Marr's Levels of Analysis
Updated 2026-08-01
INTRODUCTION
English translation pending.
CORE DEFINITION
Proposed by neuroscientist David Marr in Vision. Understanding any information-processing system requires three levels: the computational level, which specifies what problem the system solves and why; the algorithmic and representational level, which specifies how inputs are transformed into outputs; and the implementation level, which specifies the physical hardware or neural substrate. The levels can be studied independently but constrain one another, and confusing them produces category errors.
SCAFFOLDING EFFECT
Reduce cognitive load
- Start at the top: state what problem the system solves before discussing how to build it - Separate the layers: label each claim as computational, algorithmic, or implementational before arguing - Diagnose confusion: check whether a disagreement is really about goals rather than about mechanisms
Anchor fast decisions
A system can be realized in many different substrates, so implementation details do not determine what the system computes. Conversely, a computational goal constrains which algorithms can satisfy it. Separating the levels prevents explanations at one level from being used to answer questions at another, which is the error behind claims that locating a brain region has explained a cognitive function.
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/Level_of_analysisverified
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