What Should I Write

Finding a subject by following what keeps returning

I have been in reading slumps for reasons so ordinary they can feel beneath discussion. My shoulders hurt. My eyes are tired. The season changes and a book that seemed right on Tuesday becomes impossible by Friday. Writing has its equivalents. Sometimes the question what should I write sounds like a failure of imagination, although the obstruction may be exhaustion, embarrassment, an idea at the wrong scale, a form that has stopped fitting, or the suspicion that the work must justify itself before it has even begun.

When I cannot decide what to write, I begin with the material that has returned at least twice and that I still cannot explain. A phrase appears in unrelated conversations. An object survives another attempt to clear the desk. A place feels altered, though I cannot identify what changed. A book irritates me enough that I keep reopening it. I do not think recurrence is mystical evidence. Sometimes the same object is simply still sitting there. Even so, it shows me where my limited attention has already been going without instruction, and I trust that more than I trust a subject chosen because it sounds appropriately writerly.

I have never found a long list of nouns especially helpful at this stage. It gives indecision more furniture. This page follows the question into several different kinds of trouble and offers a substantial way to work inside each one. For developed exercises built around objects, place, dreams, archives, sound, reading, and chance, go to Things to Write About. For the larger practice connecting prompts to reading and research, visit Writing Inspiration. Stay here if you want to discover why no subject feels right, or why the subject you already have still will not open.

What should I write when nothing is asking to be written

There are days when the room seems to have been fully processed. The chair is a chair. The rain is weather. Every headline belongs to the same exhausted argument, and even an unusual experience arrives with a caption already attached. I do not take this as proof that the world has become uninteresting. Usually I have been moving through it too quickly, asking every detail to announce how it could be useful. Attention flattens under that demand. The recovery begins with allowing one thing to remain present before I know what I think about it.

This is one reason I read poetry when I cannot write. A poem can give ten lines to a minor perception without promising a lesson or an application. Francis Ponge can spend an astonishing amount of attention on soap, bread, an orange, or rain. The subject matters because the language keeps finding more relations inside it. I leave that kind of reading more willing to believe that an ordinary object has not exhausted itself merely because I know its name.

Look for recurrence before novelty. Something has probably been following you quietly: the same color in several places, a sentence heard badly through a wall, a building you watch from the bus, the object you touch every morning without looking at it. The point is not to declare the recurrence meaningful. Give it enough attention to see whether it develops edges. Description comes first. Meaning can arrive late, change its mind, or decline the invitation altogether.

Keep a ledger of what returns

For three days, write down whatever returns without being assigned. Include repeated words, images remembered at inconvenient moments, small irritations, subjects you keep looking up, and objects your hand reaches for automatically. Do not turn the entries into themes. On the fourth day, choose the recurrence with the largest gap between familiarity and understanding. Write what you can verify about it. Then mark the point where memory, assumption, or invention begins. If the page remains inert, bring in one external source, perhaps an etymology, map, field guide, catalog entry, weather record, or another person’s account. Let the source change the question rather than decorate the paragraph.

What should I write when choosing one idea feels like losing four

Too many ideas can create the same blankness as none. A notebook fills with beginnings, and each abandoned line seems to accuse the next one of stealing its place. I can spend an astonishing amount of time pretending the decision is the work. Looking for the best idea makes the problem worse because the material has barely begun and I am already asking it to justify weeks or months of attention. I have had better luck asking which idea can answer back.

An idea with some life in it usually contains a fact I cannot absorb immediately, a formal problem, a contradiction in my memory, or another person’s claim that I cannot dismiss. It gives me somewhere to go beyond the language I already own. An idea that produces only atmosphere may still become a poem, but I want to know what happens when the atmosphere meets an object, a document, a voice, or a rule. The encounter gives the piece consequences outside my original mood.

I am partial to constraints because they admit that choosing can be arbitrary. A rule can select the oldest note, the nearest walk, or the question another person would hand you. Chance relieves the idea of having to win a moral contest before you begin. You remain responsible for the work that follows. If the chosen idea produces no friction after an honest session, write down what failed to catch and return it to the notebook. Testing it has not harmed it.

Give three ideas the same twenty minutes

Put three possible subjects on separate pages. Give each twenty minutes and the same conditions: one physical thing, one question that requires research, one possible form, and one detail that complicates your first interpretation. Stop when the time ends, even if a page is going well. Read all three aloud. I would choose the one that changed my breathing, made me want to verify something immediately, or left the most exact kind of uncertainty. Keep the other pages. The exercise selects the current work while leaving the other ideas available.

What should I write when the subject arrives larger than the page

Some of my worst openings begin with a noble noun. Grief. Home. Climate. Faith. Love. The subject arrives as a weather system, present everywhere and difficult to touch. The first paragraph begins explaining what everyone already knows, or makes a claim that a single life cannot support. I look for the smallest material through which the pressure remains visible: a voicemail kept beyond its usefulness, one plant blooming at the wrong time, the route between two apartments, a phrase inherited from a grandparent, or an administrative form that asks the same impossible question in six boxes.

The small thing needs its own life. A repaired coat cannot exist only to symbolize a family. It has fabric, labor, weather, cost, previous repairs, and perhaps a person who remembers it differently. Following those relations prevents the object from becoming obedient too quickly. The larger subject enters through the evidence, including the parts that refuse the meaning I hoped to assign.

Scale can move outward as well. A private image may feel too slight because it has been sealed inside autobiography. The kitchen ingredient has a trade history. The childhood game has regional names. The inherited expression traveled through a language and changed along the way. Research does not reduce the intimacy of these things. It shows how much history was already sitting inside the kitchen, the game, or the sentence.

Forbid yourself from naming the large subject

Write the large subject at the top of a page, then set that word aside. Choose one object, location, or ten minute interval where the subject left physical evidence. Describe the evidence before explaining it. Add a second account from a person, document, or source that changes the scale. Draft a scene or poem that remains with the small material long enough for the larger question to become perceptible. When the draft begins announcing its importance, remove that sentence and see what the detail can carry on its own.

Read until your first question is no longer enough

I do not read around a project simply to collect proof that my first impression was correct. That usually produces a tidy bibliography and a thin piece of writing. Research begins to matter when a book, translation, field guide, oral history, catalog, or public record makes the original idea harder to keep. I have opened a book expecting support and ended up following a translator, a press, and a footnote until the first question looked embarrassingly small. I usually find more material in that change of direction than in the answer I first went looking for.

Reading also deserves time before it becomes useful. I believe people deserve thirty or forty five minutes in which work stress and productivity lose their claim on attention. A poem may return texture to the mind without producing a line for the draft. That is already enough. Sometimes the connection appears later through a rhythm, a withheld explanation, or a structure I would never have invented while staring directly at the problem.

Poetry is especially good at widening the question because it asks what a sentence means and what it means in context, then allows those answers to remain unsettled. The International Poetry Library offers routes through countries, languages, translators, presses, and themes. Choose one that sits slightly outside the project. Read a poem aloud before deciding what it means. Follow the unfamiliar word, the translator, or the press. I also maintain Bookshop lists for translated poetry for readers who want to keep moving outward through books.

A durable reading practice also includes permission to stop. I have reached the middle of poems and realized I was bored, God forbid, and moved on. A research route can end because the source has given you nothing alive to work with. Constant reading should enlarge the number of possible encounters in your life. It should not turn every book into an assignment you owe someone.

Let the source change what you are allowed to say

Write the claim your draft currently assumes, even if the piece never states it directly. Find one reliable source that supports the claim and another that complicates its language, scale, or history. Make a note of what each source permits you to say and what remains outside its frame. Revise from the complication first. If the subject changes, preserve the change. Research has become structural when it alters the order of the work, the confidence of the speaker, or the terms of the question.

Let the object keep some of its own life

I am drawn to found objects because they arrive with more life than I can account for. A receipt, loose button, damaged photograph, stone, ticket, or piece of packaging has already passed through systems of manufacture, exchange, weather, touch, and disposal. The temptation is to turn it into a symbol immediately. Researching the object slows that conversion down. Where was it made. What material is it. Who handled it before it appeared here. What does the place lose if I remove it.

Found material is never simply free material. A note written by a stranger, a photograph from an archive, or an object taken from a place may carry privacy, ownership, grief, or ceremonial use that I do not understand. Those limits can enter the work as part of its form. Photograph the object and leave it where it was. Describe the note without reproducing its intimate language. Build a catalog whose missing entries remain visible. Sometimes the gap tells the truth more clearly than possession would.

Place works the same way. Stay long enough for the location to stop behaving like scenery. Listen for the machine that turns on every eleven minutes. Notice which path people avoid, where the light catches discarded glass, or how the same corner changes after the businesses close. The writing begins when the place develops behavior and your presence becomes one fact among the others.

Ask what the material keeps doing

Strong material can still refuse every available genre. A memory will not become a continuous narrative. Research has too many gaps for an essay that promises authority. The images matter, but description drains them, while the interviews sound more alive than the paragraphs placed around them. This is where I stop asking what the material is about. I spread it out and ask what it keeps doing. It may repeat, interrupt, circle, accumulate, contradict, disappear, or return in altered language. The verbs are often more useful than the genre names.

A project built from recurrence may want a sequence. A project organized by place may want a walk, map, or set of coordinates. A history with uncertain attribution may want a catalog whose labels admit their limits. Several accounts of one event may need a chorus that preserves disagreement. Forms borrowed from weather reports, indexes, instructions, transcripts, or archival folders carry assumptions about order and authority. Using one of them gives the project a structure it can test.

Another medium can expose the form more quickly than another hour of prose. Record the draft and hear where duration becomes awkward. Arrange the notes on a floor and see which materials keep approaching each other. Photograph the central object from a distance that frustrates recognition. Write an instruction another person could perform and notice what you were trying to control. The final piece can return to poetry or prose. It will return with information about its own behavior.

Find the verb before you name the genre

Lay out the notes, quotations, objects, recordings, images, and remembered scenes. Write a verb beside each item describing what it does inside the project. Group the repeated verbs and choose the one that creates the strongest relation among unlike materials. Build a small version of the work that obeys that action. If the verb is return, bring the same image back under changed conditions. If the verb is interrupt, give another voice or medium the power to break the sequence. Let the action determine the experiment before you decide what shelf the finished work belongs on.

The private draft and the imagined witness

Sometimes what should I write means what am I willing to be seen writing. The subject has already arrived, but it feels sentimental, minor, obsessive, politically incomplete, embarrassing, or unlike the work a serious person is supposed to make. I recognize the respectable substitute when it appears. Its terms are already in order and it can defend itself immediately. The stranger and more particular material waits somewhere off the page.

I feel a version of this whenever I read poems aloud to a computer. I tend to record when I am alone because the whole act can feel fundamentally embarrassing. Then the poem begins, attention moves toward the language, and the embarrassment becomes less interesting. Writing may need the same private interval. The obvious image, excessive tenderness, unattractive detail, unresolved anger, or obsessive list can remain long enough to become more complicated than the judgment that surrounded it.

Embarrassment and ethical concern ask for different responses. Embarrassment may tell you that the material is live. Ethical concern may require research, consent, changed names, distance, or removal. A private draft gives you enough material to tell the difference before publication becomes part of the decision. The first version can leave a bracket where another person’s privacy or a missing fact prevents you from proceeding.

Write before the committee arrives

Give yourself forty five minutes and write the version that would be difficult to explain to an imagined committee. Name the detail you keep replacing with abstraction. Include the sentence that feels too earnest. Mark any place where the work enters another person’s life without permission. Close the document for a day. When you return, make two lists: what embarrassed you and what created a real ethical obligation. Revise the second list with care. Read the first list again before you cut anything.

Some ideas end by changing what you notice

A project can end without becoming the piece I imagined. The research answered the question. The image belonged to a season that has passed. The formal experiment taught me something useful and produced one page. Continuing would require manufacturing urgency, then blaming the sentences for refusing it. We count unfinished projects against finished ones, although some material completes its work by changing the next thing we are able to notice.

Before I set one aside, I write an end note. I record the question that began it, what changed, which materials remain alive, and why the current form stopped working. I save the sources and name any factual or ethical problem that prevented continuation. The note turns a vague defeat into an archive. Months later, another project may need the abandoned image, source, or structure. I can carry that part forward without pretending the old project needs to be revived intact.

Then I choose something with a different demand. After months of research, I may work from sound or observation. After an intensely private draft, I may go sit in a public place. After polished prose, I may write an instruction someone else can perform. The change gives attention another surface and keeps one difficult project from becoming a theory about my ability to write at all.

What I would write tonight if I had no idea

If I had forty five minutes tonight and genuinely did not know what to write, I would choose the thing that returned twice this week. I would spend ten minutes describing it without interpretation, including the embarrassing or apparently irrelevant details. Then I would find one source that changes its scale: the origin of a word, the material history of the object, a map of the place, a poem from another language, or another person’s memory. I would copy one sentence from that source by hand and write beside it until my own language began to argue, echo, or wander.

For the remaining time, I would give the material one rule. The rule might forbid the name of the subject, require the same image to return under different conditions, or ask the piece to alternate between verified fact and admitted invention. I would stop when the time ended and read the page aloud. I would mark the line that surprised me and the line where my voice went automatic. Those two marks are enough to begin the next session. I would not need a thesis, a title, or proof that the idea deserves to become a finished work.

Keep going according to what the page asked for

If you found a recurring subject and want to stay with it for an afternoon, choose one of the extended exercises in Things to Write About. If the material is becoming a poem, continue to Poem Ideas or Poetry Prompts. If it wants reflection, narrative, memory, or an image led route, use Journaling Ideas, or Story Ideas. Choose the next page according to the kind of attention the material requested. The genre can become clearer after you have spent more time with it.

If you found this helpful and wish to support me, follows help!