Mathematical Foundation for GUAF Collaboration
Status: proposal
Date: 2026-10-08
Context: Scott's self-assessed mathematical background includes fragments of Calculus 1-2, scattered Linear Algebra without formed mental models, and primitive Differential Equations. Symbol-to-meaning mapping and equation interpretation are identified gaps. This document proposes a foundational mental model designed not to teach all of mathematics, but to provide a universal orientation scaffold — a way to encounter any new mathematical concept and immediately know where it fits, what it does, and how to begin understanding it.
The Core Insight: Mathematics is Not Calculation
The biggest obstacle isn't missing knowledge — it's a misaligned mental model. If math feels like "manipulating symbols according to rules you don't understand," every new topic is alien. The alternative model:
Mathematics is the study of structures (collections of things with relationships) and the maps between them that preserve those relationships.
Everything else — calculus, linear algebra, category theory, the GUAF axioms — is an instance of this pattern. You don't need to master each instance. You need to recognize the pattern.
The Universal Scaffold: Five Questions
When encountering ANY new mathematical concept, ask these five questions in order. They form a complete orientation frame:
1. What are the objects? (The nouns)
Every mathematical domain has "things" it talks about.
- In set theory: sets and elements
- In linear algebra: vectors and matrices
- In topology: open sets and points
- In GUAF: contexts (Ctx), presheaves (F), questions (Q)
Your move: Don't try to understand everything. Just identify the nouns. "What kind of thing am I looking at?"
2. What are the operations/morphisms? (The verbs)
How do the objects interact or transform?
- In arithmetic: addition, multiplication
- In linear algebra: matrix-vector multiplication, linear maps
- In category theory: morphisms (arrows between objects)
- In GUAF: restriction maps (
restr), the ladder map (mu)
Your move: Find the verbs. "What can I DO with these objects?"
3. What are the rules/axioms? (The constraints)
What must always be true when objects and operations interact?
- Groups: associativity, identity, inverses
- Vector spaces: distributivity, scalar compatibility
- GUAF: empty global section, no fixed point, productivity
Your move: Axioms aren't arbitrary. They're the load-bearing walls. Ask: "What breaks if I remove this rule?"
4. What are the maps between structures? (The translations)
How does one structure relate to another while preserving its shape?
- Functions between sets
- Linear transformations between vector spaces
- Functors between categories
- The GUAF bridge (mapping a formal theory's undecidable core into a GUAF structure)
Your move: This is where deep understanding lives. "Can I translate this problem into a different structure where it's easier to solve?"
5. What is preserved? What is lost? (The invariants)
When you apply a map or transformation, what stays the same?
- Topology preserves connectedness, not distance
- Linear maps preserve straight lines, not angles (unless orthogonal)
- GUAF's representability axiom ensures internal structure faithfully mirrors external behavior
Your move: "If I transform this, what can I still trust?"
How This Maps to Your Current Gaps
| Gap | Reframe via the Scaffold |
|---|---|
| "I struggle processing equations" | An equation is just asserting two expressions name the same object (Q1) under specific rules (Q3). Read left side, read right side, ask "why are these the same thing?" |
| "I can't map symbols to meaning" | Every symbol is either a noun (object, Q1), a verb (operation, Q2), or a constraint (axiom, Q3). Classify the symbol first; meaning follows from role. |
| "Linear algebra hasn't clicked" | Vectors = objects (Q1). Matrix multiplication = operation (Q2). Ax = b means "apply verb to noun, get noun." Eigenvalues = invariants (Q5). |
| "Calculus felt like fragments" | Derivatives = operations on functions (Q2). Integrals = inverse operations (Q2). FTC = the invariant connecting them (Q5). |
Proposed Learning Protocol
Phase 0: Notation Decoding (immediate, parallel to GUAF work)
When we encounter Lean code or math notation together, I will explicitly decompose it using the Five Questions before proceeding. You build the habit of asking them; I provide the initial scaffolding.
Phase 1: Concrete Structures First
Rather than studying abstract math top-down, we start with structures you already partially know:
- Sets and Functions — the simplest objects and maps
- Posets (Partially Ordered Sets) — directly relevant to GUAF's
Site - Graphs — visual, intuitive, maps cleanly to category theory later
Phase 2: The Bridge Concepts
These are the ideas that connect your software engineering intuition to mathematical thinking:
- Types as Sets — you already think in types (TypeScript, Lean); math types are the same instinct formalized
- Functions as Data — higher-order thinking; functions that take/return functions
- Structure Preservation — the key insight behind homomorphisms, functors, and the GUAF bridge
Phase 3: Targeted Depth (only as needed for GUAF)
- Presheaves — directly required for Phase A of the collaboration plan
- Fixed-point theory — required for understanding why the ladder axiom matters
- Gödelian encoding basics — required for MT3/MT4
How We Work Together on This
| Cognitecti's Role | Scott's Role |
|---|---|
| Decompose every new concept into the Five Questions before formalizing | Judge whether the decomposition "clicks" or needs reframing |
| Translate Lean type signatures into plain English using the scaffold | Identify which translation still feels opaque |
| Provide concrete examples before abstract definitions | Test the examples against your own intuition |
| Flag when a concept requires prerequisite understanding we haven't built | Decide whether to build the prerequisite or route around it |
| Never assume prior knowledge; always make the implicit explicit | Signal when explanation is redundant vs. when it's needed |
The Anti-Pattern to Avoid
Do NOT attempt to "learn math" as a separate project from the GUAF work. The foundation is built by doing the work with the scaffold applied. Every Lean definition we write, every error we debug, every sorry we try to close is an opportunity to practice the Five Questions. Abstract study without concrete application drifts into Low Resonance.
Immediate Next Step
Before committing and pushing, confirm:
- Does the Five Questions scaffold resonate as a usable mental model?
- Should we adjust the learning protocol phases?
- Are you ready to proceed with the git commit/push for both the GUAF build artifacts and these planning documents?