The Four Examinations (望聞問切 Wàng Wén Wèn Qiè)
The oldest clinical interview, still performed at the wrist
Evidence grade 2 of 5, Thin. Read how the Atlas grades evidence.
Almost no study asks whether the four examinations improve outcomes, because the framework is treated as a given rather than a hypothesis. What has been tested is agreement. Jacobson and colleagues at Harvard Medical School reviewed 21 experimental studies in 2019 and found a mean Cohen's kappa of 0.34 across the seven studies reporting it, range 0.07 to 0.59, which they characterised as low to moderate agreement amid widespread methodological deficits. A 2014 multicentre Korean study of 168 stroke patients found kappas from 0.26 to 0.84 depending on the sign, with pulse palpation between 0.37 and 0.61. Schnyer's randomised trial found structured training raises agreement, which also confirms how unreproducible untrained practice is.
Reference only. Not medical advice.
What can go wrong
Nothing in the examination can injure a patient. The hazard is what it can miss. A pattern framework has no way to detect a pathology for which it holds no category, and a patient whose weight loss and coated tongue are read as Spleen qi deficiency may never be referred for the endoscopy that would find a gastric cancer. Most regulators permit pattern diagnosis only alongside conventional assessment rather than in place of it. The wrist is also the entry point for a good deal of commercial overreach, including practitioners who claim to detect pregnancy, the sex of a fetus, or cancer from the pulse, none of which has been demonstrated under blinded conditions. Reported agreement on far simpler pulse qualities is already poor, which makes the stronger claims difficult to credit.
Chinese medicine gathers clinical information through four channels. Wàng, inspection, covers complexion, bearing, and above all the tongue, its body colour, shape, and coating. Wén covers listening and smelling together, since one character serves for the voice, the cough, the breath, and the odour of a sickroom. Wèn is the interview, organised in classical texts as ten questions covering chills and fever, sleep, appetite, thirst, stool and urine, and in women the menstrual history. Qiè is palpation, most famously the radial pulse taken at three positions on each wrist, each assigned to a set of organs. The four are meant to be combined and never used alone, and what they produce is not a disease name but a zhèng, a pattern, which is what the prescription is then written against. The framework is systematic. Whether two trained practitioners examining the same patient arrive at the same pattern is a separate question, and the reliability literature answers it unfavourably.
Chinese medicine's four-part exam, looking, listening and smelling, asking, and palpating the pulse; a structured method whose pattern verdicts agree poorly between practitioners.
How it is supposed to work
Parts of the framework rest on real physiology. A yellow sclera, a wheeze, wasting, a fever history, an irregular pulse are real signals. What has never been established is the mapping layer, the claim that the cùn, guān and chǐ positions report on specific organs, or that a tongue coating localises a disturbance to the Spleen. No anatomical route has been shown for them. Tongue imaging and pulse-wave sensors measure something real, but what they capture is haemodynamics and mucosal colour, not the entities the scheme names.
What people get wrong
Often saidA skilled practitioner can read pregnancy, the sex of a fetus, or a hidden cancer from the pulse.
ActuallyNone of these has survived blinded testing. Agreement between practitioners on much simpler pulse qualities averages a kappa near 0.34, which is weak ground for any stronger claim.
Often saidThe four examinations are unscientific hand-waving with nothing real inside them.
ActuallyInspection, listening and inquiry collect genuine clinical signs, and Chunyu Yi's Han-era case records are among the earliest structured patient histories known. The unsupported part is the organ-mapping layer, not the act of careful observation.
Often saidWHO's ICD-11 chapter means traditional Chinese diagnosis is now validated.
ActuallyWHO states plainly that the Traditional Medicine chapter neither judges nor endorses the scientific validity of any practice, and exists so that traditional-medicine encounters can be counted and studied. A 2019 Nature editorial argued the move could still be read as legitimisation.
Who profits, and who pushes back
Eric Jacobson, PhD, MPH
Reviewing 21 experimental studies, Jacobson and colleagues reported a mean Cohen's kappa of 0.34 and concluded that with few exceptions the levels of agreement were low to moderate, with significant deficits of methodology and reporting across the literature.
Rosa N. Schnyer, DAOM
Schnyer's group found that structured training substantially improved agreement on Chinese medicine diagnoses, and has argued that conventional inter-rater reliability designs may misfit a system in which diagnosis is reassessed and revised at every visit.
Harriet Hall, MD
Hall argues that the pulse and tongue organ maps have no basis in anatomy or physiology, and that a system in which each practitioner's differing diagnosis is deemed correct places itself outside the scientific paradigm.
World Health Organization
WHO states that the Traditional Medicine chapter is neither judging nor endorsing the scientific validity of any traditional medicine practice or the efficacy of any traditional medicine intervention, and that it serves as a tool for counting and comparing conditions so that efficacy can be researched.
Where the research comes from
Most of the published research on this comes from one place. 100% of it is from Formalised in Han-era China, carried into Korean, Japanese and Vietnamese medicine, now taught wherever Chinese medicine is licensed.. Findings that no unrelated group has reproduced are weaker than the study count makes them look.
| Place | Share of published research |
|---|---|
| Formalised in Han-era China, carried into Korean, Japanese and Vietnamese medicine, now taught wherever Chinese medicine is licensed. | 100% |
How we got here
The four methods circulated separately in Han-era practice and were only later bundled and ranked, the Nàn Jīng's sixty-first Difficulty giving them their canonical order under the tag 神聖工巧. The numbered form 四診 became standard in imperial-era teaching, while the mnemonic 望聞問切 that most people know today belongs to twentieth-century curricula in the People's Republic. The scheme entered Korean medicine as sajin and Japanese Kampo as shishin, where practitioners added abdominal palpation (fukushin) to the palpation limb, and English textbooks generally render it as the Four Examinations or the Four Pillars.
The individual methods are older than the scheme that unites them. The Huángdì Nèijīng, assembled from Warring States and Han material and placed around the first century BCE, already discusses complexion, voice and vessel palpation without numbering them. Sima Qian's Shǐjì, written around 90 BCE, preserves 25 case records of the physician Chunyu Yi that read as diagnostic narratives built from looking, asking and the pulse, among the oldest structured patient histories known. The Nàn Jīng, compiled in the first or second century CE, gives the four their canonical ranking in the sixty-first Difficulty. Wang Shuhe's Mài Jīng, produced under the Western Jin in the third century, was the first text devoted to the pulse, systematising 24 types and fixing the wrist as the site. Li Shizhen's Bīn Hú Mài Xué of 1564 extended the canon to 27 qualities, and that version is what is taught today.
- c. 1st c. BCEThe Huángdì Nèijīng, assembled from earlier material, discusses complexion, voice and vessel palpation without yet numbering them
- c. 90 BCESima Qian's Shǐjì preserves 25 diagnostic case records of the physician Chunyu Yi
- 1st to 2nd c. CEThe Nàn Jīng's 61st Difficulty ranks the four methods as shén, shèng, gōng and qiǎo
- 3rd c. CEWang Shuhe's Mài Jīng, the first dedicated pulse text, codifies 24 pulse types and fixes the radial site
- 1564Li Shizhen's Bīn Hú Mài Xué expands the canon to 27 pulse qualities, the framework still taught
- 2019Jacobson and colleagues publish a systematic review finding a mean inter-rater kappa of 0.34; Nature publishes an editorial warning that WHO's traditional-medicine chapter could backfire
- Jan 2022ICD-11 Chapter 26, Traditional Medicine Conditions Module 1, comes into effect for optional dual coding
How it is understood
Diagnosis in the classical setting was a social performance as much as a technical act. The Nàn Jīng ranks the four methods by the skill each displays, calling the physician who knows by looking a shén, a spirit, and the one who must resort to the pulse a qiǎo, merely clever. Pulse-taking also solved a problem of decorum, since a physician could read a gentlewoman's wrist without examining her body. The still room, the three fingers laid along the artery, and the long silence remain part of what many patients understand a consultation to be.
At a glance
- The four
- 望 inspection, 聞 listening and smelling, 問 inquiry, 切 palpation
- Canonical ranking
- Nàn Jīng, 61st Difficulty (神聖工巧)
- Mean inter-rater kappa
- 0.34 across seven studies (Jacobson et al., 2019)
- Output of the exam
- A zhèng, a pattern, not a disease
What this entry does not yet cover
What follows is unfinished work on our side, not a finding about The Four Examinations (望聞問切 Wàng Wén Wèn Qiè). A short entry is not a safe entry.
- Cost and market. We have not documented what it sells for or who sells it.
- Supply chain. We have not mapped where it is made.
- Who uses it. We have not documented the user groups.
- What could change. No forward-looking summary yet.
- How it compares. We have not set it against the alternatives.
- Children and teenagers. We have not checked what is known about younger people.
- What people take it for. We have not catalogued the claims made for it.
Sources
The Nàn Jīng's 61st Difficulty ranks knowing by looking (shén), by listening (shèng), by asking (gōng) and by pulse (qiǎo)
Mean Cohen's kappa 0.34 (range 0.07 to 0.59) across seven studies; 21 studies reviewed; agreement low to moderate
Multicentre Korean study of 168 stroke patients; kappa 0.26 to 0.84 by sign, pulse palpation 0.37 to 0.61
Tongue-sign agreement between experts moderate to good, weakest for moist fur and spotted tongue
Structured training improves inter-rater agreement on Chinese medicine diagnoses in a randomised trial
An automated tongue diagnosis system showed higher intra-rater agreement than TCM practitioners (kappa 0.93 vs 0.64)
The pulse and tongue organ maps lack an anatomical or physiological basis; reliability studies are weak
Hall H. Pulse Diagnosis and Tongue Diagnosis in Traditional Chinese Medicine. Science-Based Medicine
The ICD-11 Traditional Medicine chapter neither judges nor endorses scientific validity or efficacy; used for optional dual coding
World Health Organization, Classifications FAQ: Traditional Medicine
Nature editorial argued WHO's traditional medicine chapter risks legitimising an unfounded framework
Bian Que regarded as the early authority on pulse lore, Wang Shuhe as the chief systematiser
Chapter 26 Traditional Medicine Conditions Module 1 came into effect in January 2022
Incorporation of complementary and traditional medicine in ICD-11. BMC Med Inform Decis Mak. 2022
Huangdi Neijing dating and content