Conditioning

The core of Pavlovian conditioning is the pairing (association) of stimuli
to elicit responses. Food (meat powder) placed in a dog’s mouth naturally
produces salivation. Pavlov called the food an unconditioned stimulus (US)
and salivation, elicited by the food, the unconditioned response (UR). When
a neutral stimulus—for example, a tone that does not naturally elicit salivation—
is repeatedly followed by presentation of food, the tone alone eventually
evokes salivation. Pavlov labeled the tone a conditioned stimulus (CS)
and the response (salivation) elicited by it the conditioned response (CR).

Pavlov’s formulation can be summarized as follows:
Before conditioning:
Food (US) elicits Salivation (UR)
Conditioning procedure:
Neutral Stimulus (Tone) plus Food (US) elicits Salivation (UR)
After conditioning:
Tone (CS) elicits Salivation (CR)
Pavlov believed that conditioned responses were identical to unconditioned
responses. That is usually not the case. For example, conditioned responses
may be less pronounced (weaker) or a bit more lethargic than unconditioned
responses.
Several phenomena turn up in studies of Pavlovian conditioning. Extinction,
generalization, and discrimination are among the most important. Extinction
refers to the procedure as well the elimination of a CR. If the CS is
repeatedly presented without the US, extinction occurs: The dog stops salivating
to the tone. During the course of extinction, the CR may return from
time to time until it is finally extinguished. Pavlov called the occasional return
of the CR “spontaneous recovery.”

Stimulus generalization refers to responding not only to a particular CS
but also to different but similar stimuli. Further, the magnitude (amount of
salivation) of a generalized response tends to decline as stimuli become less
and less like the CS. For example, a dog trained to salivate to a 5,000-cycleper-
second (cps) tone is likely to salivate also to 5,300 cps and 4,700 cps
tones without specific training to do so (stimulus generalization). Responses
tend to weaken in an orderly way as tones become more and more unlike the
CS. As the tones move away from the CS in both directions, say, to 4,400 cps
from 4,100 cps, and 5,600 cps to 5,900 cps, the flow of salivation becomes
less and less.

Stimulus generalization in effect extends the number of stimuli that elicit
a conditioned response. Discrimination procedures restrict that number by
conditioning a subject not to generalize across stimuli. The procedure involves
two processes: acquisition and extinction. The CS is paired repeatedly
with the US (acquisition) while the US is withheld as generalized stimuli are
presented repeatedly (extinction). If the dog now salivates to the CS and not
to the generalized stimuli, the dog has learned to discriminate or to act discriminatingly.

Pavlov reported that some dogs displayed a general breakdown
in behavior patterns (experimental neurosis) when called upon to
make discriminations that were too difficult.
Pavlov’s work on what he called the second-signal system implies that conditioning
principles are relevant to human as well as to animal learning.
Once, say, a tone is established as a CS in first-order conditioning, the tone
can be paired with a neutral stimulus to establish a second-order CS. Thus,
in the absence of food, a light might precede the tone (CS) several times until
the light itself begins to function as a CS. Second-order conditioning appears
to follow many of the same rules as first-order conditioning.
Pavlov’s work has clearly provided one way to study the learning process
in great detail. It has also provided the kind of data and theory that have affected
research in other areas of learning, such as instrumental conditioning
and, subsequently, cognitive science and neuroscience.