# 6. Methodology and Validation

## The methodological problem

Macro-Culture is attractive because maps, gradients, and historical patterns invite large explanations. That is also its main danger. A model can appear confirmed if it uses vague variables, flexible boundaries, selective examples, or post-hoc scores.

The project therefore needs a strict separation between:

```text
observation → measurement → model → interpretation → prediction
```

## Data architecture

Each sector should have:

- a stable identifier;
- explicit geographic boundaries;
- raw measurements or source references;
- units and time periods;
- uncertainty and missingness;
- calculated indicators;
- historical records;
- quadrant scores;
- competing explanations;
- a record of who made the judgment and when.

The model should preserve raw inputs instead of storing only a final quadrant label.

## Scoring rules

A provisional scoring procedure could use separate 0–10 scales for:

- physical constraint severity;
- comprehension or adaptation capacity;
- confidence in the score.

The score must specify whether a high number means more pressure or more capacity. It must also avoid circularity: “comprehension” cannot simply mean that the sector achieved the outcome the model expected.

## Independent validation

A stronger test would use:

- preregistered variables and thresholds;
- historical cases selected before scoring;
- blind or semi-blind scoring where feasible;
- multiple researchers or sources;
- inter-rater comparison;
- counterexamples;
- out-of-sample prediction;
- sensitivity analysis for boundary choices;
- explicit alternative explanations.

## Counterexamples are essential

A useful counterexample is not a failure to be hidden. It tells us which variable is missing or which causal story is too simple.

Potential counterexample classes include:

- high river density without centralized hydraulic state formation;
- high solar input without sedentary agricultural expansion;
- resource abundance with persistent poverty or extraction dependency;
- similar climate zones with radically different political outcomes;
- technological adaptation imported from another region;
- societies that remain resilient through social institutions rather than high-energy infrastructure.

## Proposed analysis sequence

### Phase 1: local test

Complete the Great Lakes octagon with sourced physical, historical, and infrastructure variables.

### Phase 2: matched comparison

Compare the Great Lakes with at least two regions sharing similar latitude but different history, and two regions with different latitude but similar resource complementarities.

### Phase 3: historical back-test

Score earlier periods without using later outcomes, then compare predicted trajectories with documented historical transitions.

### Phase 4: global expansion

Only after the local and matched tests are coherent should the model expand toward the full conceptual grid.

## What would weaken or falsify the thesis?

The thesis should be revised if:

- environmental variables add no explanatory value beyond borders, trade, and political history;
- same-latitude competition and north–south complementarity do not recur across carefully selected cases;
- quadrant scores cannot be assigned independently of known outcomes;
- the model fails consistently out of sample;
- boundary changes produce radically different results with no principled way to choose among them;
- the framework erases agency, colonial power, class, gender, or institutional history rather than explaining their interaction with material conditions.

## Responsible use

The map should support questions, not justify stereotypes, deterministic policy, or claims that a population has a fixed cultural essence because of its location. Macro-Culture is most useful as a comparative systems lens with visible uncertainty.