Are faces processed like words? A diagnostic test for recognition by parts

Area of Science:

• Cognitive psychology
• Neuroscience
• Visual perception

Background:

• The recognition of objects, such as faces and words, has been debated, with theories suggesting holistic (whole) or part-based processing.
• Previous research on crowding focused on interference between separate objects, not within an object's components.

Purpose of the Study:

• To investigate whether object recognition relies on processing the object as a whole or by its individual parts.
• To determine if crowding occurs between parts of an object and its impact on perception and neural activation.

Main Methods:

• A crowding test was applied to word and face recognition, analyzing critical spacing for parts.
• Familiarity effects (inversion, superiority effects) were manipulated to assess crowding test robustness.
• fMRI (functional Magnetic Resonance Imaging) was used to measure face-area activation.

Main Results:

• Internal crowding, where object parts interfere with each other, significantly impairs perception and identification.
• Critical spacing for recognition is proportional to distance from fixation and independent of object size or type.
• Both words and faces are recognized by parts, with their components (letters, features) processed holistically.

Conclusions:

• Internal crowding serves as a signature for part-based object recognition.
• The processing of words and faces is similar, influenced by both familiarity and crowding.
• Object recognition is determined by whether parts can be isolated within a critical spacing.