Some pain is a message from the body. Some is the brain pressing on the wound. Telling them apart is one of the most useful things a person can learn — and a live, sometimes bitter fight in pain science is now circling the exact question of how much of long-term suffering the brain is producing rather than reporting. None of it says your pain isn't real. The opposite: it says the brain is doing more of the work than we were taught, which is why the work can be changed.
A line held throughout: none of this means your pain is imaginary or your fault. It means the brain is more involved than the old "damage = pain" picture allowed — which is exactly what makes it changeable.
An old teaching, worth stealing because it turns out to be mechanically accurate: when something bad happens, you're struck by two arrows. The first is the event — the injury, the loss. The second is what the mind does with it: the replaying, the "this shouldn't be happening," the fear of what it means, the certainty it will never end. The first is often unavoidable. The second, it turns out, is not — and for many people most of the ongoing suffering lives there.
Modern pain science says something that sounds abstract but changes everything in practice: pain is not a simple readout of damage. It isn't a signal travelling from a hurt body part to a "pain centre" like a doorbell. Pain is something the brain produces — its best judgement of how much danger you're in — using the body's signal plus your expectations, mood, history, and what the pain seems to mean. That's why the same injury hurts more when you're frightened or alone, and less when you're safe or absorbed. The body sends an input; the brain writes the output. And for chronic pain — pain that outlasts the healing — the brain can keep writing "danger" after the damage is gone. The alarm keeps ringing in an empty building.
In 2022, a team led by Yoni Ashar published a randomised trial in JAMA Psychiatry. People with chronic back pain — real, for years — were given Pain Reprocessing Therapy, whose whole aim is to help the brain re-learn that the signals aren't dangerous. About two-thirds became pain-free or nearly pain-free — 66% (33 of 50), versus 20% on a placebo injection and 10% on usual care. At the five-year follow-up the treatment group held up well (55% still pain-free or nearly), but the honest comparison had narrowed: usual care had climbed from 10% to 36%, so the gap between the groups shrank considerably. The effect is real; the long-run advantage over ordinary care is smaller than the headline, and worth saying so. Handle it in both directions: it does not mean all chronic pain is "in your head," and it was a specific group (back pain, no clear structural cause). But it is hard evidence that when suffering is being maintained by the brain's predictions, changing those predictions can change the pain. That's not positive thinking. It's re-training an alarm.
Here a real review has to stop and show the dispute, because it's live and it's sharp.
Pain medicine recently adopted a third category, nociplastic pain (the International Association for the Study of Pain's term): pain from altered processing in the nervous system rather than from ongoing tissue damage or nerve injury. It's the official name for "the alarm is being generated centrally." Fibromyalgia is the flagship example.
And it is contested, in print, right now (see Häuser and colleagues, European Journal of Pain, 2026, "Nociplastic Pain: Facts, Controversies and Future Tasks," and the 2024 "controversy of the concept" papers). The core disagreement is exactly our question: can pain genuinely persist with no continuing input from the body? One camp says yes — the brain sustains it. Critics answer that central sensitisation may still require a continuing noxious signal, and that "no tissue cause found" too easily becomes "it's central," which too easily becomes "it's psychological" — and that road has a dark history of dismissing women, and dismissing real disease as "all in the mind."
Both sides are partly right, and holding both is the honest position: the brain clearly amplifies and can maintain pain — and "the brain is doing it" must never become a licence to stop looking for a first arrow that's still firing. The science that empowers you to work with the second arrow is the same science that, misused, blames you for the first.
The two-arrows idea comes from a non-Western, Buddhist-derived tradition, and — underseen in the pain-management mainstream — there's now a neuroscience of it. In studies of experienced Zen and mindfulness practitioners (Grant and Rainville, 2009; Fadel Zeidan and colleagues, 2011 onward), people are given a standardised painful heat and asked to rate two things separately: how intense it is, and how unpleasant. Trained meditators can pull these apart — feeling the sensation clearly while the suffering attached to it drops. Zeidan's group reported that after just four days of training, meditating during the pain cut unpleasantness by around 57% and intensity by around 40%, with brain activity suggesting the sensory and the emotional layers of pain were being processed more independently.
Two honest notes, because this is exactly where hype lives. First, whether meditation truly decouples sensation from suffering, or simply lowers the whole experience, is itself debated (see the 2020 European Journal of Pain work questioning the decoupling account). Second, this is trained skill, not a quick fix. But as a proof of principle it's strong: the first and second arrows are separable, and the separation can be learned. That is the mechanism under everything in this piece.
Widen from pain to psychological suffering and you meet the Dodo bird verdict (from Alice in Wonderland: "everybody has won, and all must have prizes"): the established talking therapies help substantially, and help about equally. A large 2025 review across 90 trials found real benefit not explained by which method was used; the finding goes back to Wampold and colleagues (1997). It does not mean therapy doesn't work — it works well. It means the specific technique explains surprisingly little; what does most of the work are the shared ingredients — a trusting relationship, a believable explanation, the expectation of improvement, and facing the thing instead of avoiding it. (The verdict is itself contested — a 2014 Clinical Psychology Review meta-analysis reports some real differences between methods — so hold it as "technique matters less than the field assumed," not "technique doesn't matter.") Either way it lands back at the two arrows: a lot of what heals is anything that reliably drains force from the second.
Put pain science, the nociplastic dispute, the contemplative decoupling and the common-factors finding together and a shape appears that none of these fields draws on its own:
Ongoing suffering is a self-sustaining prediction loop. The brain predicts danger; the prediction produces the felt pain-or-distress; the felt experience is then taken as confirmation of danger, which strengthens the prediction. What keeps second-arrow suffering going is not the first arrow but the loop closing on itself — and everything that reliably helps (reappraisal, safety, meditation's decoupling, a trusted relationship, expectation) works by breaking the loop, lowering the "danger" prior enough for it to release.
Flagged clearly as interpretation. But it's the kind you can build.
Modelled on the Optionality study already registered in this programme: build a world where the answer is known, and see whether the pattern holds.
The build. A simple predictive agent that outputs a "pain/danger" estimate from two things: a noisy body signal, and a prior (its current belief about how dangerous the situation is). Crucially, let the prior update on the agent's own output — so predicted danger can feed back and raise the prior.
Two pre-registered predictions:
Why run it. If both hold in a built world, "chronic suffering is a prediction loop with a threshold exit" survives its first real test and earns a look in real data (it also gives the nociplastic dispute a concrete, simulate-able form). If they fail, the proposition is wrong and we say so. Cheap, falsifiable, and a candidate for the next registered simulation and the paid Cross-Domain Proposition. (A simulation can't prove real pain works this way — it tests whether the idea is coherent enough to take to data.)
If suffering is partly a loop the brain runs, then the levers are unglamorous and real: reappraising the threat (changing what the sensation means — the engine of pain reprocessing and most therapy); expectation (placebo is the measurable power of expected safety — it works even when people know it's a placebo); turning toward rather than avoiding (rehearsal and avoidance feed the loop; contact drains it); and not being alone with it (nothing reduces the second arrow like company — the most robust finding in the whole area).
And the line that must be said in the same breath: the dangerous misuse of all of this is to turn it on the sufferer — if the mind adds to it, it's your fault, think your way out. False and cruel. Some first arrows are real and ongoing — disease, injury, poverty, abuse, injustice — and the honest response is to change the situation, not reframe it. Trauma is not a thinking error. "You're adding to it" is something a person can realise for themselves; it is not something to be told by someone who isn't carrying it.
(My own framework — a way of seeing, not established science, not medical advice.)
Recursive Field Theory reads chronic suffering as a system stuck in a loop it can't close — a danger-prediction that keeps re-triggering the conditions that confirm it. Healing isn't deleting the injury or the memory (nothing is deleted); it's letting a stuck loop re-settle into one that can resolve. §6–7 are that lens made testable. Offered as interpretation.
Verification pass before publishing. The Pain Reprocessing figures are the trial's own (66% vs 20% vs 10% at post-treatment); the five-year line is given with the honest caveat that the between-group gap narrowed as usual care improved (55% / 26% / 36%) — the flattering "gains held" version would have hidden that, so it's stated plainly. The meditation-and-pain "decoupling" claim was not left as settled — the 2020 EJP challenge to it is in §4, because the flattering version (that meditation cleanly separates sensation from suffering) is the one that would survive an unchecked edit. The Dodo-bird verdict is given with its 2014 counter-evidence, not as a knockout. The proposition in §6–7 is labelled mine, with stated refutation conditions.
Format & time. Reflective structured learning · ~30–45 minutes.
Learning outcomes. After this you will be able to: distinguish pain/first-arrow from suffering/second-arrow and explain why the distinction is clinical, not dismissive; describe the "pain as the brain's prediction of danger" model and the live nociplastic-pain dispute around it; and state the Dodo-bird finding with its main caveat.
Reflection prompts (write 3–5 lines each, keep for your CPD record):
CPD-eligible structured learning; self-certified, not statutory-regulator endorsement. Log only genuine time spent.
This is educational, not medical or psychological advice, and nothing here is guidance on managing a condition. If you're struggling with pain or distress, a qualified professional can help — and if you're in crisis, please reach out to a local support line or someone you trust.
Primary sources linked in the published version. The weekly Investigation is free, in full — including the disputes. Paid members get the monthly Cross-Domain Proposition, where a testable idea like §7 is developed toward a registered study. Recursive Field Theory is the original work of David Fleming. Not medical advice.
Join the Frontier. The weekly Investigation is free, in full, every week.