Lanskrit
A polyphonic environment where the intuition of natural language meets the precision of formal systems.
The world’s languages encode evidence, uncertainty, relationship, time, participation and meaning in ways conventional code often cannot. Lanskrit turns those capabilities into new computational possibilities.
Code has logic · Language has lived complexity
Code records values.
Language records how they became known.
The seven evidence classes — the working grammar of Graft 01. Each is a structural distinction the language preserves, and that conventional code silently flattens.
- 01observedthe speaker witnessed it directly
- 02reportedanother party is the source
- 03inferredconcluded from available evidence
- 04assumedheld without direct grounding
- 05estimatedquantitative approximation
- 06rememberedrecalled, possibly imperfect
- 07disputedthe record itself is contested
From grammar to runtime.
The evidence classes above are not metaphors. They compile to a deterministic runtime — parseable, executable, and reproducible on any machine.
// Graft 01 — a provenance-aware value
let risk = inferred(
value: 0.72,
source: "risk_model_v3",
evidence_age_days: 18,
confidence: "provisional"
)
decide(risk, threshold: 0.8)An analytical sequence, not a marketing grid.
- 01
Linguistic feature
What distinction does the language preserve?
→ - 02
Code deficit
What does conventional code currently flatten?
→ - 03
Computational graft
What type, operator or rule could preserve it?
→ - 04
Trial
What changes when original and grafted systems are compared?
A language is not a score. It is a library of capabilities.
Open the Atlas- 01Evidence and provenance→
- 02Inclusive and exclusive participation→
- 03Event state and completion→
- 04Authority and relationship→
- 05Obligation and possibility→
- 06Context and topic→
- 07Classification and semantic type→
- 08Analogy and metaphor→
- 09Beneficiary and affected party→
- 10Degrees of certainty→