Forms of Inhibition in Classical Conditioning

Latent Inhibition

• Pre-exposing the CS without the US slows later conditioning.
• The organism learns the CS is irrelevant → reduced attention to it.

External Inhibition

• A novel external cue introduced during conditioning temporarily suppresses the CR.
• Novelty distracts from the CS.

Conditioned Inhibition

• A CS signals the absence of the US.
• Example: A tone that predicts “no food.”

Inhibition of Delay

• CR is delayed until the expected time of the US.
• Example: Pavlov’s dogs salivate only after several seconds when CS–US interval is long.

Disinhibition

• A novel stimulus removes inhibition, temporarily restoring the CR.
• Often observed when context changes.

Conditioned Inhibition in the Rescorla–Wagner Model

• RW explains inhibitory learning as prediction learning:
  The inhibitory CS predicts no US, reducing prediction error over time.
• Associative strength becomes negative for inhibitors.
• Inhibition is expressed when the organism expects the US but it fails to occur.

Stimulus Control of Behaviour

Definition

Behaviour is said to be under stimulus control when its likelihood changes depending on the presence or absence of a particular stimulus.

Measuring Stimulus Control

• Compare behaviour under S⁺ (reinforced stimulus) vs S⁻ (nonreinforced stimulus).
• Example: Bell (S⁺) → salivation; no bell (S⁻) → no salivation.

Determinants of Stimulus Control

Factor	Explanation
Sensory Capacity	Whether the organism can detect the stimulus.
Sensory Orientation	Whether the organism is positioned or attending to it.
Stimulus Salience	More intense or meaningful cues gain control faster.
Motivational State	Hunger, fear, etc., influence what cues matter.

Example: Stopping at a traffic light requires the ability to detect red vs amber and discriminate between them.

Discrimination and Generalisation

Discrimination

• Learning to respond differently to distinct stimuli.
• Narrow responding: respond only to S⁺.

Generalisation

• Responding to similar stimuli beyond training conditions.
• Broader responding.

The two processes work in tension:

• More discrimination → narrow generalisation.
• Less discrimination → broad generalisation.

Generalisation in Classical Conditioning

Pavlov’s Findings

• Dogs trained to a 1000 Hz tone also responded to nearby tones (1100, 1200 Hz).
• Produces a generalisation gradient.

Two Major Explanations

Theory	Explanation
Pavlov (1927) — Innate Spread of Excitation	Generalisation is automatic based on physical similarity.
Lashley & Wade (1946) — Learning History Theory	Generalisation depends on lack of discrimination training; shaped by experience.

Jenkins & Harrison (1960–1962)

Studied pigeons trained to peck tones.

Findings

Training Type	Gradient	Interpretation
Nondifferential	Flat	No discrimination learning.
Presence–absence	Moderately peaked	Some discrimination.
Intradimensional (1000 Hz S⁺ vs 950 Hz S⁻)	Very sharp peak	Learning history shapes precision.

Conclusion: Stimulus control depends on how precise the discrimination training is.

Discrimination Training

Process

• Classical: S⁺ paired with US; S⁻ not.
• Operant: Reinforce behaviour only under S⁺.

Outcome

• Response becomes selective for S⁺.
• Reduced responding to S⁻.

Factors Shaping Discrimination

Factor	Effect
Value/Reinforcer Strength	Stronger reinforcers → faster control
Consistency	More trials → sharper discrimination
Similarity of S⁺ and S⁻	More similar → better discrimination learning
Type of Reinforcement	Positive/negative reinforcement shapes cue meaning

Generalisation Learning & Concepts

Generalisation Learning

Training designed to promote responses across varying examples.

Example: Teaching pigeons to peck at any image of water → stimulus equivalence.

Concept Learning

Responding to categories, not individual stimuli.
Theories include:

Theory	Mechanism
Exemplar Theory	Compare to stored examples
Prototype Theory	Compare to a category “average”
Feature Theory	Identify defining features

Example: Categorising birds vs non-birds.

Choice Behaviour

Choice involves selecting between alternatives based on reinforcement history.

Predictability

Choice is lawful and follows quantitative patterns like the Matching Law.

The Matching Law

Core Principle

Behaviour distribution matches reinforcement distribution:$$\frac{B_1}{B_1 + B_2} = \frac{R_1}{R_1 + R_2}$$B1​+B2​B1​​=R1​+R2​R1​​

• $B$B = behaviour responses
• $R$R = reinforcers obtained

Example:
If ⅔ of rewards come from left key, pigeons peck the left key ⅔ of the time.

Matching in the Real World

• Social: Talk more to peers who laugh at your jokes.
• Classroom: On-task time matches teacher attention.
• Sports: Teams allocate plays (run/pass) according to past success.

Quality and Relativity of Reinforcers

Reinforcer Quality Matters

Modified Matching Law:$$\frac{B_1}{B_1 + B_2} = \frac{Q_1R_1}{Q_1R_1 + Q_2R_2}$$B1​+B2​B1​​=Q1​R1​+Q2​R2​Q1​R1​​

Animals prefer:

• wheat > buckwheat
• richer, more immediate, or more palatable outcomes

Reinforcement Relativity

Reinforcer value depends on context and alternatives.
A mediocre stimulus becomes attractive when the only competitor is worse.

Example: Social media apps compete for behaviour allocation.

Optimisation and Self-Control

Optimization Theory

Organisms should maximise long-term reinforcement rates.
But real animals often match instead of optimise.

Self-Control Dilemma

Choosing between:

• Small immediate reward, or
• Larger delayed reward

Preferences shift as delays shrink (delay discounting).

Delay Discounting and Improving Self-Control

Delay Discounting

Value of a reinforcer declines as delay increases.
Explains impulsivity and preference reversals.

Behavioural Strategies

• Precommitment
• Self-reinforcement
• Adding punishers
• Manipulating environment

Cognitive Strategies

• Visualising long-term goals
• Distracting from temptation
• Reframing choices

Summary

• Inhibition: learned suppression of responding under specific conditions.
• Stimulus Control: cues dictate behaviour through learning history.
• Discrimination & Generalisation: complementary processes determining the precision and flexibility of behaviour.
• Choice: governed by reinforcement ratios; self-control relates to delay and value.