Beat Taxonomies from Craft Literature¶

Craft guidance for understanding the named beat types that appear across major storytelling frameworks —Save the Cat, Story Grid, the Hero's Journey, and others —and how their function labels serve as diagnostic and generative tools for interactive fiction.

What Is a Beat Taxonomy?¶

Three Levels of "Beat"¶

The word "beat" means different things depending on who uses it. Craft literature operates at three distinct levels, and confusing them causes real problems in planning and discussion.

Level	What It Describes	Example	Covered In
Prose beat	Smallest unit of prose —one action, line of dialogue, or sensory detail	"She set down the glass."	Pacing and Tension
Structural beat	Named story event with a specific dramatic function	Catalyst, Midpoint, All Is Lost	This article
Emotional beat	A moment designed to produce a specific feeling in the reader	Catharsis, dread, triumph	Emotional Beats

These are not competing definitions. They operate at different scales. A single moment in a story is simultaneously all three: a prose beat (two sentences of dialogue), a structural beat (the Catalyst), and an emotional beat (shock that reframes everything). Recognizing which level you are discussing prevents the most common planning confusion —arguing about "beats" when two people mean different things.

Example of all three levels at once: A character opens a letter revealing her father is alive. At the prose level, this is an action beat followed by an emotion beat. At the structural level, this is the Catalyst —it disrupts the status quo and launches the story. At the emotional level, this is a shock-and-wonder beat designed to produce astonishment in the reader. All three descriptions are correct. They describe different aspects of the same moment.

Why Function Names Matter¶

Consider two descriptions of the same story moment:

"The hero learns the village is threatened" —this is content
"Catalyst" —this is function

Content tells you what happens. Function tells you what the moment does for the story. A Catalyst disrupts the status quo and forces the protagonist into a new situation. That function is the same whether the content involves a threatened village, a mysterious letter, or a job offer in another city.

Function names are diagnostic tools. When a story feels flat at the midpoint, the diagnosis "your Midpoint lacks a stakes reversal" is more actionable than "the middle is boring." The name carries the fix.

Function names are also communication tools. When collaborators —human or AI —share a vocabulary of beat functions, structural discussions become precise. "Move the Catalyst earlier" is unambiguous. "Make the beginning more exciting" is not. A shared taxonomy reduces misunderstanding and wasted iteration.

For LLM-assisted generation, function names are even more valuable. Labeling a passage as "Dark Night of the Soul" tells the model what emotional register to hit, what narrative purpose the passage serves, and how it relates to surrounding passages. Without that label, models tend toward neutral, even-keeled prose that lacks structural purpose.

Save the Cat Beat Sheet¶

Blake Snyder developed his 15-beat framework for screenwriting in Save the Cat! (2005). Jessica Brody adapted it for novels in Save the Cat! Writes a Novel (2018). The framework remains one of the most widely used structural tools because it names both the function and the approximate position of each beat.

The 15 Beats¶

Beat Name	Position	Function	What It Accomplishes
Opening Image	1%	Snapshot of "before"	Establishes tone, mood, and protagonist's starting state
Theme Stated	5%	Thematic question posed	Someone states the theme (often to the protagonist, who doesn't yet understand it)
Setup	1-10%	World and character establishment	Shows protagonist's flawed world, introduces key characters, plants seeds
Catalyst	10-12%	Disruption of status quo	An event that makes the old world unsustainable —the story's inciting incident
Debate	12-25%	Protagonist resists the call	Internal or external deliberation about whether to act —raises stakes of saying yes
Break into Two	25%	Commitment to action	Protagonist makes an active choice to enter the new world or situation
B-Story	22-30%	Secondary relationship introduced	Often a love interest or mentor —carries the thematic argument
Fun and Games	25-50%	Promise of the premise	The "trailer moments" —what the audience came for. The concept in action
Midpoint	50%	Stakes shift	Either a false victory (things seem great but aren't) or a false defeat. Raises stakes permanently
Bad Guys Close In	50-75%	Increasing pressure	External forces tighten, internal flaws resurface, the team fractures
All Is Lost	75%	Lowest point	The protagonist's plan fails. Often involves a "whiff of death" —something or someone is lost
Dark Night of the Soul	75-80%	Emotional processing	Protagonist confronts despair, reflects on failure, reaches for deeper truth
Break into Three	80%	Synthesis and new plan	A-Story and B-Story merge. Protagonist finds a new approach using lessons learned
Finale	80-99%	Final confrontation and resolution	Protagonist enacts the new plan, defeats opposition, proves transformation
Final Image	99-100%	Snapshot of "after"	Mirrors Opening Image to show how far the protagonist has come

Beat Functions vs Beat Positions¶

The beat sheet prescribes both function and position. Snyder was precise: Catalyst at 12%, Midpoint at 50%, All Is Lost at 75%. These percentages come from screenplay structure, where a 110-page script has rigid timing expectations.

For interactive fiction, positions are guidelines, not rules. A branch might reach its Catalyst at 5% or 20%. What matters is that the function is present and correctly ordered. A story where All Is Lost comes before Midpoint has a structural problem regardless of position.

The ordering constraint is more important than the positional constraint:

Setup before Catalyst
Catalyst before Break into Two
Fun and Games before Midpoint
Bad Guys Close In before All Is Lost
Dark Night before Finale

Practical implication for IF: When scaffolding a branching story, check that every path through the graph respects this ordering. A branch that jumps from Catalyst directly to Finale skips the middle —it may be short, but it still needs escalation and a low point to feel complete.

Adaptation from Screenwriting to Prose¶

Snyder wrote for 110-page screenplays. Jessica Brody's adaptation for novels identified key differences:

Longer Setup. Novels have room for deeper character establishment. The Setup expands from a few pages to multiple chapters, allowing subplots, world-building, and character relationships that film compresses.

More complex B-Stories. Film B-Stories are often a single relationship. Novel B-Stories can involve multiple secondary characters and subplots, each carrying a facet of the thematic argument.

Expanded Dark Night of the Soul. In screenwriting, this might be a single scene. In novels, it can be an entire chapter of reflection, flashback, and emotional reckoning. Brody argues this is where the novel form most surpasses film —internal experience rendered in full.

Fun and Games as a larger territory. The promise of the premise occupies a quarter of the story. In novels, this section builds the world and relationships that make the second half's losses devastating.

Interactive fiction adds a further adaptation. The Debate beat —where the protagonist resists the call —often becomes the first major choice point. The reader decides whether to accept the call, and that decision creates the first meaningful branch. Similarly, Break into Two is frequently the mechanism through which the story's central branching structure activates: the reader's commitment to a path IS the break into Act II.

The B-Story also transforms in IF. Rather than a single secondary relationship carrying the thematic argument, interactive fiction can offer different B-Story characters on different branches —each embodying a different facet of the theme. A branch where the reader allies with a mentor explores the theme through wisdom and tradition. A branch where the reader allies with a rival explores it through competition and self-reliance. The B-Story becomes branch-variable content, and its variation is one of the primary ways branches feel meaningfully different.

Story Grid Beat Hierarchy¶

Shawn Coyne's Story Grid methodology (The Story Grid, 2015) takes a different approach. Rather than prescribing a fixed sequence of named beats, Coyne defines a hierarchy of structural units, each containing the ones below it.

The Hierarchy: Beat, Scene, Sequence, Act, Story¶

Level	Contains	Turns On	Example
Beat	Single action or exchange	Behavior shift	Character's expression changes from confident to uncertain
Scene	Multiple beats	Value shift (positive to negative or reverse)	"Safe" turns to "in danger"
Sequence	2-5 scenes	Larger value shift compounding scene turns	A series of discoveries escalates "curious" to "terrified"
Act	Multiple sequences	Major value reversal	Act I ends with the protagonist's world upended
Story	All acts	Global value transformation	Character moves from "cowardly" to "courageous" (or the reverse)

The key insight is that every level turns on a value shift. A beat without a shift is dead weight. A scene without a value change is a candidate for cutting. This fractal property —the same structure at every scale —makes Story Grid a powerful analytical tool.

For interactive fiction, this hierarchy maps naturally to the generation pipeline. The global story defines the overarching value transformation. Acts correspond to major story phases. Sequences map to clusters of related passages. Individual passages are scenes. And the prose beats within each passage are the smallest units. Each level can be planned, generated, and validated independently using the same value-shift criterion.

Story Beats vs Emotional Beats¶

Coyne distinguishes between external (story) beats and internal (emotional) beats:

External beats: Goal pursuit, obstacle encountered, crisis point, action taken. These move the plot.

Internal beats: Emotional reaction, subtext processing, value questioning, worldview shift. These move the character.

Coyne's central insight: every scene needs both or it fails. A scene with only external beats feels mechanical —things happen but nobody processes them. A scene with only internal beats feels static —characters feel things but nothing happens.

This maps directly to Swain's Scene-Sequel model:

Coyne's Term	Swain's Equivalent	Function
External beats	Scene (Goal-Conflict-Disaster)	Move the plot forward through action
Internal beats	Sequel (Reaction-Dilemma-Decision)	Move the character forward through processing

The vocabulary differs, but the structural requirement is identical: action followed by processing, then new action. A story that neglects either half —all plot with no reflection, or all feeling with no events —fails at the scene level regardless of how strong its macro structure may be.

The Five Commandments of Storytelling¶

Coyne identifies five elements required in every scene:

Inciting Incident —An event that upsets the balance and launches the scene
Turning Point Progressive Complication —A complication that forces a crisis (not just any complication —the one that makes the situation impossible to ignore)
Crisis —A dilemma requiring choice, typically "best bad choice" or "irreconcilable goods"
Climax —The character's action in response to the crisis
Resolution —The new state after the action, setting up the next scene

These five commandments apply at every level of the hierarchy. An act has its own inciting incident, turning point, crisis, climax, and resolution. So does the global story. This fractal structure means the same diagnostic vocabulary works at any scale: "Your scene lacks a clear crisis" uses the same framework as "Your story lacks a clear crisis."

The crisis question is particularly valuable for interactive fiction. Coyne defines two types of crisis: the best bad choice (all options have costs) and irreconcilable goods (two desirable outcomes that cannot both be achieved). These map directly to meaningful choice design —the best IF choices are exactly these crisis types. A choice where one option is obviously correct is not a crisis and will not engage the reader.

Comparative Taxonomy¶

Mapping Across Frameworks¶

Different frameworks emphasize different aspects of the same underlying story movements. This table maps approximate equivalences across five major systems:

Story Position	Swain	Save the Cat	Story Grid	Hero's Journey	Three-Act
Opening state	Setup	Opening Image + Setup	Beginning Hook (setup)	Ordinary World	Act I opening
Inciting event	Goal introduced	Catalyst	Inciting Incident (global)	Call to Adventure	Act I turning point
Commitment	Decision	Break into Two	End of Beginning Hook	Crossing the Threshold	Act II begins
Midpoint shift	Mid-scene disaster	Midpoint	Middle Build turning point	Ordeal / Innermost Cave	Act II midpoint
Low point	Disaster (major)	All Is Lost	All Is Lost moment	Death and Rebirth	Act II turning point
Climax	New goal enacted	Finale	Climax (global)	Return with Elixir	Act III climax

These mappings are approximate. The Hero's Journey's "Ordeal" is not exactly the Save the Cat "Midpoint" —Campbell emphasizes mythological death-and-rebirth where Snyder emphasizes a false victory or false defeat. But they occupy the same structural territory and serve related dramatic functions.

Notice what the table reveals: every framework has a name for the inciting disruption and the climactic resolution. The frameworks diverge most in how they describe the middle —the long stretch between commitment and crisis. Save the Cat fills this with "Fun and Games" and "Bad Guys Close In." Story Grid fills it with progressive complications. The Hero's Journey fills it with "Tests, Allies, and Enemies." These are different lenses on the same structural territory, and each highlights something the others miss.

Where the Frameworks Disagree¶

Scale of focus. Swain operates at the micro level —scene and sequel, beat by beat. Save the Cat operates at the macro level —whole-story architecture. Story Grid bridges both, defining fractal structure from beat to story. No single framework covers all scales equally well.

What drives structure. Save the Cat is prescriptive: these beats, in this order, at these positions. Story Grid is diagnostic: these elements must be present, but the framework adapts to genre conventions. The Hero's Journey is mythological: structure reflects universal human experience rather than dramatic engineering.

The role of the protagonist. Save the Cat assumes a single protagonist with a clear arc. Story Grid accommodates ensemble casts through multiple value shifts. The Hero's Journey assumes a mythic hero pattern that not all stories follow.

Genre sensitivity. Story Grid is explicitly genre-aware —a thriller has different obligatory scenes than a romance. Save the Cat is more genre-neutral, applying the same 15 beats regardless. The Hero's Journey applies best to adventure, fantasy, and coming-of-age; it strains when applied to intimate domestic stories.

Treatment of the middle. This is the most revealing disagreement. Save the Cat splits the middle into distinct phases (Fun and Games, then Bad Guys Close In) with a clear Midpoint dividing them. Story Grid treats the middle as a series of progressive complications building toward a crisis. The Hero's Journey sees it as a sequence of trials. Syd Field originally called it "the problem with Act II" —the middle is where stories sag, and each framework's attempt to structure it reveals its priorities. For interactive fiction, the Save the Cat vocabulary is most immediately useful because it names the emotional trajectory (playful exploration giving way to tightening pressure) rather than just the mechanical structure.

No single framework is complete. The richest vocabulary for structural analysis comes from combining them: Save the Cat for macro architecture, Story Grid for scene-level diagnosis, Swain for prose-level beat construction, and the Hero's Journey for mythological resonance.

Practical recommendation: Use Save the Cat beat names when planning overall story shape. Switch to Story Grid vocabulary when diagnosing individual scenes. Use Swain's Scene-Sequel language when writing or reviewing prose. This layered approach gives you precision at every scale without forcing one framework to cover territory it was not designed for.

Interactive Fiction: Beat Functions in Branching Stories¶

Mandatory vs Branch-Variable Beats¶

In a linear story, every reader encounters every beat. In branching fiction, some beats must appear on all paths while others can be branch-specific.

Beat	Mandatory / Variable	Why
Catalyst	Mandatory	Without it, no path has a reason to begin. Every branch needs its inciting disruption
Break into Two	Mandatory	The choice point that launches the branch IS the break. Often the branching mechanism itself
Midpoint equivalent	Mandatory	Every path needs a stakes shift or it sags in the middle
All Is Lost equivalent	Variable per branch	Different paths can have different low points —this is where branches feel most distinct
Dark Night of the Soul	Variable per branch	The nature of reflection depends on what was lost, which differs by path
Climax equivalent	Mandatory	Every path must resolve. A branch without a climax is an abandoned thread
Final Image	Mandatory	Every ending needs closure, though the image differs by path

The principle: every path through the story should contain a complete emotional arc. A reader who takes any single route should experience disruption, escalation, low point, and resolution. Paths that skip the low point feel hollow. Paths that skip the climax feel abandoned.

Note that "mandatory" does not mean "identical across branches." The Catalyst can be shared (all branches begin from the same disruption) or branch-specific (each path has its own triggering event). What matters is presence, not identity.

The Branch as Beat Sheet¶

Each major branch can be analyzed as its own mini beat sheet. This does not mean every branch needs all 15 Save the Cat beats —that would be impractical. But every branch needs a minimum set:

Catalyst —why this path exists, what launched the reader down it
Fun and Games —the content unique to this branch, the reason to take this path
Crisis / All Is Lost —the hardest moment on this path, where things feel most desperate
Resolution —how this path concludes, what transformation it reveals

Branches that lack a crisis or low point feel like tourism —pleasant but dramatically inert. The reader visits interesting locations without ever being challenged. Branches that lack unique Fun and Games content feel like cosmetic variations of the same story.

Test each branch independently. Read through a single path from start to finish. Does it feel like a complete story? Does tension rise and fall? Is there a moment of genuine difficulty? If a branch reads as flat, check which beat functions are missing.

Example of a hollow branch: The reader chooses to investigate the abandoned mine. The branch describes the mine's history, shows some interesting geological features, introduces a minor character, and concludes with the reader emerging safely. Nothing went wrong. No stakes were raised. No crisis was faced. The branch has Fun and Games content but no All Is Lost, no Crisis, no transformation. It reads as a side tour, not a story path.

The same branch with beat functions: The reader enters the mine (Catalyst —the entrance collapses behind them). They discover evidence of an old crime (Fun and Games —unique branch content). Their light source fails and they hear movement deeper in the tunnels (Bad Guys Close In). They realize the minor character they trusted is the one who caused the collapse (All Is Lost). They must decide whether to confront the character or find another way out (Crisis). The branch now has structural shape.

Beat Naming for LLM Generation¶

Labeling passages with beat function names produces measurably better LLM output. The function name acts as a compressed instruction set, conveying tone, purpose, and structural role in a few words. Without explicit labels, LLMs default to a neutral middle register —competent but structurally flat prose that fails to differentiate a Catalyst from a Fun and Games passage.

Beat Function Label	What to Include in Prompt	Common LLM Failure Without It
Opening Image	Sensory grounding, character state, tone establishment	Generic scene-setting with no emotional baseline
Catalyst	Disruption, urgency, status quo breaking	Events happen but nothing feels different afterward
Fun and Games	Concept in action, unique content, promise delivery	Flat exposition that fails to deliver the premise
Midpoint	Stakes reversal, false victory or false defeat	Middle sections that feel like more of the same
Bad Guys Close In	Mounting pressure, internal fractures, closing walls	Repetitive obstacles without escalation
All Is Lost	Lowest point, loss, despair	Moderate difficulty that never reaches genuine darkness
Dark Night of the Soul	Reflection, emotional processing, deeper truth	Skipped entirely —characters bounce from failure to action
Finale	Transformation applied, new approach, decisive action	Endings that feel unearned or disconnected from the journey

Example prompt with beat function: "Generate passage P12. This is a 'Bad Guys Close In' beat. The protagonist's earlier lies are catching up with her. Her ally has grown suspicious, the evidence she hid is about to surface, and she's running out of time. The tone should be increasingly claustrophobic. End with a moment that makes the next beat (All Is Lost) feel inevitable."

Without the beat function label, models tend toward monotonous pacing —every passage at the same emotional intensity, every scene at the same stakes level. The label breaks that pattern by specifying where this passage sits in the emotional arc.

Smaller models benefit most. Large frontier models can infer structural function from context, though they still benefit from explicit labels. Smaller local models (4B-8B parameters) struggle significantly with implicit structure. For these models, beat function labels are not optional —they are the primary mechanism for producing prose with appropriate emotional variation.

Combining beat labels with value shifts amplifies the effect. A prompt that says "This is an All Is Lost beat; the value shifts from 'hopeful' to 'despairing'" gives the model both structural function and emotional direction. This dual specification —what the beat does and how it should feel —produces the most structurally coherent output across model sizes.

Scaffolding with Beat Functions¶

During initial scaffolding, assign beat functions to the story skeleton before writing any prose. This creates a structural map that serves as both generation guide and validation tool.

The process:

Outline the story's passages and branching structure
Assign a beat function to each passage (Catalyst, Fun and Games, Midpoint, etc.)
Walk each branch path and verify it contains the minimum function set
Identify gaps —if no passage on a branch is labeled "All Is Lost," that branch likely lacks a genuine low point
Adjust structure before writing prose —adding, moving, or reframing passages to fill structural gaps

Beat function map as validation. After scaffolding, check:

Does every branch have a Catalyst, Midpoint equivalent, and Climax?
Is there at least one All Is Lost or Crisis beat per branch?
Do Fun and Games beats deliver content unique to their branch?
Does the ordering make structural sense (no Finale before Midpoint)?

This validation catches structural problems early, before prose has been written and rewriting becomes expensive. A missing function label is far easier to fix than a missing chapter.

Example beat function map for a three-branch story:

Passage	Branch A (Confront)	Branch B (Investigate)	Branch C (Flee)
P1-P3	Setup / Catalyst (shared)	Setup / Catalyst (shared)	Setup / Catalyst (shared)
P4	Fun and Games	Fun and Games	Fun and Games
P5	Bad Guys Close In	Midpoint	Bad Guys Close In
P6	Midpoint	Bad Guys Close In	All Is Lost
P7	All Is Lost	All Is Lost	Dark Night of the Soul
P8	Dark Night of the Soul	Dark Night of the Soul	Finale
P9	Finale	Finale	Final Image
P10	Final Image	Final Image	—

Reading down each column, you can verify that every branch contains the minimum function set. If Branch C lacked an All Is Lost passage, the map would reveal that gap immediately. The map also reveals pacing differences: Branch C reaches its low point earlier (P6) and resolves faster, creating a shorter, more intense experience. Branch B delays its Midpoint to P5 but front-loads investigation content. These are design choices, visible and adjustable before any prose is written.

Quick Reference¶

Framework	Key Contribution	Best Used For
Save the Cat	15 named beats with positions and functions	Macro story architecture, planning overall shape
Story Grid	Fractal hierarchy with Five Commandments per level	Scene-level diagnosis, genre-aware analysis
Swain (Scene-Sequel)	Micro-level beat construction and pacing	Prose-level rhythm, action-reaction patterns
Hero's Journey	Mythological resonance and archetypal structure	Fantasy, adventure, coming-of-age stories

Beat Function	What It Does	IF Application
Catalyst	Disrupts status quo, launches the story	Must appear on every branch —often IS the branch point
Midpoint	Shifts stakes via false victory or false defeat	Prevents middle sag; every path needs a stakes reversal
All Is Lost	Delivers the lowest emotional point	Branch-variable —different paths have different darkest moments
Dark Night of the Soul	Processes loss, reaches for deeper truth	Where character voice emerges most strongly in prose
Finale	Applies transformation to resolve the story	Each branch ending must feel earned through its own arc

Research Basis¶

The taxonomies in this article draw from established craft literature spanning screenwriting, novel writing, and story analysis:

Concept	Source
15-beat story structure for screenwriting	Blake Snyder, Save the Cat! (2005)
Beat sheet adaptation for novels	Jessica Brody, Save the Cat! Writes a Novel (2018)
Story Grid hierarchy and Five Commandments	Shawn Coyne, The Story Grid (2015)
Scene-Sequel structure and prose beats	Dwight V. Swain, Techniques of the Selling Writer (1965)
The Hero's Journey (monomyth)	Joseph Campbell, The Hero with a Thousand Faces (1949); Christopher Vogler, The Writer's Journey (1992)
Three-act structure and paradigm	Syd Field, Screenplay: The Foundations of Screenwriting (1979)
Story structure and value shifts	Robert McKee, Story: Substance, Structure, Style, and the Principles of Screenwriting (1997)

Snyder's beat sheet and Coyne's Story Grid represent different philosophies —prescriptive vs diagnostic —but both provide named vocabulary for structural functions. Swain's Scene-Sequel model operates at a different scale entirely, addressing the prose-level rhythm that fills these larger structural containers. The combination of all three scales —macro (Save the Cat), meso (Story Grid), and micro (Swain) —provides the most complete structural vocabulary available.

McKee's contribution is the emphasis on value shifts as the engine of story: every scene must turn on a value (safe/endangered, loved/unloved, free/trapped), and the story's global arc is a sequence of these turns. This value-based thinking complements the beat-naming approach —once you know a beat's function (Catalyst) and its value shift (safe to endangered), you have both the structural role and the emotional content defined before writing a word of prose.