Pavlov’s Studies

Even before beginning this course, you might have heard about Pavlov’s dogs. The work

of the Russian physiologist Ivan Pavlov is well known. Still, it is easy to take its true

significance for granted. Importantly, Pavlov demonstrated that neutral aspects of the

environment can attain the capacity to evoke responses through pairing with other stimuli

and that bodily processes can be influenced by environmental cues.

In the early 1900s, Pavlov was interested in the way the body digests food. In his

experiments, he routinely placed meat powder in a dog’s mouth, causing the dog to

salivate. By accident, Pavlov noticed that the meat powder was not the only stimulus that

caused the dog to drool. The dog salivated in response to a number of stimuli associated

with the food, such as the sight of the food dish, the sight of the individual who brought

the food into the room, and the sound of the door closing when the food arrived. Pavlov

recognized that the dog’s association of these sights and sounds with the food was an

important type of learning, which came to be called classical conditioning.

Pavlov wanted to know why the dog salivated in reaction to various sights and

sounds before eating the meat powder. He observed that the dog’s behavior included

both unlearned and learned components. The unlearned part of classical conditioning

is based on the fact that some stimuli automatically produce certain responses apart

from any prior learning; in other words, they are inborn (innate). Reflexes are such

automatic stimulus–response connections. They include salivation in response to food,

nausea in response to spoiled food, shivering in response to low temperature, coughing

in response to throat congestion, pupil constriction in response to light, and withdrawal

in response to pain.

An unconditioned stimulus (US) is a stimulus that produces a response without prior

learning; food was the US in Pavlov’s experiments. An unconditioned response (UR)

is an unlearned reaction that is automatically elicited by the US. Unconditioned responses

are involuntary; they happen in response to a stimulus without conscious effort. In

Pavlov’s experiment, drooling in response to food was the UR. In the case of Bob and

the flushing toilet, Bob’s learning and experience did not cause him to shriek when the

hot water hit his body. His cry of pain was unlearned and occurred automatically. The

hot water was the US, and Bob’s panic was the UR.

In classical conditioning, a conditioned stimulus (CS) is a previously neutral stimulus

that eventually elicits a conditioned response after being paired with the unconditioned

stimulus. The conditioned response (CR) is the learned response to the conditioned

● unconditioned stimulus (US) A stimulus that

produces a response without prior learning.

● unconditioned response (UR) An

unlearned reaction that is automatically

elicited by the unconditioned stimulus.

● conditioned stimulus (CS) A previously

neutral stimulus that eventually elicits a

conditioned response after being paired

with the unconditioned stimulus.

● conditioned response (CR) The learned

response to the conditioned stimulus that

occurs after conditioned stimulus–

unconditioned stimulus pairing.

Pavlov (the white-bearded gentleman in the center) is shown demonstrating the nature of

classical conditioning to students at the Military Medical Academy in Russia.

c Universal Images Group/Getty Images

stimulus that occurs after CS–US pairing (Pavlov, 1927). Sometimes conditioned responses

are quite similar to unconditioned responses, but typically they are not as strong.

In studying a dog’s response to various stimuli associated with meat powder, Pavlov

rang a bell before giving meat powder to the dog. Until then, ringing the bell did not

have a particular effect on the dog, except perhaps to wake the dog from a nap. The bell

was a neutral stimulus, meaning that in the dog’s world, this stimulus did not have any

signal value at all. Prior to being paired with the meat powder, the bell was meaningless.

However, the dog began to associate the sound of the bell with the food and salivated

when it heard the bell. The bell had become a conditioned (learned) stimulus (CS), and

salivation was now a conditioned response (CR). In the case of Bob’s interrupted shower,

the sound of the toilet flushing was the CS, and panicking was the CR after the scalding

water (US) and the flushing sound (CS) were paired. Figure 2 summarizes how classical

conditioning works.

Researchers have shown that salivation can be used as a conditioned response not only

in dogs and humans but also in, of all things, cockroaches (Nishino & others, 2015;

Watanabe & Mizunami, 2007). These researchers paired the smell of peppermint (the CS,

which was applied to the cockroaches’ antennae) with sugary water (the US). Cockroaches

naturally salivate (the UR) in response to sugary foods, and after repeated pairings between

peppermint smell and sugary water, the cockroaches salivated in response to the smell of

peppermint (the CR). When they collected and measured the cockroach saliva, the

researchers found that the cockroaches had slobbered over that smell for 2 minutes.

learning that occurs without awareness or effort, based on the presentation of two stimuli

together. For this pairing to work, however, two important factors must be present:

contiguity and contingency.

Contiguity simply means that the CS and US are presented very close together in

time—even a mere fraction of a second (Gottlieb & Begej, 2014). In Pavlov’s work, if

the bell had rung 20 minutes before the presentation of the food, the dog probably would

not have associated the bell with the food. However, pairing the CS and US close together

in time is not all that is needed for conditioning to occur. Imagine that the bell not only

rings just before the food is delivered, but it also rings many times when the food is not

on its way. In such a situation, the dog would not associate the bell with the food, and

no learning would occur. Why? Because the bell does not serve as a signal for the food.

Contingency means that the CS must not only precede the US closely in time, but it

must serve as a reliable indicator that the US is on its way (Rescorla, 1966, 1988, 2009).

To get a sense of the importance of contingency, imagine that the dog in Pavlov’s experiment

is exposed to a ringing bell at random times all day long. Whenever the dog receives

food, the delivery of the food always immediately follows a bell ring. However, in this

situation, the dog will not associate the bell with the food, because the bell is not a reliable

signal that food is coming: It rings a lot when no food is on the way. Whereas contiguity

refers to the fact that the CS and US occur close together in time, contingency refers to

the information value of the CS relative to the US. When contingency is present, the CS

provides a systematic signal that the US is on its way (Kringelbach & Berridge, 2015).


Recall from early in the chapter that classical conditioning and operant conditioning are

forms of associative learning, which involves learning that two events are connected. In

classical conditioning, organisms learn the association between two stimuli (US and CS).

Classical conditioning is a form of respondent behavior, behavior that occurs in automatic

response to a stimulus such as a nausea-producing drug and later to a conditioned

stimulus such as sweet water that was paired with the drug. Calling a behavior “respondent”

means that it happens on auto pilot.

Classical conditioning explains how neutral stimuli become associated with unlearned,

involuntary responses. Classical conditioning is not as effective, however, in explaining

voluntary behaviors such as a student’s studying hard for a test, a gambler’s playing slot

machines in Las Vegas, or a service dog fetching his owner’s cell phone on command.

Operant conditioning is usually much better than classical conditioning at explaining such

voluntary behaviors. Whereas classical conditioning focuses on the association between

stimuli, operant conditioning focuses on the association between behaviors and the stimuli

that follow them.

Defining Operant Conditioning

Operant conditioning or instrumental conditioning is a form of associative learning

in which the consequences of a behavior change the probability of the behavior’s occurrence.

The American psychologist B. F. Skinner (1938) chose the term operant to

describe the behavior of the organism. An operant behavior occurs spontaneously.

According to Skinner, the consequences that follow such spontaneous behaviors determine

whether the behavior will be repeated.

Imagine, for example, that you spontaneously decide to take a different route while driving

to campus one day. You are more likely to repeat that route on another day if you have

a pleasant experience—for instance, arriving at school faster or finding a new coffee place

to try—than if you have a lousy experience such as getting stuck in traffic. In either case,

the consequences of your spontaneous act influence whether that behavior happens again.

Recall that contingency is an important aspect of classical conditioning in which the

occurrence of one stimulus can be predicted from the presence of another one. Contingency

also plays a key role in operant conditioning. For example, when a rat pushes a

lever (behavior) that delivers food, the delivery of food (consequence) is contingent on

that behavior. This principle of contingency helps explain why passersby should never

praise, pet, or feed a service dog while he is working (at least without asking first).

Providing rewards during such times might interfere with the dog’s training.

Thorndike’s Law of Effect

Although Skinner emerged as the primary figure in operant conditioning, the experiments

of E. L. Thorndike (1898) established the power of consequences in determining voluntary

behavior. At about the same time that Pavlov was conducting classical conditioning

experiments with salivating dogs, Thorndike, another American psychologist, was studying

cats in puzzle boxes. Thorndike put a hungry cat inside a box and placed a piece of

fish outside. To escape from the box and obtain the food, the cat had to learn to open

the latch inside the box. At first the cat made a number of ineffective responses. It clawed

or bit at the bars and thrust its paw through the openings. Eventually the cat accidentally

stepped on the lever that released the door bolt. When the cat returned to the box, it

went through the same random activity until it stepped on the lever once more. On subsequent

trials, the cat made fewer and fewer random movements until finally it immediately

stepped on the lever to open the door (Figure 4). Thorndike’s resulting law of effect

states that behaviors followed by pleasant outcomes are strengthened and that behaviors

followed by unpleasant outcomes are weakened.

The law of effect is profoundly important because it presents the basic idea that the

consequences of a behavior influence the likelihood of that behavior’s recurrence. Quite

simply, a behavior can be followed by something good or something bad, and the probability

of a behavior’s being repeated depends on these outcomes. As we now explore,

Skinner’s operant conditioning model expands on this basic idea.

Skinner’s Approach to Operant Conditioning

Skinner believed that the mechanisms of learning are the same for all species. This

conviction led him to study animals in the hope that he could discover the components

● operant conditioning or instrumental

conditioning A form of associative learning

in which the consequences of a behavior

change the probability of the behavior’s


● law of effect Thorndike’s law stating that

behaviors followed by positive outcomes are

strengthened and that behaviors followed by

negative outcomes are weakened

of learning with organisms simpler than humans, including pigeons. During World War

II, Skinner trained pigeons to pilot missiles. Naval officials just could not accept pigeons

guiding their missiles in a war, but Skinner congratulated himself on the degree of control

he was able to exercise over the pigeons (Figure 5).

Skinner and other behaviorists made every effort to study organisms under precisely

controlled conditions so that they could examine the connection between the operant

behavior and the specific consequences in minute detail. In the 1930s, Skinner created

an operant conditioning chamber, also called a Skinner box, to control experimental

conditions (Figure 6).

A device in the box delivered food pellets into a tray at random. After

a rat became accustomed to the box, Skinner installed a lever and observed

the rat’s behavior. As the hungry rat explored the box, it occasionally

pressed the lever, and a food pellet was dispensed. Soon the rat learned

that the consequences of pressing the lever were positive: It would be fed.

Skinner achieved further control by soundproofing the box to ensure that

the experimenter was the only influence on the organism. In many of the

experiments, the responses were mechanically recorded, and the food (the

consequence) was dispensed automatically. These precautions were aimed

at preventing human error.