|
|
|
|
|
ABSTRACT:
Sensory preconditioning, blocking, overshadowing, second-order reinforcement, these are all terms that were coined by learning psychology more than 20 years ago and are well established using classical conditioning setups. Classical conditioning is often described as the transfer of the response-eliciting property of a stimulus to a new stimulus without that property. This association between an unconditioned stimulus (US) and a conditioned stimulus (CS) can also be established when the animal is in control of the stimulus presentation (operant conditioning). We have not found a single report as to whether the abovementioned concepts also apply for operant conditioning. We used operant pattern and color learning in the Drosophila flight simulator (see figure 1) to ask five basic questions: (1) US processing: How do variations in US strength translate into associative strength? (2) CS Processing: How do different CS’s translate into associative strength? (3) Blocking: Is it possible to systematically prevent a reinforced CS from being learned? (4) Second-order conditioning: Can a well trained CS act as a US? (5) Sensory preconditioning: Can a CS be learned without reinforcement?
(1) Increasing the amount of reinforcement from a level already sufficient to induce a significant learning response (thus assessing US processing), resulted in an increase in learning performance. With increasing strength of the reinforcer, the flies kept the ratio between avoidance and subsequent learning score constant. (2) Training the flies to a compound CS1CS2 (CS1: pattern, CS2: color. Both single CS’s are learned equally well when trained without the other) results in similar learning scores for the single CS’s in a subsequent test of patterns or colors alone (no overshadowing in compound CS processing). (3) Most error-correcting learning rules predict that pre-training one of the CS’s before a CS1CS2 compound training prevents subsequent learning of the second CS during compound training even though it receives the proper reinforcement. This blocking effect is known from most classical conditioning preparations. Our learning scores, however, were indistinguishable from the control CS processing experiment. This is at odds with all current learning theories. (4) This result can not be attributed to second-order conditioning with the pre-trained CS as second-order reinforcer since this effect is negligible in our setup. (5) Pre-exposure of a compound CS1CS2 without reinforcement before training CS2 (color) leads to a significant avoidance in a subsequent test of patterns (CS1) alone even though this CS has never been reinforced (sensory preconditioning).
|
|
|
|
STATISTICS
|
|
Click on # to view
|
|
Citations
|
|
0
|
|
References
|
|
0
|
|
Comments
|
|
0
|
|
Quality
|
|
0/0.00
|
|
Interest
|
|
0/0.00
|
|
View(er)s
|
|
2/280
|
|
|
|
|
|
|
| Prev |
Next |
|
