# Lecture 9: Learning: Classical Conditioning
## Introductory Psychology
---
## Learning Objectives
By the end of this lecture, students will be able to:
1. Define learning and distinguish between associative and non-associative learning
2. Describe Pavlov's experiments and the key elements of classical conditioning
3. Explain acquisition, extinction, spontaneous recovery, generalization, and discrimination
4. Discuss higher-order conditioning and the role of cognitive processes in conditioning
5. Apply classical conditioning principles to real-world phenomena including phobias, taste aversions, and advertising
---
## Lecture Content
### I. What Is Learning?
Learning is defined as a relatively permanent change in behavior or knowledge that results from experience. This definition distinguishes learned changes from those produced by maturation, fatigue, or temporary physiological states. At the simplest level, non-associative learning takes two forms: habituation, in which the response to a repeated, harmless stimulus diminishes over time, and sensitization, in which exposure to a strong or threatening stimulus produces an increased response. Associative learning involves forming connections between events that occur together and comes in two major varieties: classical conditioning, in which an organism learns to associate two stimuli, and operant conditioning, in which an organism learns to associate a behavior with its consequences. Observational learning, a third major form, involves learning by watching and imitating others.
### II. Pavlov's Discovery
Ivan Pavlov (1849-1936) was a Russian physiologist studying digestion in dogs when he noticed something unexpected: the dogs began salivating before food was presented — in response to the sight of the food dish or the approaching lab assistant. Pavlov termed these anticipatory responses "psychic secretions" and redirected his research program to investigate them systematically.
In his experimental procedure, Pavlov first established that a neutral stimulus (a bell) produced no salivation, while food (an unconditioned stimulus) naturally produced salivation (an unconditioned response). He then repeatedly paired the bell with the food. After sufficient pairings, the bell alone — now a conditioned stimulus — was able to elicit salivation, which was now termed a conditioned response.
### III. Key Elements of Classical Conditioning
The unconditioned stimulus (US) is any stimulus that naturally and automatically triggers a response without prior learning — food, a loud noise, or a puff of air to the eye, for example. The unconditioned response (UR) is the unlearned, natural reaction to that stimulus — salivation, a startle reflex, or an eye blink. The conditioned stimulus (CS) is a previously neutral stimulus that, after being paired with the US, comes to trigger a learned response. The conditioned response (CR) is that learned response. Although the CR is often similar to the UR, it may differ in magnitude or in its precise form.
### IV. Processes in Classical Conditioning
Acquisition is the initial learning phase during which the CS-US association is established. Conditioning is strongest when the CS precedes the US by a brief interval (forward conditioning, typically about half a second). Simultaneous conditioning, in which the CS and US occur at the same time, produces weaker learning, and backward conditioning, in which the US precedes the CS, is generally ineffective. While temporal contiguity — closeness in time — is important, it is not by itself sufficient for conditioning to occur.
Extinction is the gradual weakening of the conditioned response when the CS is repeatedly presented without the US. Importantly, the original association is not erased but rather suppressed or inhibited. Evidence for this comes from spontaneous recovery, the reappearance of an extinguished CR after a rest period. Reconditioning — re-establishing the CS-US association — typically proceeds faster than the original acquisition, further confirming that the initial learning was never fully lost.
Stimulus generalization is the tendency to respond to stimuli that are similar to the original CS. A dog conditioned to salivate to a 1,000 Hz tone may also salivate to tones of 900 Hz or 1,100 Hz, with the strength of the response diminishing as the test stimulus becomes less similar to the original — a pattern described by the generalization gradient. Stimulus discrimination is the complementary ability to distinguish between the CS and similar stimuli, typically trained by pairing one stimulus with the US (CS+) while presenting another without the US (CS-).
A four-panel diagram illustrating classical conditioning processes. Panel A: Acquisition — a graph showing conditioned response strength increasing over successive CS-US pairings, with an asymptote curve. Panel B: Extinction — the same graph continuing to show response strength declining when CS is presented alone. Panel C: Spontaneous recovery — after a rest period (gap in the graph), the CR partially reappears, then extinguishes more rapidly. Panel D: Generalization gradient — a bell curve showing response strength on the y-axis and stimulus similarity on the x-axis, with the peak at the original CS and declining responses to increasingly dissimilar stimuli.
### V. Cognitive Processes in Classical Conditioning
Classical conditioning is far more than a mechanical stimulus-response association. Robert Rescorla demonstrated in the 1960s and 1970s that conditioning depends on contingency — the degree to which the CS reliably predicts the US — rather than mere contiguity (co-occurrence). If the US occurs equally often with and without the CS, little conditioning takes place because the CS provides no useful information.
The Rescorla-Wagner model formalizes this insight by proposing that learning occurs when outcomes are unexpected. Conditioning is strongest when the US is surprising; as the CS becomes a reliable predictor, additional learning slows and approaches an asymptote. This model also explains blocking, discovered by Kamin in 1969: if one CS already predicts the US, a second CS paired alongside it will not become conditioned, because the US is no longer surprising. Mere pairing is therefore insufficient; what matters is the informativeness of the CS. Latent inhibition adds another wrinkle: prior unreinforced exposure to a stimulus makes it harder to condition as a CS later, presumably because the organism has already learned that the stimulus predicts nothing.
### VI. Biological Constraints on Classical Conditioning
Garcia and Koelling's landmark 1966 study on taste aversion demonstrated that not all associations are equally easy to learn. Rats readily associated tastes with nausea but not with electric shock, and they readily associated lights and sounds with shock but not with nausea. This finding illustrated biological preparedness — the idea that organisms are evolutionarily "prepared" to form certain associations more readily than others.
Taste aversion learning has several remarkable properties. It can be acquired in a single trial and can occur even when hours separate the taste from the onset of illness, violating the usual requirement for close temporal contiguity. These features make adaptive sense: the ability to rapidly learn to avoid poisonous foods confers an obvious survival advantage. The concept of instinctive drift, in which animals revert to biologically predisposed behaviors during conditioning, further illustrates the biological constraints on learning.
### VII. Applications of Classical Conditioning
Emotional conditioning is powerfully demonstrated by Watson and Rayner's 1920 "Little Albert" study, in which a 9-month-old infant was conditioned to fear a white rat by pairing it with a loud, startling noise. The fear generalized to similar stimuli — a rabbit, a fur coat, and a Santa Claus mask — illustrating stimulus generalization of conditioned emotional responses. The study raised serious ethical concerns because no de-conditioning was performed. While phobias can develop through classical conditioning, it is worth noting that many phobias have no identifiable conditioning event.
Systematic desensitization, developed by Joseph Wolpe, is a counter-conditioning technique used to treat phobias. The approach pairs the feared stimulus with a relaxation response, which is physiologically incompatible with anxiety. Treatment progresses from the least-feared to the most-feared version of the stimulus. Aversion therapy takes the opposite approach, pairing an undesirable behavior with an unpleasant stimulus — for example, Antabuse (disulfiram) causes nausea when alcohol is consumed, conditioning an aversive response to drinking.
Classical conditioning principles are also applied in advertising and marketing, where products (CSs) are paired with pleasant images, music, or attractive people (USs) so that the product eventually elicits a positive emotional response (CR). In a striking demonstration of the breadth of conditioning, Ader and Cohen showed in 1975 that even immune responses can be classically conditioned: rats given a taste paired with an immunosuppressive drug later showed immune suppression in response to the taste alone.
A diagram of Watson's Little Albert experiment in four steps. Panel A: Before conditioning — Albert plays happily with a white rat (NS produces no fear). Panel B: Conditioning trial — the white rat (NS) is presented and a loud steel bar is struck behind Albert's head (US), producing a fear response (UR) with Albert crying. Panel C: After conditioning — the white rat alone (now CS) produces crying and avoidance (CR). Panel D: Generalization — Albert shows fear responses to similar stimuli: a white rabbit, a fur coat, and a Santa Claus mask, with diminishing response strength shown by progressively smaller fear indicators.
---