A monk asked Zhaozhou: “Does a dog have buddha-nature or not?” Zhaozhou said: “Mu.”

That is the first koan in the Gateless Gate, and for a great many Rinzai practitioners it is the first koan they receive and the one they work with for years. It is also, on its face, baffling — particularly since Mahayana doctrine holds unambiguously that all sentient beings have buddha-nature, so the doctrinally correct answer is known in advance.

That is the point.

What a koan is #

Koan (Chinese gong’an) means “public case” — originally a legal term for a magistrate’s precedent-setting judgement. The koans are records of exchanges between Chan masters and their students, collected from the Tang period and formalised as a training curriculum in the Song.

They are not riddles with clever answers, and they are not nonsense designed to short-circuit thought — a popular characterisation that is inaccurate in both directions. Victor Sogen Hori’s work has shown that koan practice is a highly structured pedagogical system with its own literature, its own curriculum, and its own criteria for a correct response.

What a koan does is present the student with something that cannot be resolved by the ordinary operation of the discriminating mind, while requiring that they resolve it. The resulting pressure is the practice.

The great doubt #

The technical term is dai gidan, great doubt — and “doubt” here is misleading if read as scepticism. It names a state of total concentrated perplexity in which the koan occupies the whole of attention.

The classical instruction, from Wumen’s commentary on Mu: rouse your entire body with its three hundred and sixty bones and joints and its eighty-four thousand pores into a mass of doubt; concentrate on Mu day and night, without interruption; do not construe it as nothingness, and do not construe it as being or non-being. It should feel, he says, like swallowing a red-hot iron ball that you can neither swallow nor spit out.

Hakuin, the eighteenth-century Japanese master who reorganised Rinzai training and gave it its present form, made the formula explicit: great faith, great doubt, great determination. Faith that resolution is possible; doubt as the engine; determination to sustain it.

The state is deliberately uncomfortable. The discursive mind attempts solution after solution, each of which the teacher rejects, until the apparatus of solution-seeking exhausts itself.

The breakthrough #

What is sought is kensho — “seeing one’s nature” — a direct, non-conceptual recognition that resolves the koan not by answering it but by dissolving the standpoint from which it was a problem.

Two things are consistently said about this.

It is a beginning, not an end. Rinzai treats initial kensho as the opening of the path rather than its completion, and the remaining curriculum — often hundreds of koans over decades — is the work of integrating and deepening it. Hakuin was emphatic on this point and critical of practitioners who took a first experience as the finish.

It cannot be faked in the interview. Which brings us to the mechanism that makes the whole system work.

Sanzen #

Koan practice is inseparable from sanzen or dokusan, the private interview with the teacher — typically several times a day during sesshin.

The student enters, bows, states the koan, and presents their response. The teacher accepts it or rings a bell, and the student leaves. There is usually no discussion.

The interview is what makes the koan curriculum function as a training system rather than a literary exercise. The teacher has been through this koan, under their own teacher, and knows the difference between an intellectually constructed answer and a demonstrated one. Responses are frequently not verbal at all — a gesture, an action, a sound.

This also answers the obvious modern question. Published collections exist that give traditional responses to koans, and reading them is useless, because what is being tested is not the content of the answer but the state from which it comes. A student who presents a memorised response is, by all accounts, immediately obvious.

The curriculum #

Hakuin’s reorganisation produced a graded system, still used in Japanese Rinzai and in its Western branches, though the details vary by lineage.

Hosshin (dharmakaya) koans come first — the breakthrough koans, of which Mu and “the sound of one hand” are the standard openers.

Kikan (dynamic action) koans train the ability to function from the insight in particular situations.

Gonsen (explication of words) koans work with the relationship between realisation and language.

Nanto (difficult to pass through) koans are the hardest.

Goi (five ranks) koans, based on Dongshan’s schema of the relationship between the absolute and the relative, and the precept koans, complete the curriculum.

A distinctive Japanese element is the jakugo or capping phrase. Having passed a koan, the student is required to find a phrase — from a large anthology of Chinese poetry and classical literature — that expresses its point. The student searches the collection themselves. This is a substantial literary education, and it is the aspect of koan practice most invisible in Western accounts.

The main collections #

The Gateless Gate (Wumenguan, Japanese Mumonkan), compiled by Wumen Huikai in 1228. Forty-eight cases, each with a brief commentary and verse. The most accessible entry point.

The Blue Cliff Record (Biyanlu), compiled by Xuedou with commentary by Yuanwu, early twelfth century. One hundred cases with layered commentary, and literarily the more sophisticated of the two. Its history is instructive: Yuanwu’s student Dahui is said to have burned the printing blocks, on the grounds that students were studying the commentary instead of doing the practice.

The Book of Serenity (Shoyoroku), a hundred cases compiled in the Soto lineage, showing that Soto used koans too, though differently.

Rinzai and Soto #

The standard contrast — Rinzai does koans, Soto does shikantaza — is broadly accurate about emphasis and misleading about history.

Dogen, Soto’s Japanese founder, studied koans extensively and quotes them constantly throughout the Shobogenzo. Soto’s objection was not to koans but to the instrumental use of them — practice aimed at producing an experience, which Dogen regarded as a misunderstanding of what practice is. In Soto, koans are studied and illuminated rather than assigned as problems to be broken through.

Dahui Zonggao, the twelfth-century Chinese master who formalised kanhua or “observing the phrase” practice, was arguing against a Soto-precursor position he called “silent illumination,” which he characterised as quietism. The argument has been running for eight hundred years.

Working with a koan #

Practically, for someone in a Rinzai or Sanbo Zen lineage:

You are assigned one. You do not choose your koan, and working on koans from a book without a teacher is not koan practice in any sense the tradition recognises.

You hold it continuously. Not only in sitting. Hakuin’s instruction is to carry the koan through walking, working, eating, and lying down, until it is present without effort.

You do not analyse it. The instruction is explicit: do not treat it as a puzzle, do not consider what the master meant, do not construct interpretations. Wumen’s phrase for the intellectual approach is “scratching an itch through your boot.”

You present, and are refused, repeatedly. This is normal and goes on for a long time. The refusals are the practice.

It takes years. Mu is commonly worked for one to three years before passing, and the full curriculum takes decades.

Common questions #

What is the answer to Mu? There is no answer that can be transmitted in this format, and this is not evasiveness. The response is demonstrated in an interview with someone who can tell the difference between demonstration and imitation. A response read here would be exactly the kind of borrowed answer the system exists to detect.

Are koans nonsense? No. Many are perfectly coherent exchanges whose difficulty lies in what they point at rather than in their grammar. The “meaningless riddle” characterisation comes largely from mid-twentieth-century Western popularisation.

Can I do koan practice without a teacher? Not in any form the tradition recognises. The interview is constitutive of the method, not an optional check.

Is this practice only in Japan? Koans originated in Chinese Chan and are used in Korean Seon as hwadu practice, which is closely related — Korean practice typically works a single hwadu for many years rather than progressing through a curriculum. Western Rinzai and Sanbo Zen lineages use the Japanese system.

Is kensho the same as stream-entry? The traditions use different maps and do not map onto each other cleanly. Rinzai’s own account treats initial kensho as a beginning requiring long subsequent cultivation, which is at least structurally similar to the Theravada account of stream-entry.

Sources & further reading #

Victor Sogen Hori, Zen Sand: The Book of Capping Phrases for Kōan Practice (University of Hawai’i Press, 2003) — the introduction is the best scholarly account of how koan training actually works. Wumen Huikai, The Gateless Barrier, trans. Robert Aitken (North Point, 1991) — the standard collection with a Western teacher’s commentary. Hakuin, Wild Ivy, trans. Norman Waddell (Shambhala, 1999) — the reformer’s spiritual autobiography. Philip Kapleau, The Three Pillars of Zen (Anchor, 35th anniversary ed. 2000) — includes transcribed accounts of dokusan and of kensho. Yuanwu, The Blue Cliff Record, trans. Thomas Cleary & J. C. Cleary (Shambhala, 1992).