suddenly felt it spring to life the moment they opened their eyes on the mirror. And for some people, the phantom pain that had haunted them quietly drained away.

06 · The honest caveatWhy nobody's fully sure it works — or why

Now, to be fair, the mirror is not a miracle cure, and it would be dishonest to sell it as one. It doesn’t work for everyone. And when researchers pool the trials, the evidence comes out genuinely mixed: some reviews find a real benefit, while reviews that include only the most rigorous, placebo-controlled studies find no clear effect. The studies tend to be small and imperfect, so the honest verdict is “promising, cheap, low-risk, and worth trying — but unproven.”

The deeper mechanism is contested too. The tidy story — amputation, the map goes dark, the neighbours invade, and that remapping causes the pain — has been challenged. Tamar Makin and colleagues found the opposite of what you’d expect: the people with the worst phantom pain tended to have a former hand-area that was remarkably well preserved, not erased. Which flips the causal arrow, and suggests the pain may keep the old map alive rather than the remapping causing the pain. Scientists are still arguing it out.

07 · The payoffYour body lives in your head

But step back, and the deeper lesson survives every one of those caveats intact. Your arm was never really in your arm. The entire feeling of your body is built, and held, inside your brain — a map that can outlive the territory it was drawn from. That’s why a limb can go on itching, moving, and aching years after it’s gone. And it’s why, at least for some people, you can soothe a limb you no longer have simply by fooling that map with two pounds of glass. Your body isn’t quite where you think it is. It’s a map, drawn in your head — and the map can outlast the limb.

Prefer to watch?

The animated version

Same answer, told as a little animated story with the strange little stickman acting it out. Dry, quick, and — obviously — correct.

▶ Video dropping soon — subscribe on YouTube
People also ask

Quick questions

What is a phantom limb?

It's the vivid feeling that a limb is still there after it's been amputated. Studies suggest most amputees experience it — as many as around 80%. The phantom can feel like it's moving, itching, warm or cold, or held in a particular position, and for a large share of people it's painful. It usually appears within days of the amputation and can persist for years.

Why does a missing limb hurt (phantom limb pain)?

The cause is genuinely debated. The best-known idea, from V.S. Ramachandran, is a "learned paralysis": the brain keeps sending move commands to the phantom, never gets sensory confirmation it moved, and settles on the limb being stuck in a painful clench. Others link the pain to changes in the brain's body-map, and more recent work suggests the pain tracks a preserved limb representation rather than a remapped one. The peripheral nerves in the stump also contribute.

Why can stroking an amputee's face be felt in the missing hand?

Because of how the body is laid out on the brain's map. On the somatosensory cortex, the region for the hand sits right next to the region for the face. After the hand is amputated its patch goes quiet, and the neighbouring face region can effectively encroach — so touch to the cheek gets felt on the phantom fingers. Ramachandran first described this in the 1990s and it's been replicated with brain imaging.

How does a mirror help phantom limb pain?

In Ramachandran's mirror box, a vertical mirror reflects the intact hand into the position where the phantom feels like it is. The person moves the real hand and, in the reflection, watches the "phantom" move and unclench — giving the brain visual proof the limb obeyed. For some people the pain eases. But it doesn't work for everyone, and reviews of the evidence are mixed rather than conclusive.

Does mirror therapy actually work for phantom limb pain?

It helps some people, but the evidence is unsettled. Some systematic reviews find benefit; others that included only the most rigorous placebo-controlled trials found no clear effect. The studies tend to be small and methodologically weak. It's cheap and low-risk, so it's often worth trying — but it is not a guaranteed fix.

Our sources

// every claim on this page was checked before it went up

Most amputees experience a vivid phantom limb (it can seem to move, itch, and feel warm or cold), and a large share experience phantom limb pain; estimates of phantom sensation reach around 80% of amputees. Medlink Neurology; phantom limb reviews (JCI, 2018)
Cut nerves in the stump (neuromas) contribute to phantom sensations, but peripheral factors alone cannot account for the full, detailed, moving phantom; central (brain) mechanisms are central to the phenomenon. JCI review of phantom limb pain theories (2018); Ramachandran & Hirstein, 1998
The body is mapped across the somatosensory cortex (the 'homunculus,' from Penfield's stimulation studies), scaled by sensitivity, with the hand's region lying adjacent to the face's. Penfield's sensory homunculus; standard somatosensory anatomy
In many upper-limb amputees, touching the face evokes topographically organised sensations in the phantom hand — first described by Ramachandran in the 1990s and later replicated with neuroimaging. Ramachandran et al., 1990s; replicated in fMRI (see Phantom limb, Wikipedia summary + neuroimaging studies)
The leading explanation for phantom limb pain is Ramachandran's 'learned paralysis': motor commands to the phantom get no sensory feedback, so the brain reads the limb as frozen/clenched. This is the leading idea, not a settled mechanism. Ramachandran & Rogers-Ramachandran, 1996; JCI review (2018)
The classic 'maladaptive cortical remapping' account of phantom pain is contested: Makin and colleagues report that phantom pain is associated with preserved, not degraded, structure and function in the former hand area. Makin et al., "Phantom pain is associated with preserved structure and function in the former hand area," Nature Communications, 2013
Ramachandran's mirror box (developed 1996) reflects the intact hand into the phantom's position, providing visual feedback the phantom 'moved'; it eases pain for some people. Ramachandran & Rogers-Ramachandran, 1996; 'It's All Done With Mirrors' (PMC, 2016)
Evidence for mirror therapy in phantom limb pain is mixed and low-quality: some reviews find benefit, while reviews restricted to the most rigorous placebo-controlled trials find no clear effect. Systematic reviews & meta-analyses of mirror therapy for PLP (Arch Phys Med Rehabil 2021; placebo-controlled review, PMC 2023)