Learning in rich networks involves both positive and negative associations.

Area of Science:

• Cognitive Science
• Animal Behavior
• Developmental Psychology

Background:

• Adaptive behaviors are shaped by positive and negative associative learning.
• Limited understanding exists on how these processes form complex associative networks for cognition, like categorization.
• The role of negative associative learning in development (experience-dependent vs. experience-expectant) is unclear.

Purpose of the Study:

• To investigate the role of negative associative learning in the formation of rich associative networks.
• To examine how pigeons learn object-token associations in a paradigm analogous to human word learning.
• To determine if negative associative learning is experience-dependent or experience-expectant.

Main Methods:

• Pigeons were trained on a many-to-many object-token mapping task (16 objects to 16 tokens).
• The availability of incorrect token options (foils) was manipulated to control the degree of negative associative learning.
• Performance was assessed using trials with novel vs. experienced foils and when the correct token was withheld.

Main Results:

• Accuracy was lower with novel foils compared to previously experienced foils.
• When the correct token was withheld, pigeons preferred novel foils over experienced foils.
• These effects emerged only after initial positive associations were established.

Conclusions:

• Learning rich associative networks relies on both positive and negative associative learning.
• Negative associative learning appears to involve acquiring specific, experience-dependent content.
• Findings have significant implications for basic learning theories and developmental psychology in both animals and humans.