Three lines in the margin of Law 30, in a chapter otherwise made of anecdotes about other people:
A line [of poetry] will take us hours maybe; Yet if it does not seem a moment's thought, Our stitching and unstitching has been naught.1
Yeats, Adam's Curse.
Every other voice in this chapter is Greene describing somebody. Rikyū is watched. Houdini is watched. Talleyrand is watched. This is the only place where the person who actually did the concealing says what it was like.
And what he says is harsher than anything in the body.
Greene's law runs on a discount. Show the effort and the audience marks the work down — "causes everything, whatever its worth, to be discounted."2
A discount is a slope. Some concealment, some payoff. Fail partly and you lose partly.
Naught is not a slope. It is zero.
Yeats is claiming that hours of work which fail to read as a moment's thought have produced nothing — not a lesser poem, not a poem marked down for visible strain, but a null result. The labour does not partially convert. It either crosses the threshold into apparent spontaneity or it was not spent on anything.
That is a completely different payoff structure and it changes what a rational person does. Under a discount, half-polishing is half-worth-it. Under a cliff, half-polishing is the worst available allocation — you have spent the hours and bought nothing, and you would have been better off either finishing properly or not starting.
Greene never states a threshold. His chapter reads as though concealment is a dial. The poet in his own margin says it is a gate.
The compound is doing work and it is easy to read past.
Not our stitching has been naught. Stitching and unstitching.
Half the labour Yeats is describing is undoing — lines written and removed, a phrase put in and taken out, the whole afternoon's work reversed the next morning. That is not a poet polishing toward a finish. It is a poet oscillating, and the oscillation is where the hours go.
This matters because it identifies what actually has to be concealed, and it is not effort in the general sense. It is evidence of indecision. A visibly laboured surface says this was hard. A visible history of revision says something worse: the maker did not know what they wanted, and what you are reading is the version they stopped on.
Which is a more specific and more frightening exposure than the chapter's, and it explains the naught. A poem that shows its stitching does not merely look effortful. It looks arbitrary — one draft among many, arrested rather than completed — and an arbitrary poem is not a worse poem, it is not a poem.
Worth noting what Yeats called it, because Greene prints the lines without the frame.
Adam's Curse — the labour Adam is condemned to after the expulsion. The poem's surrounding argument is that writing verse is work of that kind: unglamorous, unrewarded, and, in the lines just after these, harder than "scrubbing a kitchen pavement" while the world calls the poet an idler.
So the full context is not a technique for looking good. It is a complaint. The concealment is not a strategy Yeats has chosen and is recommending; it is a condition of the form that he resents and is stuck with. The world requires poems to look spontaneous, therefore the hours must be hidden, therefore the poet is thought idle, therefore the labour is both compulsory and unacknowledged.
[SPECULATIVE] — Greene gives no indication of the poem's argument and may not have it in view; this reading is from the title and the standard text, not from anything in the source.
Extracted into a power manual, the lines become advice. In place they are a man describing a trap.
Set Yeats beside Houdini, both in this same chapter, and the difference is not artistic temperament. It is epistemics.
Houdini's concealed labour has a settlement event. The manacles open in front of four thousand people, or they do not. Whatever he hid, the hiding is vindicated on the night, publicly, in a way nobody can argue with.
Yeats has nothing of the kind. There is no moment at which a poem is confirmed to have crossed the threshold. Readers do or do not feel that a line arrived by itself, they mostly cannot say why, and they will not report it. The poet finds out, at best, slowly and by inference, and possibly never.
So the two men are running the same strategy against completely different odds:
That is the worst combination available, and it is the ordinary condition of most creative work. Greene draws every case in the chapter from the first category — escape artist, tea master before an emperor, foreign minister whose predictions come true — and addresses the advice to everyone.
You have been at something for six hours and about four of them were undoing.
First, know which kind of thing this is. Is there a moment coming when it will be tested in a way that settles it — a demo, a shipped feature, a room that will visibly respond? Or does it just go out and you never really find out?
If there is a settlement event, hide everything and stop worrying. The result will vindicate the hours whether or not anyone saw them.
If there isn't — and for most writing, most design, most quiet work, there isn't — then Yeats's terms apply, and there are only two honest responses.
Cross the threshold or don't start. Do not spend three of the six hours. Half-concealed labour in an unverifiable medium is the one allocation guaranteed to pay nothing: too worked to be fresh, not worked enough to seem unworked.
And keep the record for yourself. Not to show anyone — to know. In a medium with no verification event, the only thing standing between you and an unfalsifiable belief about your own diligence is a private, honest account of the hours. Yeats wrote his down and put it in a poem, which is one solution, and it is why we know.
Strongest evidence. A named poet, a named poem, a dated life, and a complete quoted stanza — better attributed than almost anything in the chapter's body, which is the standing pattern in this build: the margin is better sourced than the text it decorates.
Tension — Yeats's payoff structure contradicts Greene's. Naught against discounted. A cliff against a slope. Documented above; Greene prints both and does not notice.
Tension — the lines are a complaint, presented as advice. The poem's frame is the curse of Adam. [SPECULATIVE], since the source gives no context.
Tension — the chapter's cases and its readers are in different epistemic situations. Documented above.
🚩 [POPULAR SOURCE] — the quotation is accurate and attributed; no edition or date of composition is given, and Greene's bracketed "[of poetry]" is an insertion into the line. ⚠ The attribution reads "William Buller Yeats"; the poet is William Butler Yeats. A typo in the source, corrected here rather than propagated.
Open questions. Is the all-or-nothing claim true, or is it a poet's overstatement of a steep slope? It is testable in principle — readers could be shown work at graded levels of finish — and the answer changes the correct allocation of effort in every unverifiable medium.
The chapter's Japanese material and this stanza are describing two structurally different concealments, and setting them together sharpens both.
Sen no Rikyū and the Cushions in the Snow hides an apparatus that is outside the finished thing — the cushions never appear in what Hideyoshi sees, they were removed before the effect existed. Yeats's labour is inside the object; the hours are spent on the line itself, and the concealment is of revision history.
Those fail differently and are detected differently. Outside-labour is caught by inference from fit — the lemon that suited the cake too well. Inside-labour is caught by the surface itself: visible correction, the laboured passage, Uccello's perspective overwhelming his own paintings.
Greene's single instruction — conceal — treats them as one problem, and they need opposite defences. You defend outside-labour by breaking the suspicious fit. You defend inside-labour by more work, not less.
Against Greene's Mastery, the tension is the familiar one and this page is its sharpest instance: Mastery argues the hours are the point and should be understood as such; Law 30 says hide them; and the man in the margin is doing both, having hidden his hours in the poem and then written a poem about the hours.
Craft — the imperfection that is supposed to show. Imperfection Is the Beauty holds that visible roughness is often what makes work land — polish past a point strips out the evidence that a person was there, and the audience feels the absence.
Straight against Yeats, who says the roughness is a null result.
They are reconcilable, and the reconciliation is the useful part. Roughness that reads as a decision is an asset; roughness that reads as a residue is the naught. A brushstroke left because smoothing it would kill it has been chosen — the maker is present in it. A line left because the poet ran out of afternoon has been abandoned, and the reader feels the abandonment as arbitrariness.
The insight neither produces alone: what audiences actually detect is not effort or its absence but whether the final state was elected. That reframes Yeats's threshold. A moment's thought is not literally about speed; it is about the finished line reading as inevitable — as though it could not have been otherwise. Stitching and unstitching threatens that not by being visible but by advertising that it could have been otherwise, repeatedly, and nearly was. Which means the concealment being demanded is not of labour at all. It is of contingency.
Behavioural mechanics — the work nobody can price. Costly Signaling holds that a signal's credibility comes from what it demonstrably cost, and that costs which cannot be observed cannot function as signals.
Yeats's hours are, by design, unobservable. Under signalling logic they are therefore worth nothing as a signal — which is exactly his complaint, and the reason the poem says the world calls the poet an idler.
What the pairing produces is why this profession is structurally underpaid rather than merely unlucky. In most fields, hidden labour is hidden from the audience but visible to a market — the employer, the client, the guild knows roughly what the work took. Poetry's concealment runs all the way down: the labour is invisible to readers by the form's own requirement, and there is no intermediary who can see it either. So the work has a cost that can never be signalled to anyone who pays. Greene's chapter presents concealment as a lever the powerful pull. Yeats is describing the same behaviour as a condition imposed on the powerless by the medium, and that difference — chosen concealment versus compulsory concealment — is the one the law never draws, though it decides entirely whether this is a strategy or a curse.
Sharpest implication. Yeats's naught says concealed labour has a threshold, not a slope — hours that fail to produce apparent spontaneity buy nothing at all, which makes half-polishing the single worst allocation available. And what the threshold is actually testing is not effort but contingency: a line reads as a moment's thought when it reads as inevitable, and stitching and unstitching betrays not that the poet worked but that it very nearly came out otherwise.
Generative questions.