1-Back reinforcement matching and mismatching by pigeons: Implicit or explicit learning?

Area of Science:

• Comparative psychology
• Animal cognition
• Learning theory

Background:

• Human learning distinguishes between implicit (automatic) and explicit (conscious rule-based) processes.
• Assessing explicit learning in non-verbal animals is challenging due to the inability to report rules.
• Previous indirect methods, like comparing learning speeds or using delayed reinforcement tasks, have yielded debated results.

Purpose of the Study:

• To investigate whether non-verbal animals, specifically pigeons, can exhibit implicit learning in tasks previously thought to require explicit learning.
• To challenge the prevailing assumptions about the necessity of conscious rule reporting for explicit learning in animals.
• To propose alternative, simpler explanations for observed learning in non-verbal species.

Main Methods:

• Utilized the 1-back-reinforcement conditional discrimination task with delayed feedback.
• Tested pigeons on both 1-back reinforcement matching and 1-back reinforcement mismatching tasks.
• Compared pigeon performance to existing findings in monkeys and theoretical models of implicit/explicit learning.

Main Results:

• Pigeons demonstrated significant learning in both 1-back reinforcement matching and mismatching tasks.
• This learning occurred despite the delayed feedback, a condition argued to preclude implicit learning.
• Pigeons did not show the learning speed differences observed in humans between explicit and implicit tasks.

Conclusions:

• The findings suggest that pigeons can learn through implicit mechanisms even in tasks with delayed reinforcement.
• The study provides evidence against the necessity of conscious rule reporting for certain types of learning in non-verbal animals.
• Existing criteria for inferring explicit learning in animals may be flawed, potentially overestimating cognitive complexity.