Add HUF Reference Collection (science/reference/)

Nine core documents in strategic learning order (~105 min total):
01 Executive Summary (cylinder problem, fuel gauge)
02 The Four Monitoring Categories (MC-1 through MC-4, from MONITOR-001)
03 The Open-Loop Doctrine (Skydiver Principle, from LOOP-001)
03b The Ontological Foundation (pre-existing condition, from ONTO-001)
04 EITT Complete Explanation (discovery narrative)
05 EITT Finding (empirical results)
06 EITT CoDa Mathematics (formal backbone)
07 The Lineage (loudspeaker to universal instrument)
08 Vocabulary Card (CoDa terms, phrases that land)
09 Formula Reference (study card)

Docs 02, 03, 03b are new: human-readable conversions of the
JSON governance records. All others consolidated from drafts.
INDEX.md provides learning path + topic-based navigation.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: Peter Higgins

README.md

@@ -42,6 +42,7 @@ This produced three new frameworks:
 ```
 HUF/
 ├── science/                    # The unified front
+│   ├── reference/              # Core reference collection (start here, 9 docs, ~105 min)
 │   ├── chemistry/              # EITT chemistry results, HUF-IDX, PRISM
 │   ├── core/                   # EITT mathematics, complete explanation, formulas
 │   └── methodology/            # Compositional governance scale, confidence index
@@ -70,10 +71,9 @@ HUF/
 
 | Time | What | Where |
 |------|------|-------|
-| 15 min | **User Handbook** — fast summary + guided links into the full document set | [`science/core/HUF_USER_HANDBOOK.md`](science/core/HUF_USER_HANDBOOK.md) |
-| 5 min | What HUF is, in plain language | [`science/core/WHAT_HUF_IS.md`](science/core/WHAT_HUF_IS.md) |
-| 10 min | The EITT finding and the numbers | [`science/core/EITT_Finding.md`](science/core/EITT_Finding.md) |
-| 15 min | Full explanation with CoDa mathematics | [`science/core/EITT_CODA_MATHEMATICS.md`](science/core/EITT_CODA_MATHEMATICS.md) |
+| 5 min | The cylinder problem and fuel gauge — one-page HUF | [`science/reference/`](science/reference/01_EXECUTIVE_SUMMARY.md) |
+| 105 min | **Reference Collection** — 9 documents, full learning path | [`science/reference/INDEX.md`](science/reference/INDEX.md) |
+| 15 min | **User Handbook** — fast summary + guided links | [`science/core/HUF_USER_HANDBOOK.md`](science/core/HUF_USER_HANDBOOK.md) |
 | 20 min | Chemistry results (the new frontier) | [`science/chemistry/`](science/chemistry/) |
 | 30 min | The kill test — 19 documented failure modes | [`huf-gov/governance/`](huf-gov/governance/) |
 

science/reference/01_EXECUTIVE_SUMMARY.md

@@ -0,0 +1,29 @@
+# HUF — Executive Summary
+
+## The Cylinder Problem
+
+Look at a cylinder from one end and you see a circle. Look at it from the side and you see a rectangle. Both views are correct. Neither is complete. The circle doesn't know about the length. The rectangle doesn't know about the curvature. You need both projections — or better, you need to know the object is a cylinder.
+
+Complex systems work the same way. An energy grid, a wetland ecosystem, an industrial process — each produces a handful of familiar indicators. Output levels, efficiency scores, status reports. Each indicator is a projection — a shadow of a higher-dimensional object cast onto a flat surface. The projections look stable. The cylinder is rotating.
+
+**HUF watches the cylinder, not the shadows.**
+
+It monitors the full composition — all the parts of a system that must sum to a whole — and detects when the internal structure drifts, even when every individual projection still looks normal. A country's total electricity output holds steady (the shadow looks the same), but underneath, coal is replacing nuclear while gas is replacing coal while renewables are replacing gas. The parts are rearranging. The shape is changing. No single indicator catches it because each one is just a circle or a rectangle. HUF sees the cylinder turning.
+
+## The Fuel Gauge
+
+Every system that sums to a fixed total is a fuel gauge. A country's electricity mix. A wetland's species distribution. A budget. A diet. The needle shows one number — how full the tank is. But the fuel gauge doesn't tell you what's *in* the tank.
+
+HUF reads the fuel gauge and the fuel composition at the same time. It knows the tank is full. It also knows that last year it was 60% gasoline and 40% ethanol, and this year it's 50/50. The needle didn't move. The fuel changed. That's compositional drift, and it matters — because an engine tuned for one blend will fail on another.
+
+## One Sentence
+
+HUF detects structural change in complex systems by monitoring the full composition rather than individual indicators — catching the cylinder rotating before any single shadow changes shape.
+
+## For CoDa Colleagues
+
+The mathematics of compositional data analysis — Aitchison geometry, log-ratio transforms, perturbation, the simplex — provides the rigorous foundation for what HUF does operationally. HUF adds the monitoring architecture: continuous observation, drift thresholds, and the distinction between passive measurement and active governance response. The geometry is CoDa's. The monitoring protocol is HUF's. The opportunity is to join them.
+
+## For Everyone Else
+
+You know those dashboards with green lights that stay green right up until something breaks? HUF is the system that would have turned yellow three months earlier — not because any single reading was wrong, but because the *pattern* of readings was quietly rearranging itself.

science/reference/02_THE_FOUR_MONITORING_CATEGORIES.md

@@ -0,0 +1,84 @@
+# The Four Monitoring Categories
+
+**DocCode:** REF-002 (derived from MONITOR-001)  
+**Source:** `huf-gov/science/MONITOR-001.json`  
+**Status:** State of record  
+
+---
+
+## The Gap
+
+No unified taxonomy of monitoring categories exists. The first three — Magnitude, Identity, Trend — are ubiquitous but never unified into a formal framework. The fourth — Composition — has a 40-year mathematical foundation (Aitchison 1982, Compositional Data Analysis) but was never framed as a monitoring category.
+
+HUF's contribution is applying existing mathematics as a monitoring observable, not inventing new mathematics.
+
+---
+
+## The Four Categories
+
+| Category | Question | What It Reveals | What It Misses | Status |
+|----------|----------|-----------------|----------------|--------|
+| **MC-1: Magnitude** | How much? What is the scalar quantity? | Scale, size, total volume | Internal structure (deceptive drift) | Universally deployed |
+| **MC-2: Identity** | Who or what? What are the named constituents? | The names on the axes; who the players are | Internal dynamics; balance of power shifting | Universally deployed |
+| **MC-3: Trend** | Which direction? How is magnitude changing over time? | Direction and speed of aggregate change | Internal rotation (object rotates without translating; shadow unchanged) | Universally deployed |
+| **MC-4: Composition** | What is the internal balance? How are parts arranged? | Structural dynamics; zero-sum trade-offs; concentration; deceptive drift | Absolute scale (deliberately removed by normalisation) | Mathematically established; diagnostically absent |
+
+---
+
+## Operational Examples
+
+**MC-1 (Magnitude):** Total GDP, total hectares, total energy generation, total case counts.
+
+**MC-2 (Identity):** Country name, sector label, species ID, demographic category.
+
+**MC-3 (Trend):** GDP growth rate, energy change year-over-year, population growth, revenue trajectory.
+
+**MC-4 (Composition):** Energy mix (coal, gas, nuclear, renewables; sum = 1). GDP composition (agriculture, industry, services; sum = 1). Ecosystem balance (water, marsh, forest, grassland; sum = 1).
+
+---
+
+## Mathematical Home
+
+MC-1 through MC-3 operate on Stevens' measurement scales (1946): ratio scale (true zero), nominal scale (categorical), and interval/ratio scale (time-indexed).
+
+MC-4 operates on the Aitchison simplex: a bounded space where all components sum to a constant (typically 1), and the natural operations are perturbation (multiplicative change) and powering (scalar scaling). The geometry is Riemannian. Distances are measured by the Aitchison metric. Coordinates are log-ratios.
+
+---
+
+## Why MC-4 Was Invisible
+
+**The closure problem was treated as nuisance, not signal.** When statisticians encountered unit-sum data, they treated the constraint as a technical annoyance requiring correction (spurious correlations, singular covariance matrices). Aitchison showed it was geometry — but the monitoring community never adopted the result.
+
+**Magnitude dominance.** Institutional culture rewards growth. Monitoring systems are built to answer "how much?" because funding, policy, and careers are denominated in magnitude. Composition is not rewarded in the same way.
+
+**The primer gap.** The mathematics community (CoDa) had the structure. The monitoring community (WHO, OECD, UN) had the context. Neither read the other. HUF sits at the boundary where these two communities meet.
+
+---
+
+## Using Existing Methods to Define MC-4
+
+**UN Results-Based Management:** Indicator definition: Structural Concentration Index (K_eff) — dimensionless scalar on [0,1]. Deliberately marked with NO target, because MC-4 is diagnostic, not prescriptive. The instrument reads. The human decides.
+
+**OECD extension:** Seventh evaluation criterion: Structural Integrity — "did the intervention maintain or improve internal proportional balance, or introduce structural concentration?"
+
+**WHO extension:** Composition-Based Surveillance (CBS) — monitors proportional distribution across subpopulations; detects hollowing before magnitude triggers.
+
+**ISO standard:** Future ISO standard for composition monitoring instruments. Precedent: ISO 7240 (fire systems), ISO 10816 (vibration), ISO 14644 (cleanrooms).
+
+---
+
+## Academic Lineage
+
+Stevens (1946) → Shannon (1948) → Aitchison (1982) → Pawlowsky-Glahn & Egozcue (2001) → WHO/OECD/UN frameworks → Higgins (2025–2026) HUF
+
+---
+
+## Open Questions
+
+- **MONITOR-Q1:** Is the four-category taxonomy complete, or are there additional monitoring intents?
+- **MONITOR-Q2:** Can MC-1, MC-2, MC-3 be formally derived as projections from MC-4?
+
+---
+
+*Definitive source: `huf-gov/science/MONITOR-001.json`*  
+*Peter Higgins | Rogue Wave Audio | PeterHiggins@RogueWaveAudio.com*

science/reference/03_THE_OPEN_LOOP_DOCTRINE.md

@@ -0,0 +1,73 @@
+# The Open-Loop Doctrine
+
+**DocCode:** REF-003 (derived from LOOP-001)  
+**Source:** `huf-gov/governance/LOOP-001-open-loop-doctrine.json`  
+**Status:** State of record — not negotiable  
+
+---
+
+## The Skydiver Principle
+
+HUF-GOV is the altimeter. It is NOT the hand pushing you out of the plane. The skydiver jumps on their own volition.
+
+**Formal version:** HUF-GOV observes, measures, computes, and reports. It does NOT decide, act, recommend action, trigger responses, or automate downstream behaviour. The human operator receives the instrument readings and decides what to do. The loop stays open.
+
+**One line:** The instrument reads. The human decides. The loop stays open.
+
+---
+
+## Why Open Loop
+
+**Technical reason:** Closing the loop requires a controller model — assumptions about "correct" composition, "right" responses, thresholds. These are domain-specific, context-dependent, politically loaded. No universal controller exists.
+
+**Governance reason:** Closed-loop removes human agency. If the analyser detects drift and automatically triggers alerts, who decided the threshold? Who decided drift meant degradation? That is not governance — it is automation wearing governance costume.
+
+**Safety reason:** The Montreal machine did not decide to hurt anyone — a human bypassed safety. But if given authority to act autonomously, it might shut down during normal operation or fail to during abnormal operation. Passive failure (shear bolt) works because it does not decide.
+
+**Practical reason:** Different domains need different responses to the same reading. K_eff decline in electricity = policy failure (action needed). K_eff decline in a wetland = seasonal variation (no action). K_eff decline in finance = opportunity. The instrument cannot know the context. Only the human operator can decide.
+
+**Peter's reason:** The human is not a bug — the human IS the system.
+
+---
+
+## What Closing the Loop Looks Like (Violations)
+
+| Violation | Why Wrong | Correct Approach |
+|-----------|----------|-----------------|
+| Automatic alerts triggered by drift thresholds | Who set the threshold? What if drift is normal in this context? | Analyser displays drift; operator decides to escalate |
+| Recommended actions appended to outputs | Recommendations are control signals disguised as suggestions | Analyser reports measurements; operator decides response |
+| Integration with automated trading/regulatory/policy systems | Instrument becomes controller; measurement becomes action; human removed | Analyser is standalone observation instrument; feeds human judgment, not automated pipelines |
+| Predictive thresholds (tipping point warnings) | Prediction is CLS territory; requires "should be" model; GOV observes "what is" | Complexity trace shows trajectory; operator recognises pattern and assesses |
+| Sensitivity knobs adjusting "significance" | Adjusting significance is governance decision, not measurement decision | Instrument reports raw values; significance thresholds are operator's domain |
+
+---
+
+## The Engineering Paradox
+
+HUF began as a closed-loop BTL loudspeaker control system. Peter deliberately opened it — opposite of normal engineering evolution (open → closed). In governance, closing removes the human.
+
+HUF-GOV is not performance optimisation — it is a governance instrument.
+
+HUF-CLS (closed-loop) exists for domains where closing is appropriate (BTL loudspeaker control, process automation). The boundary between GOV and CLS is not a design choice — it is an ontological requirement.
+
+---
+
+## The Skydiver Test
+
+Before adding any feature to the GOV analyser, ask:
+
+**Does this push the skydiver, or show the skydiver the altimeter?**
+
+If push → CLS territory.  
+If show → GOV territory.
+
+---
+
+## The Boundary
+
+Not negotiable. HUF-CLS for performance. HUF-GOV for observation. The two architectures share mathematics but serve fundamentally different purposes. Mixing them is the single most dangerous tendency in the project.
+
+---
+
+*Definitive source: `huf-gov/governance/LOOP-001-open-loop-doctrine.json`*  
+*Peter Higgins | Rogue Wave Audio | PeterHiggins@RogueWaveAudio.com*

science/reference/03b_THE_ONTOLOGICAL_FOUNDATION.md

@@ -0,0 +1,74 @@
+# The Ontological Foundation
+
+**DocCode:** REF-003b (derived from ONTO-001)  
+**Source:** `huf-gov/science/ontological-foundation.json`  
+**Status:** State of record — foundational  
+
+---
+
+## The Claim
+
+HUF does not create the systems it measures. The ratio-state condition is pre-existing. Every compositional system — systems whose elements express as proportions of a whole — is already a ratio-state system, already redistributing, already drifting. The instrument reveals structure that was present before measurement began.
+
+**Formal:** For any system S whose state = vector of proportions p = [p₁, p₂, ..., pₙ] where Σpᵢ = 1, the ratio-state dynamics (carrier redistribution, effective dimensionality variation, structural velocity) are intrinsic properties of S. They do not require observation to occur. They require observation to be known.
+
+**In seven words:** No silent drift, only the unobserved.
+
+---
+
+## What This Means
+
+**For the instrument:** The HUF Spectrum Analyser does not generate its readings. It reads structure already there. K_eff is not a construct — it measures how many carrier elements meaningfully participate. TV distance is not a model — it is the distance the system actually travelled. Deceptive drift is not a prediction — it detects concentration building behind apparent stability.
+
+**For the operator:** The operator does not need to believe in HUF for it to apply. Their wetland is already a ratio-state system. Their energy grid is already redistributing. The operator's choice is not whether to participate — they already do. The choice is whether to observe.
+
+**For new domains:** The claim is not "we built something you should adopt." The claim is "your system already exhibits these dynamics; here is an instrument calibrated to read them." Difference between selling a product and opening a door.
+
+---
+
+## The Invisible Door
+
+HUF does NOT hold up a mirror (which shows what you already see). HUF shows a room in your house with a door that was invisible. The room was always there. The dynamics were always operating. The door was always accessible. You simply did not know it existed.
+
+**Why not a mirror:** A mirror reflects existing understanding. HUF reveals dynamics the observer had no framework to perceive. Effective dimensionality is invisible in raw data. Deceptive drift is specifically invisible to conventional monitoring. The instrument does not confirm what experts know — it shows them a structural layer they had no access to.
+
+**Practical consequence:** Not "your system has problems you don't know about" (confrontational). Not "look at your data differently" (trivial). Rather: "there is a structural layer your current instruments don't access. Here is the door."
+
+---
+
+## The Pre-Existing Condition
+
+**Step 1:** Any system with components as proportions of a whole exists on a simplex. If you say "coal is 40% of the energy mix," you have placed the system on the simplex. Not a model choice — consequence of data structure.
+
+**Step 2:** On the simplex, all proportions are interdependent. Change in one forces redistribution across all others. The unit-sum constraint Σpᵢ = 1 is mathematical, not empirical. Holds by construction.
+
+**Step 3:** Forced redistribution produces measurable dynamics: effective dimensionality changes, structural velocity, carrier concentration. These are observables, not assumptions.
+
+**Step 4:** Dynamics operate continuously, regardless of observation. A wetland does not pause ecological redistribution between surveys. An energy grid does not freeze its fuel mix between reports. The system is always in motion.
+
+**Step 5:** Therefore, every proportional system IS already a ratio-state system. The only variable is whether anyone reads the instrument.
+
+---
+
+## The Zero State
+
+**Claim:** In a ratio-state universe, the zero state does not exist.
+
+A system on the simplex cannot have all proportions equal to zero. At least one element must carry the full unit sum. The system cannot reach a zero-dynamics state — any perturbation, however small, redistributes. It may approach a stable attractor (one dominant carrier absorbing nearly all), but residual carriers remain nonzero and subject to redistribution. No ground state. No off switch.
+
+Drift is not an event that starts and stops. Drift is the continuous condition of any proportional system. What changes is rate, character, visibility — not existence.
+
+---
+
+## Three Core Sentences
+
+1. "Every system whose state can be expressed as proportions is already a ratio-state system, subject to simplex dynamics that operate continuously and independently of observation."
+
+2. "The HUF instrument does not create these dynamics — it reveals a structural layer present before measurement began, accessible through a door the domain had no reason to know existed."
+
+3. "The only state that does not exist in a ratio-state system is zero: there is no silence, no stillness, no ground state — only the unobserved."
+
+---
+
+*Definitive source: `huf-gov/science/ontological-foundation.json`*  
+*Peter Higgins | Rogue Wave Audio | PeterHiggins@RogueWaveAudio.com*