Reading focus
Syllabus: How Do We Learn What Predicts What?
Compare elemental and configural predictions for compounds, including the problems and extensions Pearce discusses. Relate generalization assumptions to the earlier blocking debate.
Why you’re reading it. Identify how changing a representation can change predictions even within the broad family of associative explanations.
A question to consider. In a blocking design, how could incomplete generalization from A to AX, and from AX to X, change the expected test result? What measurement could constrain these assumptions before observing that result?
Why this reading
What a learner represents determines what a learning rule predicts. Pearce develops an account in which a compound is represented as a configuration and uses it to reconsider effects often explained through associations with separate cues.
Before you read
An elemental account allows components to acquire associations; a configural account associates a representation of the overall pattern with an outcome. Generalization transfers responding between similar training and test patterns. A connectionist network here is a formal explanatory model, not a literal map of identified neurons.
Read for
- Compare the network diagrams and their assumptions about a cue presented alone versus within a compound.
- Track how generalization affects blocking, summation, and discrimination, rather than assuming that a test cue is psychologically unchanged by its context.
- Examine why the later extensions are needed and which findings remain difficult for configural or elemental accounts.
Consider afterwards
- A and B each predict an outcome when presented alone. Why need presenting AB together produce neither twice the response nor the same response? Identify a representational assumption that changes the prediction.
Editorial context
This develops Pearce’s earlier configural theory while retaining problems and competing explanations. The conclusion acknowledges difficult summation findings and increasingly flexible elemental accounts. Adding representational flexibility can explain more patterns, but also increases the need to specify independently when a proposed representation should occur. No single configuration model is established as the final account.