MLMN is not a theory presenting ten completed answers. It is a clinical reasoning framework under research, designed to preserve original examination information and allow comparison of how the current pathological state developed and how it changed after intervention. This page publishes the process of derivation and revision as well as the conclusions.
Starting Point: The “Two-Thousand-Year Problem of East Asian Medicine”
Through about 30 years of practice, Koichi Ishihara questioned why even sophisticated systems of classical Chinese medical theory could sometimes leave unclear the relationship between individual patients' life histories, illnesses, reactions, treatment courses, and specific treatment choices. Meanwhile, some Japanese Koho-school and early modern clinical texts preserve concrete accounts of physique, abdominal findings, daily life, changes before and after medication, failures, and recovery sequences, making the relationship between patient and treatment visible in greater depth.
This distinction is not framed as a simple ranking of China versus Japan or deduction versus induction. It is treated as a testable hypothesis about textual forms, the density of retained clinical information, and how treatment reasoning is shown. Where deductive systems in classical Chinese medicine connect with accumulated inductive clinical observations in Japan is examined through Yoshimasu Todo's toxin theory, Utsugi Kontai's Igaku Keigo and Kokun Iden, and the formation history leading to MLMN.
Current Structure: Ten Layers, L0–L9
Separating L0 from L1
L0 concerns the formation of life and initial conditions before and at birth. L1 is the baseline formed through postnatal living, development, illness, and treatment history. Prenatal origins are not automatically inferred backward from current findings.
Ministerial Fire as a Mechanism across Layers
The ministerial-fire theory associated especially with Zhu Danxi is not fixed as an independent layer. It is recorded with sources and possible refutations as a candidate mechanism spanning inputs, zang-fu function, qi, blood, and fluids, whole-body distribution, and temporal change.
L0–L9 are shelves for information, not ten stages every patient passes through in order. Branching, cycles, coexistence, changes over time, and responses to intervention crossing the layers are retained to compare multiple formation pathways. Definitions of each layer appear in MLMN Theory for reference.
Required Information When Publishing Each Layer
| Item | Published resources |
|---|---|
| Need | Why is a separate shelf necessary? What information does it prevent us from losing? |
| Original Source | Title, volume, chapter, passage or page, base edition, surrounding context, and original-image verification status |
| Attribution | Distinguish original text, later interpretation, historical practice, Ishihara's clinical interpretation, and MLMN reconstruction |
| AI Processing | Methods for searching, extracting, classifying, and detecting repeated passages, work boundaries, and OCR errors |
| Ishihara's Reconstruction | Clinical necessity, boundaries with adjacent layers, and propositions to test through cases |
| Refutation | Nonapplicable cases, competing hypotheses, and rules against inferring past causes from present findings |
| Verification Status | For example: OCR draft, checked against original images, confirmed by Ishihara, or not clinically validated |
Roles of AI and the Practitioner
AI searches classical materials beyond what one person can handle, extracts and classifies candidates work by work, and identifies repeated passages, work boundaries, OCR errors, and competing explanations. It does not establish a source merely from matching search terms or make final diagnostic and treatment judgments. Base editions, context, and original images are checked, and Ishihara reconstructs clinical necessity, boundaries, and falsifiable propositions in light of his experience.
Why We Do Not Hide the Early Paper
The early manuscript candidate dated May 16, 2026 includes the former L1–L10 structure, ministerial fire as the former independent L8 layer, strong expressions such as “complete definition” and “established,” and provisional extraction counts dependent on editions, OCR, and image comparison. Rather than deleting it as an embarrassing old version, we publish it unchanged as the starting point from which the theory began, was criticized, and evolved.
Check the Formation History through the Original and Its Changes
The early paper is an unreviewed record of the theory's formation, not the authoritative current theory.
How Patterns, Toxin, and Classics Are Handled
Pattern
Patterns are respected as powerful summaries of clinical judgment, while preserving the original observations, life history, medical history, and treatment responses that precede the summary.
Toxin
Yoshimasu Todo's toxin theory is neither identified with all classical medicine nor treated as the only correct answer. It is approached as a historical and clinical candidate for examining formative relationships across layers.
Classics
The Suwen, Lingshu, Nanjing, Shanghan Lun, Jingui Yaolue, Jin–Yuan medicine, and early modern Japanese medical works are not mechanically mapped one-to-one. Each work's context and verification status are retained.