📘 Reproducibility as a Structural Basis for Cognition

From Physical Laws to Representational Algorithms

This repository contains the manuscript and supplementary materials for the paper:

“Reproducibility as a Structural Basis for Cognition:
From Physical Laws to Representational Algorithms.”

Zenodo DOI: https://doi.org/10.5281/zenodo.23259601

The work reconstructs foundational concepts in cognition, computation, and language —
kichi/michi (known/unknown), algorithms, symbol manipulation, representational states —
through a single unifying principle: reproducibility.

---

🔧 Core Idea

Cognition is modeled as ordered state‑transition sequences grounded in two layers of reproducibility:

1. Physical Law Layer (E)

Reproducibility emerges from the triad:

• Structure — finite interaction possibilities
• Interaction — causal constraints
• Energy — cost driving transitions

This yields stable physical trajectories.

2. Representational Layer (I)

Biological sensors cannot map trajectories; they map only states.
Thus reproducibility becomes partial:
only non‑breaking transitions between representational states are guaranteed.

This partial reproducibility defines:

• kichi (known) — reconstructible entity
• michi (unknown) — non‑reconstructible or unobserved entity
• algorithms — reconstructible state sequences
• symbol / language operations — socially constrained reproducible sequences

---

📚 Repository Contents

• paper_en.md — English manuscript
• README.md — overview and context

---

🧩 Relation to Previous Work

This paper extends two prior structural theories:

• Michi‑Drive Hypothesis — DOI: 10.5281/zenodo.20990020
• Categorical Internal World Model — DOI: 10.5281/zenodo.22870778

Together, these works form a unified framework for cognition grounded in state transitions and reproducibility.

---

✔ Zenodo Integration

This repository corresponds to the Zenodo record:
https://doi.org/10.5281/zenodo.23259601

---