Phantom Limb

Abstract

A phantom limb, which MeSH classifies under postoperative pain, is the vivid sensation that an amputated or missing limb is still present, frequently accompanied by pain felt in the limb that no longer exists. Silas Weir Mitchell named the phenomenon in the nineteenth century from his observations of Civil War amputees, and it became a central test case for how the brain builds and maintains a model of the body. Three accounts compete to explain it: a peripheral account rooted in the severed nerves, a cortical-reorganization account in which neighbouring body maps invade the deafferented territory, and a preserved-representation account in which the missing limb's cortical map persists largely intact. This article defines the phantom limb and surveys its phenomenology, prevalence, and mechanisms, examines its significance for the body schema, and reviews mirror therapy and measurement, with interactive demonstrations.

Keywords: phantom limb, cortical reorganization, body schema, mirror therapy

What a Phantom Limb Is

A phantom limb is the persisting sensation that a limb remains attached and functioning after it has been amputated or was never present at birth. The experience is not a memory or an idea of the limb; it is a felt presence, located in external space, with a position, a posture, and often the capacity to move at the person's will. Most people who lose a limb report a phantom of some kind, and for many it carries pain, the condition termed phantom limb pain. The phenomenon is striking because the sensation is referred to a region of space where there is now nothing at all, which makes the phantom limb one of the clearest demonstrations that the body as we feel it is a construction of the nervous system rather than a direct readout of the flesh.

Silas Weir Mitchell, an American neurologist working with amputees during and after the American Civil War, gave the phenomenon its name and its first systematic clinical description, drawing attention to how real and how troubling the sensations could be (Mitchell, 1872). The phantom limb should be distinguished from stump pain, which is felt in the residual limb itself and often has a straightforward peripheral cause, and from the non-painful phantom sensation, the mere awareness of the absent limb without discomfort. Phantom limb pain is the painful subtype, and it is the one that has driven most of the theoretical debate, because its persistence and its resistance to peripheral treatment pointed early investigators toward the central nervous system.

Four accounts recur through the literature, and because they make different and sometimes opposite predictions it helps to hold them side by side before the detail. Table 1 summarises what each claims, the evidence most often cited for it, and where it runs into difficulty; the sections that follow take them in turn.

Phenomenology and Telescoping

The phantom limb is not a single experience but a family of them. The phantom may feel complete, as if the whole arm or leg were still there, or it may be incomplete, with the hand or foot vivid but the intervening segment faint or absent. Many amputees can voluntarily move the phantom, flexing phantom fingers or clenching a phantom fist, while others feel it locked in a fixed and sometimes painfully contorted posture. The sensations include not only pain but pressure, temperature, itch, and the impression of a ring or a watch still worn, and they can be triggered or modulated by touching the stump, by emotional state, and by attention to the limb.

A characteristic change over time is telescoping, in which the phantom gradually shortens so that the phantom hand or foot is felt to approach, and eventually to occupy, the end of the stump, the intervening limb having seemingly retracted. Telescoping is one of the regularities that any adequate theory must explain, because it shows that the phantom is plastic rather than fixed, reshaping itself in the months and years after amputation. The malleability of the phantom is also demonstrable on a shorter timescale: experimental manipulation of sensory input can change the vividness and even the felt existence of the phantom, showing that it is continuously maintained by the balance of signals reaching the brain rather than passively persisting (Hunter et al., 2003).

Prevalence and Time Course

Phantom sensations are the rule rather than the exception after amputation, and painful phantoms are common. A prospective study following amputees through the first six months after surgery mapped how phantom limb, phantom pain, and stump pain emerge and evolve, establishing that phantom pain often appears within the first days and that its prevalence is high in the early postoperative period (Jensen et al., 1983). That early work set the pattern for the clinical understanding of the condition, including the observation that pain present before amputation can be echoed in the phantom.

The best current estimate of how common phantom limb pain is comes from a systematic review and meta-analysis pooling prevalence across many amputee samples, which found that a substantial majority of people with amputations experience phantom limb pain, confirming that it is a major and persistent clinical problem rather than a rare curiosity (Limakatso et al., 2020). Clinical reviews synthesising the mechanisms, risk factors, and management of the condition have repeatedly emphasised this high prevalence together with the difficulty of treatment, which is part of what has kept the phenomenon at the centre of both clinical and theoretical attention (Nikolajsen & Jensen, 2001).

The Peripheral Account

The earliest and most intuitive explanation locates the phantom in the periphery. When a nerve is severed, the cut axons do not fall silent; they can form neuromas, tangled sprouts at the stump that generate spontaneous and ectopic discharge, and this abnormal input, arriving over pathways that the brain still interprets as coming from the lost limb, could be read as sensation in the phantom. On this view the phantom is driven from below, by the irritable remnants of the damaged nerves, and the attraction of the account is that it explains stump pain well and predicts that local treatment of the neuroma should help.

The peripheral account cannot be the whole story, and the reasons are instructive. Phantoms occur in people born without a limb, who have no severed nerve to generate the signal, and they persist after complete spinal transection above the level of the amputation, when no peripheral input can reach the brain at all. Local anaesthesia of the stump often fails to abolish phantom pain, and the phantom's plasticity, telescoping, and susceptibility to purely central manipulations such as mirror therapy all point beyond the periphery. Peripheral input clearly contributes and can trigger or worsen a phantom, but the phenomenon requires an account of what the central nervous system is doing, which is where the modern debate is joined (Collins et al., 2018).

The Cortical-Reorganization Account

The most influential central account for two decades was cortical reorganization. The primary somatosensory cortex contains an orderly map of the body surface, the sensory homunculus, in which adjacent body parts occupy adjacent cortical territory. When a limb is amputated, its territory in the map is deafferented, and neighbouring representations, the face and the upper arm for a missing hand, were found to expand into the vacated zone. The decisive human evidence came from magnetoencephalographic mapping showing that the amount of this invasion correlated with the magnitude of phantom limb pain: the more the mouth and chin representation had shifted into the former hand area, the worse the pain (Flor et al., 1995).

A vivid behavioural counterpart was the discovery of referred sensations. Touching the face of an amputee could evoke precisely localised sensation in the phantom hand, with a topographic map of the fingers found on the cheek, because the facial input was now driving the reorganised hand cortex (Ramachandran et al., 1995). These findings were drawn together in a widely cited synthesis that presented the reorganised cortex as the substrate of phantom phenomena and framed phantom pain as its perceptual correlate (Ramachandran & Hirstein, 1998). The account was consolidated into the maladaptive-plasticity model, which held that phantom limb pain is a consequence of this abnormal cortical remapping, a case of central plasticity gone wrong (Flor et al., 2006).

The Preserved-Representation Account

The maladaptive-plasticity consensus was challenged by a programme of work showing nearly the opposite pattern. Using functional and structural neuroimaging in amputees with and without phantom pain, investigators found that phantom pain was associated with preserved, not degraded, structure and function in the former hand area of cortex: the people with the worst pain were those whose missing-hand representation remained most intact (Makin et al., 2013). This directly contradicts the prediction of the reorganization account, in which pain should track the loss of the original representation and its invasion by neighbours.

The preserved-representation finding was reinforced by evidence that the cortical map of the missing hand can persist in remarkable detail decades after amputation. By having amputees move their phantom fingers in the scanner, researchers recovered the fine, finger-by-finger topography of the absent hand, a neural fingerprint that should have been overwritten if wholesale reorganization were the rule (Kikkert et al., 2016). The two bodies of evidence, reorganization on one side and preservation on the other, set up the central tension of the contemporary field, which has prompted a direct attempt at reconciliation (see Current Directions).

The Neuromatrix and the Body Schema

Ronald Melzack proposed a framework broad enough to encompass the phantom's independence from peripheral input: the neuromatrix. On this account the brain contains a genetically specified, widely distributed network whose ongoing activity generates a unified sense of the bodily self, a continuously produced signature of the body that Melzack called the neurosignature. Perception of the body, including a limb, is the output of this network, which is normally modulated by sensory input but does not depend on it; the phantom is what the neuromatrix continues to produce for a limb that is no longer sending signals (Melzack, 1990).

The neuromatrix idea connects the phantom limb to the broader cognitive construct of the body schema, the brain's internal model of the body's configuration and extent that underlies action and the sense of ownership. A limb that is congenitally absent can still have a place in this model, which is why such individuals can experience phantoms, and the model is what telescoping reshapes and what mirror therapy addresses. The phantom limb, in this light, is not an error about the periphery so much as a window onto the standing representation of the body that the brain maintains in order to act in and perceive the world.

Cognitive Implications

Although phantom limb pain is a clinical and perceptual phenomenon rather than a traditional topic of higher cognition, it bears directly on several core questions in cognitive science. The first is the nature of the body schema and the sense of body ownership. The phantom shows that the felt body and the physical body can come apart, and the ease with which the phantom can be manipulated, by mirrors, by touching the face, by changing sensory input, demonstrates that body ownership is an inference the brain continuously updates rather than a fixed given. The same multisensory integration that produces the rubber-hand illusion in intact participants, in which synchronous seen and felt touch make a model hand feel like one's own, is implicated in how a phantom is sustained and how it can be altered (Botvinick & Cohen, 1998).

The second is the role of top-down expectation and attention in perception. A phantom can be strengthened by attending to it and weakened by competing input, and mirror therapy works by supplying a visual prediction, an image of an intact, moving limb, that the perceptual system reconciles with the felt phantom. This places the phantom limb within the same predictive framework used to explain other constructive perceptual phenomena, in which perception is the brain's best reconciliation of prediction and evidence. The phantom limb thus serves cognitive science as a natural experiment in how the brain builds a coherent percept of the self, and why that construction, once built, can be so difficult to revise.

Mirror Therapy and Treatment

The most celebrated treatment to emerge from the central accounts is mirror therapy. A mirror is placed vertically in the sagittal plane so that the reflection of the intact limb appears in the location of the missing one; when the person moves the intact limb and watches its reflection, they see and feel a restored, moving phantom. The technique grew out of the mirror-box experiments, which showed that this visual feedback could capture the position of the phantom and, by letting patients unclench a painfully clenched phantom hand, relieve pain and even produce the temporary disappearance of the phantom (Ramachandran & Rogers-Ramachandran, 1996).

Mirror therapy was subsequently tested in controlled form. A randomised trial in people with lower-limb amputations found that those who used a mirror had significantly greater reduction in phantom pain than those using a covered mirror or mental visualisation, providing experimental support for the technique (Chan et al., 2007). Mirror therapy sits alongside a range of other approaches, pharmacological, surgical, and behavioural, whose rationale and evidence have been surveyed in reviews of the competing theories and their associated therapies; no single treatment is reliably effective for everyone, which is itself a reflection of the unsettled mechanistic picture (Weeks et al., 2010).

Assessment and Measurement

Because the phantom is a private experience, it is measured through report and through the neural correlates that report can be tied to. Clinically, phantom limb pain is assessed with the same self-report pain instruments used for other chronic pain, rating intensity, frequency, and quality, supplemented by questions that distinguish phantom pain from stump pain and from non-painful phantom sensation, a distinction that matters because the three have different mechanisms and treatments (Nikolajsen & Jensen, 2001). Characterising the phantom also means charting its phenomenology: its completeness, its mobility, its posture, and whether it has telescoped.

The scientific study of mechanism depends on linking these reports to the brain. The reorganization and preserved-representation accounts were adjudicated precisely by measuring the cortical map, with magnetoencephalography quantifying the shift of the lip representation into the hand area (Flor et al., 1995) and functional MRI quantifying the integrity of the missing-hand representation during phantom movement (Makin et al., 2013; Kikkert et al., 2016). The methodological lesson of the field is that the same phenomenon can appear to support opposite theories depending on exactly what is measured and how, which is why the reconciliation of the two accounts has turned on careful, convergent measurement rather than on any single decisive experiment.

Figure

Figure 1

Two Accounts of the Deafferented Hand Area

Interactive Demonstrations

The three demonstrations below make the core ideas manipulable. The first models referred sensation, letting the reader touch points on a schematic face and watch the sensation appear on the phantom hand, as predicted by cortical remapping. The second is a mirror-box simulation, letting the reader move the intact hand and see the mirror restore a congruent phantom, with phantom pain easing over simulated sessions. The third plots the two rival accounts directly, letting the reader vary the state of the former hand area and see how the reorganization and preserved-representation models make opposite predictions about phantom pain.

Worked Example

Consider the two competing accounts cast as predictions to be tested against data, the way the field actually adjudicated them. Imagine a study of ten amputees in which each person's phantom pain is rated from 0 to 10, and two cortical measures are taken: the distance, in millimetres, that the lip representation has shifted into the former hand area (the reorganization measure), and the integrity of the missing-hand representation, scored from 0 to 1 (the preservation measure).

The reorganization account predicts a positive correlation between pain and the lip shift: the more the face has invaded, the worse the pain. The preserved-representation account predicts a positive correlation between pain and representational integrity: the more intact the hand map, the worse the pain. These cannot both be the dominant effect in the same dataset, because the two cortical measures tend to move in opposite directions, more invasion generally means less preservation. Suppose the data show that pain correlates at about +0.7 with representational integrity and near zero with the lip shift. That pattern is the signature reported by the preserved-representation studies, and it is why a finding that looked, under the reorganization lens, as though cortex had simply been overwritten turned out, when the missing-hand map was measured directly, to show the opposite: the people in the most pain were those whose hand representation had survived. The worked lesson is methodological as much as theoretical, that which account a dataset appears to support depends on which quantity is placed on the axis, so progress required measuring both in the same people.

Discussion

The phantom limb earns its place in cognitive science because it exposes, with unusual clarity, that the body we feel is something the brain builds. A percept as compelling as a whole arm can be sustained for a limb that is gone, which means the machinery responsible must be the same machinery that, in the intact person, constructs the ordinary and seemingly self-evident sense of having a body. Every serious account, peripheral, reorganization, preserved-representation, and neuromatrix, is in the end an account of that constructive machinery, differing over where the critical activity sits and what maintains it.

The unresolved tension between reorganization and preserved representation is not a failure of the field but a sharpening of its central question. The maladaptive-plasticity model gave a clean story, pain as the cost of a remapped cortex, and motivated therapies aimed at reversing the remapping. The preserved-representation findings forced a reconsideration, suggesting that the persistence of the missing limb's representation, not its erasure, is what hurts. Both cannot be straightforwardly true in their strong forms, and the attempt to reconcile them has become the organising problem of contemporary phantom-limb research, with direct consequences for which treatments should be expected to work and why mirror therapy, which restores rather than removes a representation, helps the patients it does.

Current Directions

The leading edge of the field is the explicit reconciliation of the reorganization and preserved-representation accounts. In a joint review, the chief proponents of the two positions set out what each body of evidence does and does not show and proposed how brain reorganization following amputation should be understood in light of both, moving the debate from a contest of rival models toward an integrated account of what happens to cortex after limb loss and how that relates to pain (Makin & Flor, 2020). That the two camps co-authored the synthesis is itself a marker of where the field now stands.

A second thread is the drive toward mechanism-based treatment. Contemporary reviews integrate peripheral, spinal, and central contributions rather than treating them as competitors, and they survey emerging interventions, from targeted neuroma treatment and prosthetic embodiment to brain-computer interfaces, that aim to act on the specific level at which a given patient's pain is generated (Collins et al., 2018). The open questions are no longer whether the phantom is peripheral or central but how the levels interact, how the preserved representation and the residual peripheral input jointly produce pain, and whether a measurement of the former hand area could one day predict which treatment a particular amputee will respond to.

Common Misconceptions

A phantom limb is imagined or psychological.

The phantom is a genuine perceptual phenomenon generated by the nervous system, with measurable neural correlates in somatosensory cortex; it occurs even in people born without the limb (Melzack, 1990; Flor et al., 1995).

Phantom limb pain comes from the cut nerves in the stump.

Peripheral input contributes, but phantoms occur after complete spinal injury and in congenital absence, and local anaesthesia often fails to abolish the pain, so a central account is required (Collins et al., 2018).

Phantom limb pain is rare.

A meta-analysis of prevalence finds that a substantial majority of amputees experience phantom limb pain, making it a common and persistent clinical problem (Limakatso et al., 2020).

Phantom pain is simply the result of the brain forgetting the lost limb.

Evidence suggests the reverse for many patients: pain is associated with a preserved, detailed representation of the missing hand rather than its erasure (Makin et al., 2013; Kikkert et al., 2016).

Glossary

Key Researchers

Herta Flor

(Central Institute of Mental Health, Heidelberg University, Mannheim). Neuropsychologist whose 1995 study tied phantom limb pain to the extent of somatosensory cortical reorganization, and who later refined and partly revised the maladaptive-plasticity account she had helped establish. Google Scholar · Wikidata · Wikipedia

Tamar R. Makin

(MRC Cognition and Brain Sciences Unit, University of Cambridge). Cognitive neuroscientist whose work showed that phantom pain tracks preserved rather than reorganized hand representation, reframing the plasticity debate she then co-reviewed with Flor. Faculty · Google Scholar · Wikipedia

Ronald Melzack

(McGill University, Montreal). Psychologist who, with Patrick Wall, proposed the gate control theory of pain, and whose neuromatrix theory recast the phantom limb as the output of a genetically specified network generating the bodily self. Wikipedia · Wikidata

Silas Weir Mitchell

(Turner's Lane Hospital, Philadelphia). American neurologist who coined the term phantom limb and gave the first systematic clinical description, from his observations of Civil War amputees. Wikipedia · Wikidata

Lone Nikolajsen

(Aarhus University Hospital). Anaesthesiologist and pain researcher whose clinical reviews defined the mechanisms, risk factors, and management of phantom limb pain. ORCID · Faculty

Vilayanur S. Ramachandran

(Center for Brain and Cognition, University of California, San Diego). Neuroscientist who introduced mirror therapy and the cortical-remapping account of referred sensations in phantom limbs, synthesised in his widely cited 1998 review. Faculty · Google Scholar · Wikipedia · Wikidata

Frequently Asked Questions

What is a phantom limb?

A phantom limb is the vivid sensation that an amputated or congenitally absent limb is still present. The limb is felt in external space with a position and a posture, and it can often be moved at will. Most people who lose a limb experience a phantom of some kind.

What is phantom limb pain?

Phantom limb pain is pain experienced as located in the missing limb. It is distinct from stump pain, which is felt in the residual limb, and from non-painful phantom sensation. A meta-analysis finds that a substantial majority of amputees experience phantom limb pain, so it is common rather than rare.

Why does a limb that is gone still hurt?

The pain is generated centrally, in the nervous system, not simply by the cut nerves. Phantoms occur even in people born without the limb and after complete spinal injury, which shows that the brain maintains a representation of the limb that can produce sensation and pain without any input from the periphery.

Who first described the phantom limb?

Silas Weir Mitchell, an American neurologist, named the phenomenon and gave its first systematic clinical description in the nineteenth century, based on his work with amputees from the American Civil War.

Does phantom pain come from the cortex being remapped?

One influential account says yes, that neighbouring body representations invade the deafferented cortex and that pain tracks this reorganization. A competing body of evidence finds the opposite, that pain is associated with a preserved, intact representation of the missing limb. Reconciling these accounts is the central problem of current research.

What is the body schema, and how does the phantom relate to it?

The body schema is the brain's internal model of the body's layout and extent. The phantom limb shows that this model can retain a limb that is no longer there, which is why the felt body and the physical body can come apart and why the phantom can be reshaped by changing sensory input.

Does mirror therapy work?

Mirror therapy uses a mirror reflection of the intact limb to create the image of a restored, moving phantom. A randomised controlled trial found greater pain reduction with a real mirror than with control conditions. It helps many patients, though no treatment works for everyone, which reflects the unsettled understanding of the mechanism.

How is the phantom limb relevant to cognitive science?

It is a natural experiment in how the brain constructs the sense of having a body. It bears on body ownership, multisensory integration, and the role of expectation in perception, demonstrating that a compelling percept of a limb can be built and maintained entirely by the brain.

References

Botvinick, M., & Cohen, J. (1998). Rubber hands ‘feel’ touch that eyes see. Nature, 391(6669), 756. https://doi.org/10.1038/35784

Chan, B. L., Witt, R., Charrow, A. P., Magee, A., Howard, R., Pasquina, P. F., Heilman, K. M., & Tsao, J. W. (2007). Mirror therapy for phantom limb pain. New England Journal of Medicine, 357(21), 2206–2207. https://doi.org/10.1056/NEJMc071927

Collins, K. L., Russell, H. G., Schumacher, P. J., Robinson-Freeman, K. E., O'Conor, E. C., Gibney, K. D., Yambem, O., Dykes, R. W., Waters, R. S., & Tsao, J. W. (2018). A review of current theories and treatments for phantom limb pain. Journal of Clinical Investigation, 128(6), 2168–2176. https://doi.org/10.1172/JCI94003

Flor, H., Elbert, T., Knecht, S., Wienbruch, C., Pantev, C., Birbaumer, N., Larbig, W., & Taub, E. (1995). Phantom-limb pain as a perceptual correlate of cortical reorganization following arm amputation. Nature, 375(6531), 482–484. https://doi.org/10.1038/375482a0

Flor, H., Nikolajsen, L., & Jensen, T. S. (2006). Phantom limb pain: A case of maladaptive CNS plasticity? Nature Reviews Neuroscience, 7(11), 873–881. https://doi.org/10.1038/nrn1991

Hunter, J. P., Katz, J., & Davis, K. D. (2003). The effect of tactile and visual sensory inputs on phantom limb awareness. Brain, 126(3), 579–589. https://doi.org/10.1093/brain/awg054

Jensen, T. S., Krebs, B., Nielsen, J., & Rasmussen, P. (1983). Phantom limb, phantom pain and stump pain in amputees during the first 6 months following limb amputation. Pain, 17(3), 243–256. https://doi.org/10.1016/0304-3959(83)90097-0

Kikkert, S., Kolasinski, J., Jbabdi, S., Tracey, I., Beckmann, C. F., Johansen-Berg, H., & Makin, T. R. (2016). Revealing the neural fingerprints of a missing hand. eLife, 5, e15292. https://doi.org/10.7554/eLife.15292

Limakatso, K., Bedwell, G. J., Madden, V. J., & Parker, R. (2020). The prevalence and risk factors for phantom limb pain in people with amputations: A systematic review and meta-analysis. PLOS ONE, 15(10), e0240431. https://doi.org/10.1371/journal.pone.0240431

Makin, T. R., Scholz, J., Filippini, N., Henderson Slater, D., Tracey, I., & Johansen-Berg, H. (2013). Phantom pain is associated with preserved structure and function in the former hand area. Nature Communications, 4, 1570. https://doi.org/10.1038/ncomms2571

Makin, T. R., & Flor, H. (2020). Brain (re)organisation following amputation: Implications for phantom limb pain. NeuroImage, 218, 116943. https://doi.org/10.1016/j.neuroimage.2020.116943

Melzack, R. (1990). Phantom limbs and the concept of a neuromatrix. Trends in Neurosciences, 13(3), 88–92. https://doi.org/10.1016/0166-2236(90)90179-E

Mitchell, S. W. (1872). Injuries of nerves and their consequences. J. B. Lippincott.

Nikolajsen, L., & Jensen, T. S. (2001). Phantom limb pain. British Journal of Anaesthesia, 87(1), 107–116. https://doi.org/10.1093/bja/87.1.107

Ramachandran, V. S., Rogers-Ramachandran, D., & Cobb, S. (1995). Touching the phantom limb. Nature, 377(6549), 489–490. https://doi.org/10.1038/377489a0

Ramachandran, V. S., & Rogers-Ramachandran, D. (1996). Synaesthesia in phantom limbs induced with mirrors. Proceedings of the Royal Society of London B, 263(1369), 377–386. https://doi.org/10.1098/rspb.1996.0058

Ramachandran, V. S., & Hirstein, W. (1998). The perception of phantom limbs: The D. O. Hebb lecture. Brain, 121(9), 1603–1630. https://doi.org/10.1093/brain/121.9.1603

Weeks, S. R., Anderson-Barnes, V. C., & Tsao, J. W. (2010). Phantom limb pain: Theories and therapies. The Neurologist, 16(5), 277–286. https://doi.org/10.1097/NRL.0b013e3181edf128