9 min read

Notes From A Masterclass in Fiction

George Saunders, in his 2021 book A Swim in a Pond in the Rain, guides the reader through seven classic Russian short stories that he teaches in the Syracuse University MFA program.

Saunders is a New York Times bestselling, Booker Prize-winning author. And so in the book, we're learning how fiction works from an expert.

What follows are the interesting and useful ideas I found in Saunders' book, in my words, separated by lines. I took the time to compile them because well, this book is a gold mine.

Of course, these notes are but a fraction of the book's power.

original creation


What makes a reader keep reading?

Actually, better question: If some things better keep a reader reading than others, how would we know?

Well, we're readers. We can track our own minds as they move from line to line.

Saunders asks us to track our minds as we read the stories in the book.

When we read a piece of text, a set of expectations arises.

A man stood on the roof of a seventy-story building.

Aren't we already expecting to him fall, jump, or be pushed?

We'll be pleased if the story takes that into account.

But we'll be displeased if it addresses it too neatly.

A story is simply a series of such expectation/resolution moments.

While reading a story, continually ask:

  1. What do we know?
  2. What are we curious about?
  3. What do we expect to happen?

Whatever you answer, that's what the writer has now to work with.

The writer is like a juggler, throwing bowling pins into the air. The rest of the story is the catching of those pins.

As a writer, the term "structure" is intimidating.

Let's make it less intimidating. Let's imagine structure in fiction simply as a form of call-and-response.

"If we want to make good structure, we just have to be aware of what question we are causing the reader to ask, then answer that question."

For example, in one story, a character finds a nose in a loaf of bread. The reader thinks, where in heck did that nose come from?

The question can also be more subtle. In another story, we are introduced to a peasant driving a cart. We wonder, who is this peasant and why has the author put him here?

The writer should respond to these questions.

When responding to expectations (or using or exploiting or honoring them), the writer should respond not too tightly, but not too loosely either.

Don't respond to expectations in a way that feels too linear. But don't ignore the expectations and take the story in an entirely unrelated direction, either.

Imagine a meter mounted in your forehead with a P on one side (for "Positive") and an N on the other (for "Negative").

Try and read your draft line by line like a first-time reader might. Where's the needle on the P/N meter? Does it drop into the N zone or the P zone? If it drops into the N zone, maybe a fix presents itself.

"I go through the draft like that, marking it up, then go back and enter that round of changes, print it out, read it again, for as long as I feel sharp—usually three or four times in a writing day."

"Over time, like a cruise ship slowly turning, the story will start to alter course via those thousands of incremental adjustments."

"[Trust] that those big thematic decisions are going to be made, naturally, by way of the thousands of accreting micro-decisions at the line level."

"Since we're deciding over and over again, we assume that, eventually, all our decisions will be good ones."

"Given infinite time, anything can happen."

Saunders starts with blocks of loose sloppy text, but as he revises, the blocks get better. At some point, a block will start working and none of it will drop into the N zone.

"The word that sometimes comes to mind [for these blocks] is 'undeniable' as in 'All right, this bit is pretty much undeniable,' which means that I feel that any reasonable reader would like it and would still be with me at the end of it."

Imagine Saunders gives you an apartment in New York City, but it's furnished the way he likes it, not how you like it.

Let's say you can furnish it (at his expense) and make all the changes you want in one day.

Alternatively, let's say you can furnish it, one change at a time, over the course of two years.


"By the end of that two years, that apartment will have had the benefit of the opinions of literally hundreds of manifestations of you... Your intuition will have been given thousands of chances to do its best work."

"That's how I see revision: a chance for the writer's intuition to assert itself over and over."

"My point is that it's not the flavor of your taste that matters; it's the intensity with which you apply your taste that will cause the resulting work of art to feel highly organized."

"The difference between a great writer and a good one (or a good one and a bad one) is the in the quality of the instantaneous decisions she makes as she works. A line pops into her head. She deletes a phrase. She cuts this section. She inverts the order of two words that have been sitting there in her text for months."

When I was a kid I had this Hot Wheels set: lengths of plastic tracks, metal cards, a couple of battery-powered plastic "gas stations." Inside each station was a pair of spinning rubber wheels. The little car went in, then got shot out on the other side. If you arrange the gas stations right, you could urge a little car into one as you left for school and come back hours later to find the car still going around the track.

The reader is the little car. The writer's task is to place gas stations around the track so that the reader will keep reading and make it to the end of the story. What are those gas stations? Well, manifestations of writerly charm, basically. Anything that inclines the reader to keep going. Bursts of honesty, wit, powerful language, humor; a pithy description of a thing in the world that makes us really see it, a swath of dialogue that pulls us through it via its internal rhythm—every sentence is a potential little gas station.

"The writer spends her whole artistic life trying to figure out what gas stations she is uniquely capable of making."

Saunders prefers to refer to plot as "meaningful action".

If a story begins, and introduces a boy who is afraid of water, perhaps a pond or river or ocean will appear. The characterization, and the increased specificity, have created the opportunity for this meaningful event/action.

More generally, as a particular person gets made in a story, through details and characterization, the potential for meaningful action increases.

The writer asks, "Which potential person is this, anyway?" and answers with a series of facts that have the effect of creating a narrowing path: ruling out certain possibilities, urging others forward.

"In specificity lies nascent plot."

"Good writerly habit might consist of continually revising towards specificity, so that specificity can appear and produce [meaningful action]."

Ideally, nothing in a story (especially a short story) exists by chance or to simply serve a documentary function.

Every element should be a little poem, freighted with subtle meaning that is in connection with the story's purpose.

Let's call this Ruthless Efficiency Principle (REP). When we read the elements of the story, we should ask, "What is the purpose of this element?"

For example, when our cart enters a town, we ask, "What is the purpose of this town?"

A question Saunders likes to ask is, "What is the heart of you, dear story?"

All of the other parts of a story are beautiful and necessary to the extent that they serve the heart.

As we read a story, we're dragging along a cart labeled "Things I Couldn't Help Noticing" (TICHN).

As we read, we add aspects of the story that call attention to themselves through some sort of presentational excess. These are the "non-normative" aspects of the story.

When a writer subjects us to a non-normative event... he pays a price: our reading energy drops.


The goal is not to keep the TICHN cart empty and thus writer a "perfectly normal" story. A story that approaches its ending with nothing in its TICHN cart is going to have a hard time ending spectacularly. A good story is one that, having created a pattern of excesses, notices those excesses and converts them into virtues.

When you are reading a story and stumble on something to add to our TICHN cart, ask "How is [that something] going to earn its keep?"

The author hopefully finds a way to convert it into a virtue.

Our evolving rather hard-ass model of a story says that every part of it should be there for a reason. The merely incidental ("this really happened" or "this was pretty cool") won't cut it.

Stuart Cornfeld once told me that in a good screenplay, every structural unit needs to do two things: (I) be entertaining in its own right and (2) advance the story in a non-trivial way.

Aka the Cornfeld Principle.

In their minds, readers naturally juxtapose elements of a story.

A story is not like real life; it's like a table with just a few things on it. The 'meaning' of the table is made by the choice of things and their relation to one another.

For example, in one story, we had three characters, and we can't help but compare and contrast them. Or if there is a description of a ravine or a horse or a pond, we can't help but compare their descriptions to the human happenings of the story.

I've worked with so many wildly talented young writers over the years that I feel qualified to say that there are two things that separate writers that go on to publish from those who don't.

First, a willingness to revise.

Second, the extent to which the writer has learned to make causality.

"For most of us, the problem is not in making things happen ("A dog barked," "The house exploded", "Darren kicked the tire of his car" are all easy enough to type) but in making one thing seem to cause the next.

This is important because causation is what creates the appearance of meaning."

In his youth, Tolstoy published a story called "The Snowstorm". Forty years later, using the same material, he published a much better story called "Master and Man".

One big difference: the events in the first story were happenstance, and the events in the later story were causation.

What transforms an anecdote into a story is escalation. Or, we might say: when escalation is suddenly felt to be occurring, it is a sign that our anecdote is transforming into a story.

You might be familiar with the following:

Freytag's triangle

I sometimes joke that with my students that if they find themselves trapped in exposition, writing pages and pages in which their action doesn't rise, all they need to do is drop this sentence into their story: 'Then something happened that changed everything forever.' The story has no choice but to respond.

In the story "Master and Man", the characters get lost and have to circle back to a town. Every time they do, they pass a clothesline, and they pass that clothesline four times.

Every time, the clothesline looks different. The clothes hang more and more desperately to the line. At one point a shirt has come loose and is hanging by one sleeve. In the last passing, the clothesline entirely disappears.

Escalation is "that which results when we refuse to repeat beats". Each time we pass that clothesline, the laundry has gone some small change in its condition. We read this as an escalation.

"Always be escalating," then, can be understood as "be alert, always, to the possibilities you have created for variation."

I find escalation a bit confusing 😥 Need to practice it.

One effective way of creating an expectation is the enactment of a pattern.

I liked all the stories in the book, but the one I just could not put down was The Darling, by Anton Chekhov. This story followed a simple pattern, where a woman falls in love and that love comes to an end. This pattern recurs three times, with some variation. (I'm reminded of a similar pattern in The Namesake, as well as fairy tales, like the Three Little Pigs or Goldilocks.)

Let me add my own thoughts here. I know humans loves variable rewards. It's why checking your notifications or gambling at a casino is so addicting. Patterns in stories (with an appropriate amount of variation) trigger a similar response.

Saunders creates a pattern in a story of his with a slot in a zoo that is meant to deliver food. On reading this, I thought of slot machines in a casino.

In a story in the book, Chekhov has a chance to respond to our woman protagonist's loneliness with a man, another character in the story. This is the easy solution. But instead, Chekhov waves him away and the man exits the scene!

This is a storytelling move Saunders calls "ritual banality avoidance."

If we deny ourselves the crappo version of our story, a better version (we aspirationally assume) will present itself.

We read a story by Turgenev, in which his excessive descriptions bother us.


The contemporary reader feels [Turgenev's] method of description as old-fashioned. Per our current understanding of fiction, people are to be described selectively—in the sense that not everyone is described and that not everything about them needs to be described. We expect description to be somewhat minimal and serve a thematic purpose, whereas Turgenev seems to be describing things just because they're there.

Many young writers start out with the idea that a story is a place to express their views—to tell the world what they believe. That is, they understand the story as a delivery system for their ideas. I know I felt that way. A story was where I got to set the world straight and achieve glory via the sheer originality of my advanced moral positions.

But as a technical matter, fiction doesn't support polemic very well.


Hope you enjoyed!