Yang-shan has seen the Tao. This isn't in dispute — his teacher agrees it happened. And he isn't at peace. So he does the reasonable thing and asks his teacher what he should be doing now.
Kuei-shan's answer isn't an instruction. It's a refusal to give one:
"All that's important is that your eye is correct. I won't talk about your practice."1
the passage glosses both halves in brackets, and the glosses are his: your eye means your perception of the Tao; your practice means your subsequent work of cultivating realization.1 So a student who has crossed the threshold, and is uncomfortable, and asks for a method, is told that the method is none of the teacher's business — and that the only thing worth checking is whether he is looking at the right thing.
The refusal isn't evasion. There's a doctrinal reason a teacher can't answer how should I practise now, and it's stated in the next breath: “that's why the Stage of True Cultivation Practice doesn't take on any fixed form, and why it only begins after you succeed in Seeing Truth.”3 Two things are said there, and the page rests on both.
It has no fixed form. Not that the form is secret, or advanced, or reserved. There isn't one. Everything a manual could specify — posture, count, sequence, object — belongs to the stages before this one, and those get named as “just the principles to help you get there.”3
And it starts after. Everything a practitioner spends years on is the approach, and the thing the word practice usually refers to hasn't begun yet. Which reverses what the question assumed. Yang-shan asked what to do now that he'd seen the Tao, as though seeing were the end of one programme and the start of another with its own instructions. It's the start of the one that can't be written down.
The consequence for anybody with a manual in their hands is stated flatly, and it's more permissive than it sounds: “if you know the correct principles of practice, you can even invent new cultivation techniques yourself.”3 Techniques are downstream of principles, so somebody holding the principles is licensed to build. Somebody holding only techniques is licensed to copy.
That's also why the Zen houses left no manuals on this material — they “take the route of ignoring these things entirely,” and most cultivated quickly.3 Not a boast about temperament. A consequence of where they thought the instructions run out.
The temptation is to read eye as understanding — get your theory right and the practice follows. The gloss given blocks that, and the block matters.
His expansion at p 247 makes the eye a matter of realization, not comprehension:
As long as your perception of the true state (Tao) is correct, I won't speak of any particular practice that you should engage in. If your realization of the Tao is wrong or incomplete, however, then no matter what you do in your practice it's all wrong.2
Realization, not opinion. The eye isn't a set of beliefs about the Tao that could be corrected by reading; it's the direct perception that seeing the Tao consists of. Which is why the instruction can only be given to someone who has already crossed — you can't check an eye that hasn't opened, and Kuei-shan isn't offering Yang-shan a doctrinal audit.
the passage makes the prerequisite explicit by stacking three authorities who all draw the same line at the same place, on the same page:
The Fifth Patriarch to the Sixth, on his awakening: Before you see your original nature, cultivation will not benefit you.2 A Tibetan formulation: Only after recognizing your intrinsic "awareness-without-content" can the disciple practice Mahāmudrā correctly. Until he has done so, he will find it difficult to escape from the subject-object entanglement.2 And Milarepa, that experiencing the bliss, luminosity and no-thought of the samādhi just isn't enough.2
The second half of the text's gloss is the part with teeth, and it's a strong claim about the structure of error:
If your realization of the Tao is wrong or incomplete, however, then no matter what you do in your practice it's all wrong.2
Not less effective. Wrong. The eye gates the practice rather than adding to it — method contributes nothing at all if the perception it serves is off, and contributes almost everything if it's not. That's why the teacher declines to discuss method. Discussing method with a student whose eye is uncertain would be prescribing a dosage without a diagnosis, and worse, it would let the student substitute technique for the thing technique can't fix.
And it explains the otherwise strange sequencing in the text's own presentation. He spends hundreds of pages on technique — channels, breath, the dhyāna ladder, the completion-stage yogas — and then, at the point where a reader most wants the post-awakening manual, hands them a refusal. The refusal isn't withholding. On his own logic there is nothing there to withhold, because the only variable at that stage is one no manual can carry.
You've been practising for years and you keep changing methods. Breath, then mantra, then a koan, then back. Each one works for a while. On this account the churn is a symptom rather than a search: you're adjusting the variable that is available instead of the one that matters, because the available one has a literature and the other doesn't. The uncomfortable question the passage puts to you isn't which method but what are you actually looking at when you sit — and if that question has no clear answer, no method will supply one.
Now the case Kuei-shan is actually addressing, which is rarer and more specific. Something has happened, it holds, and you're still not at peace. The reflex is to conclude the practice is wrong and go looking for a better one. The instruction here is to do the opposite: leave the method entirely alone, and check the seeing. Not check whether you had the experience — check what you are perceiving now, today, in the ordinary way, and whether it's the same thing.
And the discipline that follows from the refusal. When someone asks you what to practise and they're past the threshold, the honest answer is that you don't know and it isn't yours to say. That will feel like a failure of teaching. On this account it is the teaching — practise according to your own situation and according to your own wisdom1 — and the failure mode is the teacher who fills the silence with a technique because a refusal is hard to charge for.
The Kuei-shan exchange arrives at p 573 as the punchline to an attack on a genre, and the setup is what makes it argumentative rather than merely pious.
the passage has been criticising extensive manuals containing all sorts of complicated instructions with sub-branches and sub-definitions — his description of a good deal of the Tibetan literature. His complaint isn't that they're wrong. It's that they're unnecessary:
As to how an individual should guide themselves on the spiritual path, the overall process is like cooking rice over a hot fire: once you know the basic principles and instructions for this simple process, it just requires natural wisdom to know how and when to adjust the fire.3
Then: Perhaps even this explanation says too much — and the Kuei-shan quotation follows.3 The rice image is doing precise work. Cooking rice has a small number of principles and an unbounded number of situations. No manual can enumerate the situations, because the variables — the pot, the heat, the grain, the altitude — don't decompose. What transfers is the principle plus the judgement to apply it, and the judgement isn't written down because it can't be.
the passage draws the radical consequence himself, and it's the most permissive sentence in the book: if you know the correct principles of practice, you can even invent new cultivation techniques yourself.3 That's a striking thing for a lineage teacher to print. It makes technique downstream and disposable — a local solution to a local situation, generated by whoever has the eye, rather than a transmission that must be preserved. And it follows directly from the gating claim. If the eye determines everything and method determines nothing, then method can be improvised at no risk, and the manuals aren't merely excessive but misdirected: they're the elaboration of the variable that doesn't matter.
The tension this creates with the rest of the book isn't small, and It doesn't address it. He has just written six hundred pages of exactly the thing he is objecting to.
the text's comparative move here is unusually restrained and therefore unusually convincing. He isn't claiming four traditions describe the same state. He is claiming four traditions impose the same ordering constraint:
What they agree on is an order of operations, which costs a tradition far less than agreeing on what a state is like — and cheap agreement is likelier to be real. Four traditions could describe awakening in four incompatible ways and still all find that method stops helping until something else has happened. Then extends the point beyond method to the whole enterprise: spiritual cultivation is nonsectarian, quite scientific in nature, and fundamentally natural — you don't have to follow any artificial route, and the artificial cultivation techniques that have been espoused over time have only been invented to help you speed your progress.3 Techniques are accelerants, not the road. And a page later, without apparent irony: Benchmarking is not just proper in business, but proper if you want to speed yourself along the path of spiritual progress.4
The last observation is the one that explains an absence in the literature, and the passage states it as a historical fact rather than a doctrine:
Most of the Zen masters never dared speak of this gate of practice, for while it's easy to talk about seeing the Path, or seeing the Tao, it's very difficult to talk about cultivating realization.1
Then the example he attaches, which is the Second Patriarch found in the red-light district, answering I'm learning to pacify my mind. What is it to you?1 Put the two together and you get the shape of the silence. Post-awakening practice is situation-specific and improvised; a teacher describing their own would be describing something non-transferable; and described from outside, it looks like misconduct. A practice that is individual, unwritten and externally indistinguishable from error isn't a thing a tradition can transmit, and so what got transmitted instead was everything up to the threshold — which is exactly the shape of the literature The passage is complaining about at the start of the chapter.
The complaint and the explanation are on facing pages, and he doesn't connect them.
Source criticism, and this is the whole budget. The Kuei-shan exchange is quoted three times across the book and rendered three ways — “dharma eye of enlightenment” at p 247, “perception of the truth” at p 573, “your eye [your perception of the Tao]” at p 574 — with no locator on any of them.1 2 3 The three aren't equivalent: the first is a technical term with a doctrinal history, the second a general claim about knowing, the third neither. The shape is the same in all three, and that's what this page carries.
Strongest support. The rice analogy is the best argument on the page because it identifies why the manual fails rather than merely asserting that it does.3 Where a field has few principles and situations that won't break into parts, no procedure will cover it — and that's about the shape of the field, which you can look at without touching doctrine.
The tension the source does not close. the passage attacks extensive manuals as needless complication3 in a six-hundred-page manual of extensive complication. He never addresses the reflexive problem, and the closest he comes — that the artificial techniques exist to help you speed your progress — would license the Tibetan literature he is objecting to just as well as his own.
A second tension, on the three renderings. The same exchange is quoted three times with materially different wordings — dharma eye of enlightenment, perception of the truth, your eye — none with a locator.1 2 3 The differences aren't stylistic: the first is a technical term, the second a general claim about knowing.
A third, on the licence to invent. If you know the correct principles of practice, you can even invent new cultivation techniques yourself3 is the most permissive statement in the book, and it's unguarded. Elsewhere The passage is severe about practitioners improvising — guiding chi with the mind is condemned at length on the immediately preceding page. The licence and the condemnation are adjacent and unreconciled.
A fourth, on the gating claim. No matter what you do in your practice it's all wrong2 is stated without qualification and would, taken literally, mean that all pre-threshold practice is worthless — which contradicts the entire Path of Preparation the book spends five chapters describing.
And the standing scope limit. The eye being checked is perception of a stage that footnote 518 calls a false stage, with three real jobs named for the scheme — rest stops, verifications and motivational assists.5 The bite here is specific and sharp: verification is one of the three, and this passage is the book's clearest instance of the scheme being used for exactly that and nothing else.
Open questions. If technique is improvisable by anyone with a correct eye, what's a lineage for — transmission of the eye, or of the licence to improvise? And the gating claim says a wrong eye makes all practice wrong: is there any diagnostic for a wrong eye that doesn't require an already-correct one?
Read against What a Guru Is Actually For, this page supplies the sharpest possible answer and it's a subtractive one. The teacher's function here is not instruction but verification of the eye, and the refusal to prescribe is the service. That narrows the guru's job to something checkable and removes most of what the role is usually paid for.
With The Legacy of the Waves the fit is exact and the two pages explain each other. That page says the post-awakening stage has no fixed method because its whole content is withholding energy. This one says the teacher therefore declines to discuss method. Same silence, two derivations — one from what the stage consists of, one from what a teacher can honestly say about it.
Against The Book That Warns Against Its Own Genre, the reflexive problem is at its most acute here. Elsewhere the book criticises a genre it belongs to; here it criticises manual length at page 573 of a manual, and offers the rice analogy as proof that the whole business was avoidable.
Against Verification and Its Broken Chain, there's a live difficulty. That page's problem is that verification requires a verified verifier, and the chain doesn't close. Kuei-shan's rule makes the eye the only thing that matters and offers no way to check it except another eye. The rule maximises the load on precisely the mechanism that page shows to be broken.
The variable you cannot delegate — Taste, Judgment, Labor Framework. Simmonds's framework splits creative work into three roles: Taste (the aesthetic anchor — what "good" means here, codified as constraint), Labor (the physical generation of the artefact), and Judgment (the pass, where every output is checked against the Taste before it ships). The framework's whole point is that Labor is fully delegable — to a team, to a model — and Taste isn't, and that a creator who delegates Taste loses the thing that made the work theirs. Set Kuei-shan's rule beside it and the correspondence is joint-for-joint: the eye is Taste, the practice is Labor, and the teacher's refusal is the assertion that only one of the two is worth a conversation. What the pairing produces that neither side produces alone is a diagnosis of why both fields are drowning in the wrong literature. Simmonds's context is a market flooded with prompt-technique content and almost nothing on developing taste; the text's is a tradition flooded with manuals of sub-branches and sub-definitions and almost nothing on the post-threshold stretch. In both cases the delegable component is the one that generates a literature, and it generates one precisely because it's delegable — a thing that can be handed to someone else can be written down, sold, standardised and taught in a weekend. The non-delegable component generates almost nothing, not because it's ineffable but because it can't be packaged. Which yields a general rule with real predictive force: the volume of instructional material in a domain is inversely proportional to how much the instructed variable matters, and a field's most confident literature will reliably concentrate on its least decisive input. That explains the shape of the text's complaint better than his own explanation does, and it explains it without needing anyone to be dishonest.
One handshake. A second — that judgement is unwritable because situations don't decompose — is the rice analogy restated in another vocabulary rather than a connection that produces anything new.
Sharpest implication. A domain's instructional literature grows on whatever component can be handed to somebody else, which means the most thoroughly documented part of any practice is by construction the part that matters least — and the silence around the decisive variable reads as mystification when it's actually just the absence of anything packageable.
Generative questions.