Everything you think you know about pain is probably wrong. Pain is not a signal from damaged tissue — it is an output of the brain, a decision about protection, a prediction. You can have severe tissue damage and no pain (soldiers wounded in battle who feel nothing until safe). You can have agonising pain with no tissue damage at all (phantom limb pain in an amputated limb, chronic pain that persists decades after healing is complete). Pain and damage are not the same thing. Understanding the difference is the difference between a model of pain that makes chronic pain inevitable and one that makes it treatable.
René Descartes drew, in 1664, a diagram of a man with his foot near a fire. A thread ran from the foot, up through the hollow nerve tubes, to the brain — where pulling it opened a pore and "animal spirits" flowed into the nerve, producing the sensation of pain. Pain, in this model, was an alarm signal sent from the site of damage to the brain. More damage = louder alarm. The brain was the passive receiver of signals generated by the body's tissues.
This model — the "cartesian" or "specific" model of pain — dominated medicine for three hundred years. It is still the implicit model most people use when they think about pain: something hurts, so something is damaged there, and the pain signal is being sent from that place. Remove the damaged tissue, fix the structural problem, and the pain will stop. It seemed obvious. It is wrong.
The evidence against the cartesian model accumulated throughout the twentieth century. Surgeons who cut pain pathways to relieve chronic pain found that the relief was often temporary — pain returned through new pathways. X-rays showed severe spinal degeneration in people with no pain, and pristine spines in people with debilitating back pain. Amputees experienced excruciating pain in limbs that no longer existed. Placebos reduced pain — demonstrably, measurably, neurologically — without any physical intervention at the site. The model could not explain these findings. A better model was needed.
In 1965, Ronald Melzack and Patrick Wall published a paper in Science that changed pain science permanently: the Gate Control Theory. Their central proposal: the transmission of pain signals from the periphery to the brain is not automatic. It is modulated — gated — at the level of the spinal cord by a complex system of neural interactions that can increase or decrease the signal before it reaches consciousness. The "gate" can be opened (amplifying pain signals) or closed (reducing them) by multiple factors: other sensory signals, brain signals descending from above, emotional state, attention and context.
This is why rubbing an injury reduces pain (tactile signals from the rubbing close the gate to pain signals from the injury), why distraction reduces pain (descending signals from the prefrontal cortex close the gate), why anxiety increases pain (arousal and threat appraisal open the gate), why social support reduces pain (safety and connection close the gate), and why placebos work (expectation of relief triggers descending inhibitory pathways that close the gate). The gate is not metaphor. It is a real neural mechanism, now understood in considerable anatomical detail, operating in the dorsal horn of the spinal cord.
Why the 1965 paper was rejected by reviewers: the Gate Control Theory paper was initially rejected by the journal Brain because reviewers found it impossible that pain could be modulated from above — by psychological and emotional factors — at the level of the spinal cord. The idea that the nervous system was not a simple alarm cable but a processing system with bidirectional communication was too radical for the model dominant at the time. Melzack and Wall published in Science instead. The paper went on to become one of the most cited scientific publications in history and launched an entirely new understanding of pain, the nervous system and the mind-body relationship.
Melzack extended his work in the 1990s with the Neuromatrix theory — a more radical proposal that pain is not a signal received by the brain from the body but an output produced by the brain to protect the body. The brain continuously integrates sensory information from the body, contextual information (threat level, past experience, expectation, emotional state, social support) and descending modulation — and produces pain when it concludes that protection is warranted. Pain is the brain's best prediction about whether the body needs defending.
This model, now the scientific consensus among pain researchers, explains findings that the cartesian model cannot:
Phantom limb pain — a brain that has a well-established neuromatrix representation of a limb continues to produce pain from that limb even after amputation, because the neural pattern that generates the limb-experience can be activated without sensory input from the limb itself. Nocebo effects — people told that a procedure will be painful experience more pain than those told it will not, even under identical physiological conditions, because the expectation primes the brain's pain-production system. Chronic pain after healing — tissue heals but pain persists, because the brain's protective prediction has become entrenched — a learned pattern that continues to produce pain even when the original threat is resolved. Distraction reducing pain during surgery — documented cases of patients undergoing surgical procedures under hypnosis or deep distraction with minimal anaesthesia, because the brain's attention and threat-appraisal systems are redirected away from the pain-production decision.
Acute pain is the nervous system working correctly — detecting a threat, producing a protective signal, withdrawing the body from harm and promoting healing behaviour. Chronic pain is the nervous system that has learned to protect too much. The gate is stuck open. The brain's threat-prediction system, calibrated by repeated painful experience, produces pain with stimuli that would not produce pain in a sensitised system — light touch becomes agony, normal movement triggers severe pain responses, areas of the body remote from the original injury become painful.
This process — central sensitisation — involves actual neurological changes: the amplification of pain-signalling circuits in the spinal cord and brain, the development of new synaptic connections that lower the threshold for pain production, and changes in the descending inhibitory pathways that would normally close the gate. The nervous system has, in effect, turned up the gain on its pain-production system, and the recalibration of this gain — not the treatment of tissue damage — is the primary challenge of chronic pain management.
Why chronic pain is not "in the mind" — and why it also is: chronic pain is real. The brain changes in central sensitisation are neurological, measurable and not voluntarily controllable. People with chronic pain are not imagining their pain or exaggerating for psychological reasons — their brains are genuinely producing pain signals, and those signals are genuinely experienced as severe. At the same time, the brain's threat-prediction system is the mechanism that maintains chronic pain, and addressing that system through pain education, graded exposure, psychological approaches and nervous system regulation is now the most evidence-based approach to chronic pain treatment. Both statements are true simultaneously: chronic pain is a real neurological condition AND it is maintained by the brain's protective mechanisms AND it can be treated by changing those mechanisms. This is not mind over matter — it is understanding matter accurately enough to change it.
The language we use for emotional pain — heartache, gut-wrenching, a broken heart, feeling crushed — is not metaphor. It reflects a neurological reality: the brain processes social and emotional pain through overlapping systems with physical pain. Matthew Lieberman and Naomi Eisenberger's research at UCLA demonstrated that social exclusion activates the dorsal anterior cingulate cortex — the same brain region activated by physical pain — and that paracetamol (acetaminophen) reduces both physical pain and the distress of social rejection in randomised controlled trials.
This overlap evolved for good reasons: social connection is as essential to primate survival as physical safety, and the pain system's job is to motivate protective behaviour. Being excluded from the group was, for most of human evolutionary history, as dangerous as a broken limb — both warranted a strong motivational signal to change the situation. The social pain system shares circuitry with the physical pain system because the function is the same: protect the organism from conditions that threaten survival.
This has direct therapeutic implications. Chronic pain that does not resolve with physical treatment alone frequently has a significant social and emotional component — not because the pain is "psychosomatic" in the dismissive sense, but because the brain's threat-detection system registers social and emotional threat alongside physical threat, and chronic psychosocial stress maintains central sensitisation. Treatment that addresses only the physical, leaving the social and emotional dimension unaddressed, is treating half the system.
The neuroscience of pain is established and clinically transformative. Gate Control Theory and the Neuromatrix model are mainstream pain science, taught at every contemporary pain research institution. The distinction between nociception and pain, the role of central sensitisation in chronic pain and the evidence base for multidisciplinary pain treatment (combining physical, psychological and social approaches) are not fringe positions — they are the current scientific consensus, and departing from them in clinical practice produces worse outcomes for patients.
Understanding pain physiology reduces pain. One of the most consistent findings in pain education research is that patients who understand the neuroscience of pain — who understand that pain is a brain output, not a tissue signal, and that central sensitisation maintains chronic pain — show significant reductions in fear-avoidance behaviour, improved function and often reduced pain itself. Lorimer Moseley and David Butler's pain education programs (Explain Pain) produce clinically meaningful outcomes through education alone, without any other physical intervention. The information about pain is itself therapeutic.
None of this means pain is not real. The brain-output model of pain is sometimes misunderstood as implying that chronic pain is "imaginary" or that people should simply "think their way out of it." This is a serious misreading. Chronic pain is a genuine neurological condition. The brain changes in central sensitisation are real and measurable. What the modern model offers is not the dismissal of chronic pain but a more accurate understanding of where it lives — in the nervous system's threat-prediction circuitry — and therefore more precise targets for treatment.