The Reader Who Reads Between Lines Reads Harder

A reader in Tehran once wrote to me about a poem I had published in 2007, explaining that the pigeon in the third stanza was clearly the Central Bank governor, based on a rumour about the man's habit of feeding birds outside his office window. The pigeon was, in fact, a pigeon. I had not built any code into that stanza at all. The letter taught me something I have used ever since, and it was not about pigeons.

A reader trained under censorship, or under any regime where meaning is routinely hidden on purpose, does not read a text the way an untrained reader does. She reads every image as a possible container, every plain noun as a candidate for encryption, because for years the texts she was given to read actually worked that way. This produces a reader who works far harder than the writer asked her to, and the ordinary response to this, among writers I know, is mild irritation. I want to argue the opposite. That extra labour is not a misreading to be corrected. It is a resource sitting unclaimed in the room, and a writer who designs for it gets a poem that pays out more than the one written for a reader who reads once and stops.

The second reading budget

Here is the operational version. Before a piece is finished, I estimate its second reading budget: the amount of additional, legitimate interpretive work a suspicious reader can spend on it and be rewarded, as opposed to punished, for spending. A poem with a budget of zero rewards nothing on a second pass, every image resolves exactly as it did the first time, and a reader trained to hunt walks away either bored or, worse, inventing a pigeon that was never there because the piece gave her nowhere real to put her attention. A poem with a healthy budget has at least one image built with enough excess capacity that a second, harder reading finds something true that a first, casual reading missed, without that discovery contradicting the first reading. The budget is not infinite. A poem asked to reward unlimited suspicion becomes a puzzle box, and a puzzle box is a different, lesser object than a poem.

I set my own budget now at exactly one image per piece, built on purpose to hold a second load. In a poem about my mother's kitchen, the kettle she used for thirty years appears twice, once in a line about routine and once, six stanzas later, in a line about the year she stopped using it. A first reading takes the kettle as furniture. A second, more suspicious reading notices that the kettle is the only object in the poem that changes state without comment, and that its silence is doing the same work the missing subject of the poem, my father's absence, is doing everywhere else. Both readings are correct. Only the second one is complete, and I built it that way on purpose, for exactly the kind of reader who wrote to me about the pigeon.

Why most writers underfund this and pay for it later

Most drafts spend their entire attention on the first reading, because the first reading is the one the writer actually experiences while revising, over and over, until it feels finished. The second reading belongs to a reader the writer has never watched work. It is easy to forget a budget exists for someone whose labour you cannot see happening. I only started tracking it after enough letters like the pigeon one taught me that a trained reader's suspicion does not go away just because I decline to feed it, and a poem with nothing legitimate to reward that suspicion sends her looking in places I never built, which is a worse outcome for both of us than simply building somewhere real for her to land.

The case against this

You could argue that designing for a hypothetical suspicious reader is a way of writing for the wrong audience entirely, and that most readers, in most countries, have no training in this kind of hunting and will simply take the kettle as a kettle, meaning the whole exercise serves a shrinking population of readers shaped by a specific political history. That is true, and I do not think every poem should carry a second reading budget for this exact kind of reader. Some pieces are built to resolve completely on a first pass and gain nothing from a second one. What I am arguing against is the assumption that the harder reader's extra labour is automatically wasted or automatically an error, an assumption I held myself until enough letters arrived proving that some of that labour, redirected onto real capacity, produces a truer poem than the one built only for a single, trusting pass.

I wrote back to the reader in Tehran and told her the pigeon was only a pigeon. She wrote again, unconvinced, certain I was protecting myself by denying the code. I have never been able to tell her that the real code, in a different poem entirely, was in a kettle she had probably never even noticed.