Ellari Institute · SLA Framework Report · 2026

Shape Language
Alphabet

A medium-agnostic grammar for structural coherence and collapse — derived from a lyric essay, confirmed across 10 governed public domains, receipted.

Anchor specimen
Pulled Too Late
Domains mapped
10 governed + 2 candidate extensions
Specimens scanned
46+
Status
R1-D · ongoing
Receipt
SLA-SESSION-001
Claim Status: Corpus-aligned within this report · Purposive sample, not random · Public explainer · Not a clinical tool · Not a universality claim
How To Use This Report

Read this report as a public map of the SLA framework.

The formal papers carry the proof burden. The scanners teach the grammar through use. The receipts preserve the claim boundaries. The Institute maintains the publication record.

01

The Grammar

Three axioms generate the entire framework. They contain no reference to any specific domain — which is why the same grammar appears in music, software architecture, film, and relationships without modification.

I
Coherence Axiom
Any meaning-producing system is in one of three states: coherent, destabilized, or collapsed. The 10 L1 primitives specify these states and transitions. R1 is a process, not a state.
States: {F1, L2…D1, T3} · R1 = process · not domain-specific
II
The T3 Rule
A collapsed system cannot transition directly to coherence. Bypassing the return process produces Type C: the appearance of coherence without its structural conditions. Type C degrades — it resurfaces as T3.
T3 → F1 BLOCKED · bypass(T3, R1) → Type C · Type C → T3
III
R1-D Principle
Movement toward coherence (R1-D) is categorically more valuable than claimed coherence (Type C). R1-D keeps the return process alive. Type C closes it by claiming completion.
R1-D > Type C · R1-D ≠ R1 achieved · R1-D maintains generativity

The ten L1 primitives.

GlyphPrimitiveDescription
F1Linear FlowCoherent progression, expected output. The paradigm producing normal science.
L2Feedback LoopRecursive process, output feeds input. The citation network. The looping motif.
P1ProjectionA system oriented toward a target not yet reached. Sillage. Leading tone. The anomaly the field still believes it can solve.
I1Identity ShellThe stable organizing framework. The paradigm as I1. Chanel No.5 as I1.
T1Fracture LineA single event that cannot be unfelt. The Michelson-Morley experiment. The hang-up.
T3Collapse BasinThe state from which the system cannot self-correct. T3→F1 blocked. Requires R1.
S1Shearing PlanesTwo systems in unresolvable structural tension. Meta + Facebook running simultaneously.
S3Attractor HubAn organizing center exerting gravitational pull. The groove. The drone. Post-revolutionary paradigm.
R1Return ArcThe process of return — not the arrival. Direction is confirmed; completion is not claimed.
D1Affect WarpDistortion without systemic collapse. Blue notes. Rubato. Doublespeak.
02

The Return Arc

Not all paths from T3 are the same. Four subtypes determine the governance note, the compositional implication, and the diagnostic response. Distinguishing them prevents the most common misapplication of the framework.

R1-A Unavailable
R1 is structurally unavailable. The conditions producing T3 have not changed and cannot be changed within the current scope.
Governance note: witness, do not prescribe return
Strange Fruit · Lysenkoist science · Infrastructure (collection)
R1-B Available, Not Taken
Return is structurally available. The system has the resources to return. It has not.
Governance note: R1 modeling may be appropriate
Pulled Too Late · Priestley / oxygen · Exit Music
R1-C ✕ False Repair
T3→F1 attempted. Coherence is claimed without the return process. The structural source is unchanged.
Governance note: avoid — models bypassing T3
Fix You · Tropicana rebrand · pre-registration without incentive reform
R1-D Directional
Movement toward return is real and verifiable. Arrival is not forced or claimed. Most therapeutically and compositionally useful.
Governance note: most generative — keep the process alive
Cologne Concert · A Love Supreme · Walking (collection)
03

The Collection

Six pieces sharing three structural positions across the R1 taxonomy. The 11:02 collection is the first creative work in the corpus with cross-domain receipts confirmed across text, music, and visual design simultaneously.

0
The Week Before
F1 + I1
Baseline. Shoes as just shoes. Descent begins from coherence, not failure.
#1a9e78
ii
Just Before
◇ not sealed
Same 168-second window. The ◇ approached, not crossed. The near-miss invisible to I1.
#3b9e1a
iii
Pulled Too Late
R1-B
The anchor. ◇ sealed. R1 available via the boy. Not taken. T3 terminal.
#c44040
iv
Fine
R1-C ✕
False repair. Prose claims F1. Content confirms T3. The grammar error is visible to the reader, not the character.
#c4704a
v
Infrastructure
R1-A
No characters. No timestamp. A system running as designed. "Still running. As you read this."
#888480
vi
Walking
R1-D
The boy. Undated. Direction without arrival. Movement initiated. Arrival not claimed.
#b87514
04

Cross-Domain Evidence

The grammar arc was mapped across text, music, and visual design independently from the same anchor specimen. All three analyses produced the same arc. This is the corpus's first confirmed three-domain alignment.

Text Pulled Too Late Visual 11:02 color system Music 11:02 score SLA-CROSS 001 F1 → D1 → S1 → T1 → L2 → T3 confirmed independently · same arc · no coordination
05

Twelve Active Domains

The grammar applies to any domain that produces meaning over time. Each domain extension was mapped independently. The convergence across domains is the system's primary evidence claim — not a universality claim.

#DomainT3 specimenR1 specimen
01Text / NarrativeVoicemail loop degradingR1-D: the boy walking
02MusicTristan Prelude — T3 as methodR1-D: A Love Supreme Psalm
03Visual DesignTwitter→X — I1 destroyedR1-D: Apple 1997 return
04TypographyCompeting weights, no hierarchyR1-D: type system governance
05Motion / AnimationVestibular-triggering animationR1-D: shared element transitions
06Spatial / BuiltHospital wayfinding failureR1-D: courtyard as S3
07Code / SoftwareDistributed monolithR1-D: strangler fig pattern
08Film / CinemaLa La Land finale — T3 claimed as R1R1-D: Manchester by the Sea
09Relational / SocialSame fight, no resolutionR1-D: "I don't know, but I'm trying"
10Scientific ParadigmsReplication crisis (active T3)R1-D: Registered Reports
11AI / AlignmentRLHF → approval not valuesR1-D: Constitutional AI
12Language / LinguisticsDoublespeak (Orwell 1946)R1-D: neologism that names the gap
13*Ecology (candidate)Invasive monocultureR1-D: keystone reintroduction

* Domain 13 (Ecology) is a candidate extension pending further mapping. Not counted in the active domain set.

06

What the Corpus Shows

46 specimens mapped across 10 governed domains. The distribution is consistent with the T3 Rule: Type C is dominant in commercially successful work; R1-D is dominant in work that endures. This is a pattern within a purposive sample — not a statistical claim about all creative production.

R1-C ✕ — False Repaircommercially dominant · Type C
43%
R1-D — Directionalartistically respected · most honest from T3
30%
R1-A — Unavailablepractice-relevant for witnessing
17%
R1-B — Available, Not Takenentry point for reform or therapeutic work
9%
07

Tools Built

Four interactive governance scanners apply the grammar across domains. Each returns a receipt: R1 type, grammar arc, confusion check, and governance note — not a diagnosis.

Pop Scannerworking
Submit any song. Receive its grammar arc, R1 type, confusion check, and governance note. Type C detector built in — ask why a piece does not resolve.
Domains: Music · 46+ specimen receipts
Design Scannerworking
Submit a brand, rebrand, or logo system. Receive an SLA-V visual grammar receipt. Canonical False Repair audit for design decisions.
Domains: Visual Design · Typography · Motion
Unified Scannerworking
Three-domain scan from one specimen. Runs text + music + visual independently and compares grammar arcs for alignment. Cross-domain receipt generated.
Domains: Text · Music · Visual (simultaneous)
SLA Learninteractive
Interactive grammar primer. Teaches all 10 L1 primitives through use across specimens. Generates a personal grammar map from responses.
Domains: All · beginner-facing
08

Session Receipt

The session that produced this report is itself an SLA specimen. It entered at F1 (a single lyric essay) and moved through P1 as each domain implied the next. The three axioms were the destination the session was always pointing toward.

SLA-SESSION-001 R1-D
Starting specimen"Pulled Too Late" — lyric essay, eight sections
Cross-domain receiptSLA-CROSS-001 confirmed — text + music + visual, same arc, independent analyses
Total specimens46+ across 10 governed domains + 1 candidate ecology extension
Domains mappedText · Music · Visual · Typography · Motion · Spatial · Code · Film · Relational · Scientific · AI · Linguistics
Grammar arcF1 (initial scan) → P1 (each domain implied next) → S3 (3 axioms) → R1-D (ongoing)
T3 Rule statusHONORED — no Type C claims made in this report
DAVAR statusCandidate — pending 3 authorial confirmations
Session grammarR1-D · direction shown · arrival not claimed
notAllowedToProve
Cannot claim the grammar is universal — the corpus is purposive, not randomly sampled. Cannot diagnose specific people, companies, or systems without their participation. Cannot claim the framework is complete. Cannot claim that independent confirmation (Kuhn, raga tradition) constitutes peer review. Cannot claim this report substitutes for authorial confirmation, clinical judgment, or institutional review. A well-structured system is not automatically a true system.