Latticework

Overview

Latticework is the practice of cross-wiring mental models from multiple disciplines on the same situation. Power comes from inter-connection: independent lenses converging = high-confidence signal; lenses diverging = unknown to investigate. When multiple forces align simultaneously they amplify — the lollapalooza effect (Munger, 1994). Composes with first-principles, second-order-thinking, probabilistic-thinking, and map-is-not-the-territory.

When to Use

Stakeholders keep raising non-overlapping objections — each is right from their model
Post-mortem shows failure was "outside the model we used"
"Our analysis is solid" — but only one framework was applied
Situation looks like a classic X but has anomalous features X cannot explain
Designing a strategy/product where market, psychology, operations, and incentives all interact

Not when: problem is contained in one discipline; crisis triage (no time); decision too small for multi-model overhead.

Coaching Novices (Adaptive Front Door)

Engine mode: user has a concrete case → run The Process directly.
Coach mode: user is unfamiliar → guide step by step.

In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.

One-liner: facts don't become knowledge until they hang on a latticework of theory (Munger's rule #1).
Check fit: has one model already failed or felt incomplete? If yes, proceed.
Elicit: what models applied so far? What disciplines are missing?

[WAIT — do not advance until user responds]

Run The Process one step at a time with their input.

[WAIT — do not advance until user responds]

Close: name the convergence map, blind spots, and any lollapalooza effects found.

[WAIT — do not advance until user responds]

The Process + Output Template

# Latticework Analysis: <situation>

## 1 — Phenomenon
  Core question:
  Prior single-model framing + its known blind spot:

## 2 — Lenses (3–5, genuinely independent disciplines)
| # | Discipline | Key Prediction | Force (+/-/0) |
|---|-----------|---------------|--------------|
| 1 | Economics | | |
| 2 | Psychology | | |
| 3 | Systems | | |

## 3 — Convergence Map
  ≥2 lenses agree (higher-confidence):
  Lenses disagree (live unknown — investigate):
  Lollapalooza: multiplicatively aligned forces?

## 4 — Blind Spots
  What no lens covers:

## 5 — Calibrated Conclusion
  Recommendation + Confidence:
  Key residual uncertainty:
  Information that would most change the picture:

→ Method in Action: Charlie Munger 1994 USC Business School Address

Pack: Latticework Across Domains

| Domain | Typical single lens | Key missing lens | |---|---|---| | Startup PMF | Customer interviews | Systems (adoption loops) + History | | Pricing | Demand curve | Game theory (competitive response) | | M&A | Financial synergies | Psychology (culture) + History (base rates) |

Applying It Well

Independence matters: 3 re-labeled versions of the same model is not a latticework
Divergence = information; stop adding lenses when marginal new predictions cease (3–5)

→ Primary sources: references/sources.md

Common Rationalizations

[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.

| Fake move | Reality | |---|---| | [D] "We already did a full analysis" | One framework applied thoroughly is a single-lens deep dive — not a latticework. | | [D] "Adding more models adds confusion" | Confusion from diverging models is information — it shows where understanding is incomplete. | | [D] "We consulted multiple advisors" | If all advisors share the same disciplinary lens, that is triangulation within one model. | | [D] "The model has worked before" | A model that predicted correctly in past contexts may be in a regime where its assumptions no longer hold. | | [D] "Convergence is confirmation bias with extra steps" | Confirmation bias seeks evidence for a pre-held view. Latticework compares independent predictions — divergence check is the anti-bias mechanism. | | → Add [O] entries here after each real use — paste the actual failure pattern | What went wrong and why |

Red Flags

Only one discipline's vocabulary used throughout
Stakeholder objections dismissed without checking if they represent another model's prediction
Decision called "rigorous" because the single model was applied thoroughly
Diverging data forced into the primary model instead of triggering a model-check
The analysis cannot name its own blind spots

Verification

[ ] ≥3 genuinely independent disciplinary lenses applied
[ ] Each lens produced an explicit, falsifiable prediction (not just "we considered X")
[ ] Convergence zones marked higher-confidence; divergence zones named as live unknowns
[ ] At least one blind spot named (phenomenon no lens covers)
[ ] Lollapalooza check: any convergent forces multiplicatively aligned?

Part of deciqAI Knowledge Skills — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.