Wagner's "SOP" Model

Question 1

what phenomena does SOP fix?

Answer 1

• one-trial overshadowing (e.g. James and Wagner, 1980)
• Latent inhibition and its “content specificity”

Question 2

what phenomena does SOP fail at?

Answer 2

• downshift unblocking (e.g., Dickinson, Hall and Mackintosh, 1976)
  inhibition:
• extinction isn’t learning
• spontaneous recovery

Question 3

what model can Resorla and Wagner explain that SOP can’t?

Answer 3

I have never been able to think of any learning phenomena that SOP can’t explain that Rescorla & Wagner can explain. (Jasper-lecturer)

Question 4

what does SOP stand for?

Answer 4

as in common usage it stands for “Sometimes Opponent Processes
“

Question 5

what does SOP assume?

Answer 5

SOP assumes that events (including CSs and UCSs) are not represented like intact photographs or recordings

Question 6

what did Hebb (1949) suggest?

Answer 6

what fires together, wires together

Question 7

what are the issues with Hebb’s suggestions?

Answer 7

• Hebb cannot explain any of the things that RMW can
• it also predicts that learning will be ceaseless; there is no way to reach λ

Question 8

what is wagner’s solution to the Hebb/Konorski problem?

Answer 8

• Wagner’s SOP can be thought of as Hebb’s model with two different activity states
• A1, A2(primed), I

Question 9

what dod Jerzy Konorski do?

Answer 9

• made the same suggestion as Hebb but one year earlier
• however argued that it could not be true because it would wrongly predict that forward (cs → UCS) and backward (UCS → CS) conditioning would be similar

Question 10

what is the capacity of state A1?

Answer 10

limited
- helps to explain one-trial overshadowing
- when 2 conditioned stimulus’s (tome and light) are paired, each one is less active then they would be alone
- this means less will be learned about them
- the association between light and the unconditioned stimulus is worse than when light accompanied the tone than if the tone is on its pwn
- so limited capacity feature of A1 activity created overshadowing of the tone

Question 11

what system did wagner propose instead of the on-on → extinction?

Answer 11

• Wagner referred to a stimulus’ elements being in their A2 states as their ‘primed’ and likened it to the ‘refractory period’ in nervous transmission
• this gives it all the RMW features it needs

Question 12

what is the representational activity cycle (and “self-generated priming”)

Answer 12

• An inactive stimulus’ elements (I), become fully active (A1) when is presented, and fade (A2), eventually returning to I

Question 13

what is self generated prining?

Answer 13

when a stimulus goes back to its normal state (like a refractory period)

Question 14

what is retrieval-generated priming?

Answer 14

• an associatively-activated representation goes into from inactive, but only to its A2 state
• it cannot immediately enter A1 without passing back through inactive.
• This takes some time (Wagner likened this to the refractory period in nervous transmission) without this feature, this model would not work
• at λ the presentation of the CS will “prime” most (but usually not all) of the UCS’s representational elements into their A2 states.—This is what generates the CR in Pavlovian conditioning.

Question 15

How does SOP explain blocking?

Answer 15

• During stage-2, the tone primes many of the shocks elements into their A2 states.
• This means that the blocking test stimulus, the light won’t be learned about normally.
• This is because lights’ elements are in their A1 states while a lot of the shock’s elements are in their A2 states.