The Grammar We Found

24 narrative archetypes, derived from 459 films. And what happens when you map every major story framework onto them.

From shapes to registers

The first post worked with intensity curves: plotting the emotional arc of a film as a time series and clustering those curves to find recurring shapes. That work produced six provisional arc shapes and a lot of open questions about validation and cross-genre stability.

We kept going. And the methodology evolved.

The key shift: we started by asking "what shape does this film make" and added "what structural register is this film operating in at each moment."

This is a meaningful distinction. Two films can have the same overall shape - a deep valley in the middle, two high bookends - but be doing completely different structural things. One might be in a sustained crisis register for that valley. Another might be in a revelation register that looks like a valley on an intensity curve but is doing the opposite structural work. The shape is the same. The grammar is different.

We defined eight structural registers:

Sa (Stability) - the ordinary world. The protagonist's baseline before disruption.

Ri (Incitement) - disruption. Something breaks the ordinary state.

Ga (Pursuit) - directed forward motion. The protagonist actively moving toward a goal.

Ma (Reversal) - structural redirection. A complication that reorients the trajectory.

Pa (Crisis) - sustained pressure. Not a discrete event but a condition the protagonist inhabits.

Dha (Revelation) - epistemic arrival. The protagonist or the audience reaches understanding.

Ni (Climax) - peak confrontation. The final collision with the primary obstacle.

Sa' (Resolution) - new equilibrium. The opening question answered.

These map directly to the swaras of Indian classical music. The analogy is not decorative. In the Melakarta system of Carnatic music, a raaga's identity is defined by which notes dominate, which are secondary, where it opens, where it closes, and which notes are absent entirely. We apply the same logic to narrative: which register dominates screen time (the vadi), which is secondary (the samvadi), where the story opens (the graha), where it closes (the nyasa), and which registers are structurally absent. A lot of these have been institutionalized best practices, where we pay attention to the first 15 mins and last 15 mins of the film - this is now codified.

From these five parameters, applied to 459 films, 24 structural archetypes emerged.

The 24 archetypes

I will not list all 24 here - they are documented at arc.quanten.co/archetype with full structural descriptions, film examples, and disambiguation notes. But let me describe the shape of what we found.

The archetypes cluster into seven register families based on which register is dominant:

Ga-vadi (Pursuit dominant): Sadhana (The Dedicated Path), Vichar (The Open Road), Spanda (The Living Pulse), Vyakula (The Turbulent Pursuit), Gambhira (The Weighted Path), Anubodha (The Guided Pursuit).

Ma-vadi (Reversal dominant): Uttara (The Crossing), Chanchal (The Quick Pivot), Vichitra (The Echo), Griha (The Pull of Home), Viveka (The Discerning Arc), Vedana (The Unhealed Wound), Vikrama (The Pyrrhic Arc).

Pa-vadi (Crisis dominant): Tapasya (The Crucible), Dhairya (The Siege), Sahana (The Graceful Endurance).

Ri-vadi (Incitement dominant): Rahasya (The Hidden Question), Utsaha (The Forward Drive), Pravas (The Exile), Anishchit (The Suspended Note).

Positional (Revelation as structural hinge): Bodha (The Awakening). This archetype required a different classification approach -- more on that below.

Ni-vadi (Climax dominant): Samara (The Living Battle).

Sa-vadi (Stability dominant): Sthira (The Still Life).

Each archetype has a distinct structural identity, a characteristic movement pattern (the pakad), a typical nyasa, and a set of films that instantiate it. Manchester by the Sea is Vichar. Parasite is Vedana. The Batman is Bodha. Moonlight is Spanda. Gone Girl is Uttara. Arrival is Sahana.

These are not genre labels. Genre cuts across archetypes freely. What the archetype captures is the structural grammar of how the story moves - which registers dominate, in what sequence, for how long.

The Bodha problem and what it taught us

One archetype required us to rethink the classification logic entirely.

Bodha (The Awakening) is the archetype of the story where revelation is the structural hinge. The entire second half of the arc is about understanding - not action, not endurance, not pursuit. The protagonist (or the audience) is arriving at comprehension.

The problem: Dha (Revelation) is a traversal register by nature. Stories pass through revelation rather than dwelling in it. A single powerful Dha beat - the moment Neo understands he is The One, the moment Ellie's understanding of time arrives in Arrival - can feel like it defines an entire film even when it represents 10% of screen time. If you classify by which register has the most total screen time, Dha almost never wins. The result was a classification that was technically accurate but structurally dishonest.

We solved it by making Bodha the only positionally-classified archetype in the system. Instead of asking which register dominates by duration, we ask: does the Dha beat arrive in the second half of the film, does it occupy more than 30% of the remaining runtime, and does it convert directly into climax or resolution without returning to another register? If yes, that is Bodha - regardless of what the overall vadi says.

The lesson: some structural patterns are defined by what they do rather than how long they last. The grammar had to be flexible enough to capture that.

Mapping the frameworks

Once we had the archetype system, we did something that has not been done before: we mapped every major published screenplay framework onto the archetype system.

Save the Cat, The Hero's Journey, McKee's Story, Dan Harmon's Story Circle, John Truby's 22 Steps, Michael Hauge's 6-Stage structure, Linda Seger's Turning Points, the Fichtean Curve, Freytag's Pyramid, Kishōtenketsu, Ho-Ha-Kyu, Aristotle's Poetics, the Virgin's Promise, and more.

What we found is both obvious in retrospect and genuinely surprising in its specifics.

The dominant frameworks produce linear arcs, not archetypes.

Save the Cat, The Hero's Journey, Nigel Watts' 8-Point Arc, the Pixar Story Spine, the Bollywood three-act with interval - all of them produce what we call LINEAR_PATH classifications. These frameworks advance through all eight registers in a clean progressive sequence without dwelling in any one. No register dominates. Every register gets its time.

This is not a failure of these frameworks. It is their structural achievement. They were designed to produce complete, balanced arcs that pass through every emotional register. The problem is that when an entire industry is trained on these frameworks for 40 years, the result is a library where 32% of all produced films have no dominant structural register - where the grammar has been averaged out of existence.

A third of our 459-film library classifies as LINEAR_PATH. These are not badly structured films. Many of them are very good films. But they are films where the structural identity has been smoothed into a progressive linear advance that the vadi-based system cannot meaningfully classify.

This is empirical evidence of framework monoculture captured in beat data.

Only six archetypes have corresponding published frameworks.

Of the 24 named archetypes, only six have a major published framework teaching the structural grammar that produces them:

• Tapasya (The Crucible): McKee's Story, the Fichtean Curve
• Vyakula (The Turbulent Pursuit): Vogler's Hero's Journey, Michael Hauge's 6-Stage
• Bodha (The Awakening): Dan Harmon's Story Circle, Truby's 22 Steps
• Chanchal (The Quick Pivot): Seger's Turning Points, Kishōtenketsu
• Dhairya (The Siege): Dan Wells' 7-Point, The Virgin's Promise
• Sadhana (The Dedicated Path): Gulino's Sequence Approach

The remaining 18 archetypes exist abundantly in great films. Vichar (The Open Road) is the most common archetype in our library at 13% of titles. It describes films like Manchester by the Sea, Roma, Before Sunset, Nomadland - some of the most critically admired films in the corpus. There is no published framework that teaches a writer how to construct a story that deliberately does not resolve. The structural grammar exists. The pedagogy does not.

Vogler and Hauge are structurally identical.

This one surprised us. The Hero's Journey (Vogler) and Michael Hauge's 6-Stage Structure both classify as Vyakula (The Turbulent Pursuit) with a Rebounding pakad modifier. The same vadi, the same samvadi, the same trajectory shape. A writer who has studied both frameworks has studied the same structural grammar twice in different vocabulary.

Harmon and Truby both resolve through revelation.

Dan Harmon's Story Circle and John Truby's 22 Steps both produce Bodha (The Awakening), despite being very different in granularity and approach. Harmon removes the battle beat entirely. Truby places four revelation beats after the battle. Both arrive at the same structural conclusion: knowing is the true climax, not fighting. The frameworks arrived at the same parent form by different routes.

McKee and the Fichtean Curve are structurally identical.

McKee's Story and the Fichtean Curve both produce Tapasya (The Crucible) with a Consolidating pakad. Both open in Ri (disruption), have Sa absent, are Pa-dominant, and resolve fully at Sa'. McKee arrives at this through conflict theory. The Fichtean Curve arrives at it through escalation mechanics. The structural outcome is the same.

What the frameworks do not teach

The most consequential finding from the framework mapping is not what the frameworks produce. It is what they leave untaught.

No major published framework teaches the open-ending archetypes. Vichar (The Open Road), Anishchit (The Suspended Note), Vedana (The Unhealed Wound), and Sahana (The Graceful Endurance) all close below Sa' - in crisis, irresolution, or qualified acceptance. Together they account for a substantial portion of the most critically respected films in our library.

Save the Cat explicitly prohibits them. The Final Image rule requires a transformed but resolved Sa'. Any writer following Snyder is structurally prevented from writing Manchester by the Sea or Arrival or Parasite. Not because those films are uncommercial - all three were commercially successful - but because the framework's grammar does not include the registers those films operate in.

No major framework teaches the distinction between Tapasya (The Crucible) and Dhairya (The Siege) - both are crisis-dominant with full resolution, but Tapasya is about endurance (the protagonist holds under sustained pressure) and Dhairya is about systematic advance (the protagonist moves through escalating crisis with discipline). The structural difference is real, the experiential difference is significant, and no published pedagogy names it.

No major framework teaches Spanda (The Living Pulse) - the structural form of the persistence narrative, where the protagonist keeps moving forward against a world that keeps generating fresh disruption rather than reversals or crisis. Moonlight. Sound of Metal. The Father. Poor Things. These films share a structural grammar that has never been named or taught.

The frameworks have been teaching the same six structural forms for decades. The other 18 have been discovered by writers through instinct, experience, and luck.

Why this matters now

The argument from the first post still holds: if AI commoditises content production, the differentiator is narrative compulsion. The story that makes someone feel something they will seek out again.

The framework monoculture is part of what produced the current landscape of remakes, sequels, and franchise extensions. Not because these are bad stories - some of them are excellent - but because an industry trained on six structural forms will unconsciously reach for those forms when under pressure. The development ecosystem rewards the recognisable shape. The recognisable shape is what the frameworks teach.

The 18 archetypes without corresponding frameworks are not exotic or difficult. They are the structural forms of some of the most beloved and enduring films in the library. They are simply untaught - which means they are underexplored, underproduced, and, in a world where AI will increasingly generate the recognisable forms by default, potentially the territory where genuinely original human authorship is most distinctive.

We are building a system that can identify which archetype a script is attempting and tell the writer whether the structural grammar supports that attempt. Whether the registers are in the right positions, whether the transitions are earned, whether the arc is doing what the logline promises.

The grammar exists. We found it. The question now is what writers do with it.

Note: As we expand our library, we are fairly confident we will find more archetypes, and some of them will lead to finding their parent archetype and help us collapse and expand. So we do expect to see this evolve as we continue to benchmark more films.

Note: While most of the corpus has been English films, we are excited to get to regional language films - be it Bollywood (Hindi), Korean, Japanese, European Art Cinema - to see what patterns we will uncover.