**Motivations:**
- Reduce ambiguity between transmissible memory, hidden variables, and non-Markovian projections
- Make the scope of quantification/interpretation explicit (indexation by mu/P/d/c/L) and improve scientific navigability
**Root causes:**
- Layer boundaries (ensemblist vs measured/probabilistic vs decision vs thermodynamic) were not always explicit, enabling over-interpretations
- Some passages mixed interpretative readings with core results without a uniform "hypotheses -> statement -> optional reading" discipline
**Correctifs:**
- Add an explicit "memory" note (transmissible register vs omitted state) and require state/projection clarity when invoking memory
- Clarify the default status of transition cost c as a graph weight and separate physical instantiations as optional, declared dictionaries
- Reframe cosmological/ontological sections as conditional interpretations and tighten wording that could imply teleology
**Evolutions:**
- Add per-chapter "Hypotheses and results" markers to separate assumptions, results, and status
- Introduce symbol/dependency indices, a quantification choices registry, and a robustness protocol; add discrete<->continuous repères and explicit non-transferability warnings
**Pages affectées:**
- v0/livre.md
Co-authored-by: Cursor <cursoragent@cursor.com>
