Transfer of a serial representation between two distinct tasks by rhesus macaques

Area of Science:

• Cognitive science
• Animal behavior
• Neuroscience

Background:

• Investigating whether animals form task-specific or generalizable representations is crucial for understanding cognitive flexibility.
• Prior research suggests varying degrees of transferability between different learning paradigms.

Purpose of the Study:

• To determine if rhesus monkeys' knowledge of stimulus list orderings is task-specific or general.
• To assess if experience in one task enhances performance in a qualitatively different task.

Main Methods:

• Rhesus monkeys learned stimulus list orderings using two distinct paradigms: simultaneous chaining and transitive inference.
• Knowledge acquired in one paradigm was tested using the other to evaluate transfer of learning.
• Performance was analyzed for enhancements and error patterns indicative of underlying representations.

Main Results:

• Performance was enhanced from the initial stages of transfer training between tasks.
• Transitive inference performance exhibited symbolic distance effects, correlating with response accuracy.
• Error patterns in both tasks strongly supported a spatially coded representation of list items.

Conclusions:

• Animal cognitive representations of learned sequences are not strictly task-specific but possess a generalizable form.
• A spatially coded representation underlies performance across different serial learning tasks in rhesus monkeys.
• This spatial representation facilitates flexible cognitive transfer, enhancing performance in novel but related tasks.