Feature Integration Theory
Updated 2026-08-01
INTRODUCTION
English translation pending.
CORE DEFINITION
Feature integration theory was proposed by Anne Treisman and Garry Gelade. It posits two stages: a preattentive stage that registers elementary features such as color, orientation, and size in parallel and independently, and a focused attention stage that binds features at the same location into a unified object representation. The theory explains illusory conjunctions, in which unattended displays produce objects with the wrong feature combinations. The binding problem is the core puzzle: nothing in the separate feature maps specifies that a particular color belongs with a particular shape.
SCAFFOLDING EFFECT
Reduce cognitive load
- Split the display: separate which features are available preattentively from which require binding. - Budget attention: give viewers enough focused attention to bind correctly. - Prevent misbinding: reduce competing features so illusory conjunctions cannot form.
Anchor fast decisions
Feature maps encode color, orientation, and motion separately, with no record of which features co-occur. Object perception requires exactly that co-occurrence, which no single map supplies. Attention provides it by selecting a location and linking whatever features are registered there into one representation. This is why perceiving a coherent object costs attentional resources, and why those resources are the scarce factor. Under time pressure or divided attention the linking step is skipped or misapplied, so features get bound to the wrong locations.
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/Feature_integration_theoryverified
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