2. Pavlov, Watson, Rescorla & Wagner, Wagner, and Miller all had different theories of Classical Conditioning.

            (1) Briefly describe each theory, and what makes it different from the one before it.

            (2) Tell me why each theory can or cannot handle the following findings:

                        (a) release from latent inhibition

                        (b) UCS pre-exposure effect

                        (c) blocking

Point 1: “Watson” in the question above was a type (it was supposed to be “Hull” so I simply ignored anything having to do with Watson.

(1). Brief descriptions of the theories:

Pavlov: With temporal contiguity, a CS links directly to a UCS center in the brain, and triggers it (causing the UCR to go off). Also, if you have more than one CS, each forms its own separate link to the UCS.

RW: They look at signal value and contingency (predictability) rather than contiguity: Items presented more often or with higher salience capture most of the association (so a CS can compete with another CS for lambda). Unlike Pavlov, RW predicts context effects like blocking, OS, etc. Also, they claim inhibition is the opposite of excitation (one single association that can be positive or negative, but not both), unlike Pavlov, who has both positive and negative associations between a CS and a UCS.

W: For excitation, CS and UCS need to be rehearsed together so that later presentation of the CS will prime the memory that the UCS is coming next. As a CS or UCS becomes more familiar, it gets rehearsed less often. Can handle most of the findings RW does, but also the CS salience effects that RW falls apart on. Inhibition occurs when a CS is rehearsed alone.

Miller: all associations form regardless, but they’re compared to determine which will control the CR (typically, the more frequently presented or more distinct association). Claims most of what the other theories view as a learning deficit is really a response deficit. Also, claims there is no real inhibition.

(2). How they handle release from LI, UCS pre-exposure, and blocking.

Release from LI: P and RW don’t predict LI (CS pre-exposure effect), because presenting CS by itself initially should be irrelevant. So, can’t predict release. W handles LI (CS is familiar, so no longer rehearsed with UCS in Phase 2), but claims it is a learning deficit (not enough rehearsals means weaker learning). So, without new learning there can’t be release from LI. Only Miller handles LI: Weaken the context/comparator association, and the CS-UCS association now wins the comparison. (LI due to a strong CS-context link due to pre-exposing CS).

UCS pre-exposure: P can’t handle this, since UCS by itself will form no associations, so there should be no effect on Phase 2, where CS is introduced. In RW, this is handled as blocking: Context acts as CS1, captures a lot of lambda, so later, when CS2 is introduced, there’s little association left for it to capture (Learning deficit). W handles it by familiarity: UCS becomes familiar, so it isn’t rehearsed much together with the CS in Phase 2 (also learning deficit). In Miller, UCS-context link is more frequent, so the indirect path wins the comparison.

Blocking: P can’t handle, since CS1 and CS2 should connect independently of one another. In RW, the first stimulus captures most of the association, leaving little for the second to capture when it’s finally introduced (learning deficit). Wagner: UCS and CS1 are familiar from Phase 1, so UCS isn’t rehearsed much anymore once CS2 is introduced (learning deficit). Miller: UCS and CS1 form a more frequent association than UCS and CS2, so the latter loses the comparison (response deficit).