You sit down to write and discover that the thing you wanted to say has become strangely difficult to say. A thought that was perfectly clear on a walk now looks thin on the page. You try an opening sentence, delete it, and wonder whether you should have read another book before attempting this.
Part of the trouble is how much we ask of a beginning. It has to give us a subject, sound like us, and somehow prove that the rest will be worth writing. That's a lot of work for a sentence we haven't written yet.
A more forgiving place to begin is with something you have already noticed. A conversation you keep replaying. The house across the street with its curtains always closed. Something you meant to ask your father, then didn't. You can put that on a page before you know what it means.
Give yourself something to write about
In On Keeping a Logbook, Austin Kleon describes recording ordinary facts about his days: where he went, what he did, who he met. He found lists easier to sustain than a prose diary. They also let him record things without deciding, on the spot, whether they mattered.
That last part is useful when you feel you have nothing to say. The demand for an interesting subject can make you discard your material before you've looked at it. An uneventful Tuesday still has particulars. Someone said something oddly. You took the long way home. You ate standing up, though there was a chair.
Try making a short list of things you remember from yesterday. Keep it factual for now. Then choose the detail you feel an urge to explain.
My father called while I was making dinner. I let it ring. Later, I listened to his voicemail twice.
This is an invented example, but notice what it gives a writer to work with. Why let it ring? What was in the message? Why listen twice? You don't need answers to all of those questions before you start. One of them may be enough to carry the next paragraph.
If you're writing for yourself, you can stay with what happened and how you remember it. If you want to try fiction, change a detail and follow the consequence. Suppose the caller leaves a message intended for someone else. Now you have a different question to write toward.
Make the first attempt smaller than the ambition
A beginner once asked Ursula K. Le Guin how to start a novel when the possible characters and plots felt paralyzing. In her Navigating the Ocean of Story workshop, she suggested practicing with a paragraph or a page: describe a scene, or put a character in a situation and see what they do and feel. The novel could wait.
You can borrow that scale even if you never intend to write fiction. Instead of writing about your relationship with your father, write about the few minutes around the phone call. Where was the phone? What were you doing with your hands? What did you do when it stopped ringing? The larger subject will still be there.
Specific details give you choices that a broad topic doesn't. A pan beginning to burn might explain why someone missed a call. Turning down the stove and watching the phone ring would suggest something else. The same event changes as you pay closer attention to it.
For a first attempt, a paragraph is a reasonable amount to ask of yourself. Let it be incomplete. You can decide later whether it belongs in a story, an essay, or nowhere beyond today's page.
Write far enough to find out what you're saying
There is a point, often quite early, when a draft starts to disappoint you. The words repeat themselves. A sentence sounds borrowed. You can see how much better it ought to be, but correcting the first line hasn't helped you reach the second.
The University of North Carolina Writing Center's guide to brainstorming describes freewriting as a way to separate generating material from judging it. For a limited stretch, you keep writing without correcting style or grammar. Afterwards, you read back for something worth developing. It's a practical exercise, rather than a promise that everything you write will be useful.
If that sounds worth trying, give the phone-call scene five minutes. When you get stuck, write the question you're stuck on. You might end up with something like this:
I don't know why I listened twice. There wasn't any news in it. He wanted to know whether the rain had reached us. I think I wanted to hear the part where he said my name.
That last sentence gives the writer somewhere to go. It also changes how the earlier details might be read. You would have had a harder time reaching it if every sentence needed to be presentable before the next one could exist.
When the five minutes are up, read what you have. Find a line you'd like to understand better and spend another paragraph on it. If nothing interests you, leave the attempt alone and try a different detail tomorrow. This is a way to discover material, so some attempts will give you very little.
Le Guin makes a related point in the same workshop when a writer asks how to judge a story's first page. She argues that you need the rough whole before you can know whether the opening works. Those early pages may have helped you find the story even if you eventually remove them.
Freewriting won't suit every writer or every day. If continuous writing makes you feel rushed, use a list. If you already know what you want to say, sketch an outline. Give yourself a method that helps you continue, and change it when it doesn't.
Leave yourself a way back
Starting to write again brings a particular temptation: explaining the whole absence before you're allowed to write about today. An unfinished notebook can begin to feel like correspondence you've neglected.
You can turn to a fresh page and begin with whatever is happening now. There is no missing chapter you have to supply. If an older piece still interests you, read the last paragraph and write a note about what you want to explore next.
The same small courtesy helps at the end of a session. Before closing the page, leave yourself a question. For the scene above, it might be: What happened when I finally called back? Next time, you have a place to begin without rereading everything or inventing a new subject.
Eventually you may want to revise a piece for a reader. That's a different kind of attention: deciding what needs explaining, what can go, where the piece actually begins. For today, you can leave the voicemail on the page and follow the thought it brings up. You don't have to know yet how the piece ends.