Some of my favorite writing happens at the edges of the day.

The world gets quieter. Fewer notifications arrive. Nobody expects an immediate response. There is something peaceful about working while the rest of the world appears to have collectively agreed to stop talking for a few hours. In a strange way, writing early in the morning and writing late at night share a lot in common. Both happen before the demands start or after they’ve ended. Both offer a rare feeling of uninterrupted attention.

But there is one crucial difference: Midnight Ember and Morning Ember are two completely different people.

I trust Morning Ember. Morning Ember drinks water (albeit after a comically large cup of coffee), reads instructions before beginning assignments, and generally makes decisions supported by evidence.

Midnight Ember, meanwhile, has a very different set of strengths.

Her primary qualification is that she is still awake. Beyond that, she has a laptop, feels strongly about things, and specializes in perseverance under the most dire of circumstances. She can continue working long after success and stubbornness become indistinguishable.

I appreciate Midnight Ember’s effort. I admire her commitment. But I absolutely do not trust her unsupervised.

What makes Midnight Ember particularly dangerous is that she has never successfully completed a project in her life, yet she remains extraordinarily confident in her ability to manage them. Morning Ember always has to deal with the consequences. Their relationship has deteriorated considerably over the years.

I’ve come to suspect that Morning Ember and Midnight Ember are merely the most visible expression of a much larger disagreement; they are two competing instincts that accompany every writing project from beginning to end.

Writing creates a strange conflict of interest, as the writer and editor often occupy the same body. The problem is that both jobs demand opposite things. One requires commitment; the other requires skepticism.

Unfortunately, after spending three hours wrestling with a paragraph, it becomes difficult to tell whether you’ve improved it or just grown fond of it. Somewhere around the twentieth reread, every sentence begins to earn the benefit of the doubt. And when you’ve spent three hours fighting for a paragraph, of course you think it’s important. You’ve probably invested more time in it than anything else that day.

The opposite happens too. A writer can become just as invested in cutting a paragraph as keeping one. More than once, Morning Ember has opened a document to discover that Midnight Ember deleted a paragraph for being “redundant,” only for Morning Ember to spend twenty minutes writing the exact same paragraph again.

The thing is, Midnight Ember isn’t wrong about everything.

Some of my favorite ideas arrive after midnight. Some of my best drafts begin there. Midnight Ember is willing to chase ideas that Morning Ember would dismiss immediately. She is reckless, but creativity occasionally benefits from recklessness.

The problem is when either insists on being both writer and editor.

Those are different jobs.

Creating something requires a willingness to be wrong. You have to follow bad ideas, write clumsy sentences, and occasionally convince yourself that the history of human attention can be explained entirely through a TikTok of someone chugging a BuzzBall (which, if my thesis advisor is reading this, was less ridiculous in context).

Editing requires the opposite mindset. It demands distance, restraint, and the willingness to delete things you spent an hour writing—but only when distance and restraint suggest you should.

This is one of the biggest lessons I have learned from editing other people’s work. As an editor, I encounter the finished product rather than the struggle that created it. I don’t know which sentence took five minutes and which one took five hours. I don’t know which paragraph consumed an entire evening or which idea survived six drafts. All I see is the page.

That separation is incredibly useful.

The more editing I’ve done, the more I’ve realized that editing your own work often relies on pretending, however briefly, that the work belongs to someone else. That’s what Morning Ember accidentally provides.

(This may explain why every writer eventually tries to trick someone else into reading their work. Not because they are incapable of editing it themselves, but because another person possesses a superpower they don’t: They haven’t seen it before.)

The writer and editor may share a body, but they should not share a shift.

I usually do my best writing when I’m not simultaneously trying to evaluate it. That’s easier said than done, of course. The editor is an enthusiastic backseat driver. Halfway through a draft, she’ll start pointing out repetitive phrasing, awkward transitions, and inconsistencies in curly and smart quotation marks. Gee, thanks.

These observations aren’t unhelpful, they’re just not the emergency she believes them to be. A first draft is a poor environment for judgment because there is rarely enough context for the judgment to be useful.

More than once, I’ve returned to a document convinced I would need to rewrite an entire section, only to discover that it was the strongest in my paper. Other times, a paragraph that seemed pointless while I was writing it turned out to be the missing piece that connected everything else. Many writing problems become easier to solve once there is more writing around them. You can’t always tell what belongs until you’ve finished building the thing it’s supposed to belong to.

And many of the paragraphs I eventually cut are the same paragraphs that allowed me to discover the ones I keep.

So in a way, I appreciate the conflict between the writer and editor. For years, I thought the goal was to make them agree—to become the kind of writer who always knew exactly what to say and exactly how to say it.

Now I’m not so sure. In fact, I suspect the conflict is a good sign. The moment a writer becomes completely satisfied with every sentence they produce is probably the moment they stop improving. The editor exists because the writer is imperfect. The writer exists because the editor is a perfectionist, and someone has to be willing to write the bad first draft.

Rather than trying to make Morning Ember and Midnight Ember agree, I settle for keeping them from talking over each other.