Ladder of Causation
Updated 2026-08-15
INTRODUCTION
English translation pending.
CORE DEFINITION
The ladder of causation proposed by Judea Pearl has three rungs: association, seeing that A and B move together; intervention, asking whether B changes if A is changed; and counterfactual, imagining what B would have been had A not occurred. Each higher rung requires a richer causal model and yields a stronger kind of understanding, and most machine learning today still sits on the first rung.
SCAFFOLDING EFFECT
Reduce cognitive load
- Grade the question: locate any causal inquiry on rung one, two, or three - Match evidence to rung: statistics for association, experiments for intervention, structural models for counterfactuals - Expose the shallow claim: reject any associational answer given to an intervention question
Anchor fast decisions
The three rungs ask different questions, and each can only be answered with its own kind of evidence. Association needs joint observation, intervention needs a way to change the world or faithfully emulate the change, and counterfactuals need a model of how the world is generated. Climbing the ladder buys explanatory and planning power at the price of stronger assumptions.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- en.wikipedia.orghttps://en.wikipedia.org/wiki/Causal_modelverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS