The Flow Problem: How to Make Your Writing Connect
Maybe you played this game when you were younger.
You sit in a circle with a piece of paper. You write the opening line of a story—something about spaceships, or a magical monkey. Then you pass it to the person next to you.
They read your line. They write the next line. Then they fold the paper so only their line is visible and pass it on.
By the time the paper comes back around, the spaceship captain has become a bumblebee who is also your math teacher, and everyone is fighting a sentient cloud named Steve. The story is ridiculous. That's the point.
But here's what's interesting: each individual connection works. The person who added the bumblebee was responding to whatever came before. Everyone followed the thread. The thread just kept changing color.
There's a name for what each writer achieved: cohesion. The sentence-to-sentence connection, the local "flow" from one idea to the next.
And there's a name for what the group failed to achieve: coherence. The sense that all the parts add up to something. The feeling that someone is steering the ship.
The game is fun because it mechanically separates these two things. Each player has cohesion. No one has coherence. The laughter comes from the gap.
Two Different Problems
When someone tells you your writing "doesn't flow" or feels "choppy," that feedback isn’t very helpful. "Flow" is a vibe. It doesn't distinguish between two very different issues.
It might be cohesion: your sentences don't connect to each other. Each one starts fresh, asking your reader to reorient.
Or it might be coherence: your sentences connect fine, but they're not accumulating toward anything.
These problems feel similar—both produce that "something's off" sensation—but they require completely different revisions.
How Information Wants to Move
We prefer to move from what we already know to what we're learning: begin with familiar information, then end with new. Cohesion is the sentence-level version of this idea.
Read these two sentences as a pair:
Haidt's argument depends on large-scale survey data about teen mental health. Neurodivergent children often have very different relationships to screens than their neurotypical peers, so their experience disappears in survey data.
Now this version:
Haidt's argument depends on large-scale survey data about teen mental health. Such data tends to reflect majority experiences—which means neurodivergent children, who often have very different relationships to screens, can disappear from the findings.
The second pair flows. The first doesn't. Why?
Look at where the reference to “data” appears in each example’s second sentence. The second example begins with "Such data"—picking up exactly where we left off. The old information reappears at the start, and new information extends from it.
In the choppier version, the second sentence starts cold: “Neurodivergent children.” That's new. Your reader has to hold the previous sentence in mind while processing something unrelated, and is then asked to work backward to see the connection.
Notice that the smoother version starts with the “known” and builds somewhere new. By linking the sentences, we develop an idea—that the data itself has a built-in blind spot. It allows readers to build their understanding methodically.
This doesn’t mean all sentence-to-sentence connections need this exact type of cohesion—starting the next sentence with information from the previous and building on it—but it’s an important principle to recognize. The easiest way to follow an idea is through this kind of cohesive “build,” where A leads to B, then B leads to C.
Steering the Ship
You can see the trap. It's possible to have perfect cohesion and still produce incoherent writing. All of a sudden you’re at Z and there’s no connection to A. That's what makes the children's game funny.
Coherence provides direction. Coherent writing returns to its central ideas, reminding readers why they're reading. Coherence isn't just something you build into a text; it's how you help readers ground themselves in an idea.
Writers often fear this repetition. They worry that returning to their main idea will bore their reader or feel condescending. So they keep introducing new topics, new angles. The main idea drifts away. This leaves the reader feeling less like a fellow explorer through an idea, and more like a blindfolded follower, left to guess at connections.
The five-paragraph essay is a good example of this lack of coherence. It often acts as a listicle—three random examples of a thing. The sentences might have cohesion within each paragraph, but without a central inquiry that evolves, the paragraphs don't build. It’s a list of semi-related ideas, slightly more coherent than the children’s game, but still lacking a driver.
The Problem with "Add Transitions"
The intuitive solution to choppy writing is to add transitions. Throw in some "howevers" and "therefores." But studies of college writing find something counterintuitive: heavy use of conjunctions often correlates with lower quality writing. Expert writers depend less on explicit markers and more on the actual arrangement of information.
Phrases like "It is important to note that" or "This brings us to the consideration of" signal transition without making one. They create the appearance of connection without the substance (often called “fake cohesion”).
This happens a lot in student rough drafts. For instance, a student writes: However, it is important to note that in arriving at such a conclusion, we must recognize that Haidt's argument focuses primarily on neurotypical children.
There's a real idea buried in there, clearer after some revision: Haidt's data comes almost entirely from studies of neurotypical children, raising questions about whose experiences his recommendations actually address.
The first version announces you're about to say something, hedges, then says a fragment of it. The second says the thing, then leaves room for analysis.
Genre Changes Everything
Different genres have radically different relationships to flow. What counts as "flowing" in a lab report would be deadening in a poem.
Scientific writing treats cohesion as nearly sacred. The conventional structure—Introduction, Methods, Results, Discussion—provides a macro-coherence framework. Within that, your job is to build airtight local connections. A gap in cohesion isn't just confusing—it's suspicious.
Poetry can abandon both cohesion and coherence as traditionally understood. Consider Ezra Pound: "The apparition of these faces in the crowd; / Petals on a wet, black bough." There's no logical connector between faces and petals. But something coheres—through image, rhythm, and interpretation. The reader makes the connection: faces become petals, anonymous and beautiful, glimpsed and gone.
Narrative journalism does something stranger still: it breaks cohesion on purpose. Think about how a magazine feature opens. You're dropped into a scene. A nurse is driving through a wildfire. You don't know why you're watching. You don't know what the story is "about."
Then comes what journalists call the nut graf: the “nutshell paragraph” that zooms out and tells you what you're actually reading. The nurse is part of a story about California's evacuation failures. Journalists at the Philadelphia Inquirer used to describe the nut graf this way: "You may have wondered why we invited you to this party."
That gap is a promise. I've shown you something vivid. Now I'll tell you why it matters.
Years ago I wrote an essay about paddling a tiny inflatable raft down the Bronx River with a friend—twenty-four miles through one of the most polluted urban waterways in America. The lede was comic: two guys arguing about whether to spring for the raft with three air compartments. We pop two of those bladders in the first few miles. For several paragraphs, it's an adventure story. It's funny.
Then the pivot: "To call the Bronx River 'fresh' is a bit misleading. Through the nineteenth and twentieth centuries, it became a natural sewer for industrial waste."
Suddenly we're in a different essay—one about environmental racism, about which communities got factories and runoff, and which ones got parks. Comedy and tragedy were just one paragraph apart. Readers were trusted to understand that I would eventually connect these seemingly disparate tones.
That's not how you'd write a philosophy paper. But it's exactly how narrative journalism handles coherence: through structure, through trust, through a gap the whole piece works to close.
Academic writing is messy to discuss, because it's not one genre but many.
A philosophy paper constructs a chain-link argument. If one link breaks, the logic falls. Every move must connect explicitly to the last. Coherence is essential.
A textual analysis builds a constellation of observations. You might move from one passage to another, letting juxtaposition generate insight before you explain the connection. But there's always the returning build—the moment you show how this new evidence helps deepen the main idea.
Part of becoming a skilled writer is learning to read the genre—to understand what kind of flow the situation demands.
The Real Question
The children's game is funny because no one chose to write an incoherent story. It just happened. The rules of the game guaranteed it.
But you're not playing that game. You can see the whole paper. You decide whether sentences connect tightly or jump. You decide whether to create gaps or close them. Flow isn't magic. It's craft—and the craft changes depending on what you're building.
The question isn't whether your writing flows. The question is: what kind of flow does this piece need? And are you achieving it on purpose, or just passing the paper to the next kid in the circle?
Brice Particelli is a lecturer in UC Berkeley's College Writing Programs. His scholarship on writing pedagogy has been published in Pedagogy and the Journal of Response to Writing, and his essays appear in magazines including Harper's and Guernica.