Every Poem Is a Negotiation With Silence
A student in a workshop I ran in 2014 handed me a poem with a note pinned to it: I want this to say everything. I told her it could not, and that the more useful question was which forty percent of everything she was willing to give up first.
Writers talk about silence in poems as though it were a mood, an atmosphere the poem sits inside, something achieved by short lines and white space on the page. That is decoration, and it is available to anyone with a page break. The silence I mean is structural. It is the set of true, relevant things the poem knows and declines to say, chosen in advance, not stumbled into by vagueness or running out of room.
The silence quota
Before I draft anything longer than a couplet now, I set a silence quota: a fixed proportion of the poem's actual content, decided before a word is written, that will be left unstated and carried entirely by inference. For a short lyric, my quota tends to run around a third. For a longer elegy, closer to half. The number itself matters less than the act of setting it before drafting, because a quota set in advance changes which sentences you write first. You stop trying to say the true thing and then trim it down after. You start choosing, before drafting, which true thing gets the silence and which gets the ink, and that choice turns out to be the actual composition of the poem, more than any line by line decision that follows it.
For the student's poem, which was about her grandmother's dementia, we ran the exercise together. The poem knew four things: that the grandmother no longer recognised her, that she had once been unrecognisably cruel to the student's mother, that the cruelty and the forgetting felt, to the student, like a kind of justice she was ashamed of wanting, and that she visited anyway, every Sunday, out of a duty she could not name. She wanted to say all four. We assigned two to silence, the cruelty and the shame, and let only the forgetting and the Sunday visits appear on the page. The finished poem never mentions why the visits feel dutiful rather than loving. Readers ask her about it constantly, which is the quota working exactly as intended: the two silent facts are what readers reconstruct, unprompted, from the two stated ones, and a reconstructed fact lands with more force in a reader's mind than a stated one ever does, because they believe they found it themselves.
Why silence has to be a quota and not an instinct
The reason this needs to be a number, or something close to a number, rather than a feeling you develop with experience, is that unassigned silence defaults toward whatever the writer is least comfortable saying, which is not the same as whatever the poem most needs withheld. Left to instinct, that student would have silenced the shame and kept the cruelty, because shame is harder to sit with on a page than an accusation against someone else. A quota set in advance, argued over with another reader, forces the question open: which silence actually serves the poem, not which silence protects the writer. Those are different questions and instinct alone answers the second one by default.
The obvious objection
Critics of this approach will say that deciding silence quantitatively in advance is exactly the kind of mechanical thinking that kills a poem's spontaneity, that you cannot pre plan what a poem discovers about itself mid draft. I take this seriously because it is often true of other formal constraints and I would be dishonest to claim silence is exempt. My answer is that the quota is a starting position, not a contract. The student's poem changed its silent pair twice during revision, once when a third draft revealed the shame was actually the poem's real subject and the forgetting was scaffolding. The quota did not prevent that discovery. It gave her a fixed thing to revise against, the way a ghazal's radif gives a poet something fixed to write around rather than something that forecloses the writing.
What the workshop taught me, watching this happen across a hundred student poems since, is that a reader can feel the difference between a poem that ran out of things to say and a poem that decided, with discipline, what it would not say. The first reads as thin. The second reads as a poem that trusts you to be a person capable of finishing a thought it deliberately left open, which is a form of respect most poems never risk extending.