Section 1
Basic Concepts
Classical (Pavlovian) conditioning: a neutral stimulus comes to predict a biologically significant event. The fundamental formula is CS + US → CR.
| Term | Definition & Example |
|---|---|
| US (Unconditioned Stimulus) | Naturally evokes a response without prior training. Examples: food, airpuff, electric shock. |
| UR (Unconditioned Response) | Reflexive, automatic response to the US. Occurs without any learning. |
| CS (Conditioned Stimulus) | Formerly neutral stimulus that, after pairing with the US, predicts the US. Example: bell, tone, light. |
| CR (Conditioned Response) | Learned anticipatory response to the CS — occurs before the US arrives. Protective and adaptive. |
- Appetitive conditioning: US is positive (food, sex) → CR reflects approach behavior.
- Aversive conditioning: US is unpleasant (shock, airpuff) → CR reflects defensive preparation.
- CER (Conditioned Emotional Response): Estes & Skinner — rats freeze (CR) to a tone (CS) predicting shock (US). Measurable as suppression of lever pressing.
- Eyeblink conditioning: tone CS + airpuff US → anticipatory eyeblink CR that peaks just as the US is expected to arrive.
Common Misconception — CR vs. UR
Both can be eyeblinks! The key distinction: the UR is an automatic reflex triggered by the US itself. The CR is a learned anticipatory response to the CS that occurs before the US arrives, providing adaptive protection. Same behavior, different eliciting stimuli and different timing.
Section 2
Extinction & Spontaneous Recovery
Extinction: CS presented repeatedly without the US → CR diminishes. But extinction does NOT erase the original learning.
Extinction creates new inhibitory learning — a "do not respond" association that competes with the original excitatory CS-US association. The original excitatory memory remains latent.
- Evidence 1 — Spontaneous recovery: After a rest period following extinction, the CR returns without any new CS-US pairings.
- Evidence 2 — Context dependence: Extinction performed in context A does not eliminate CRs in context B. The CR can return in a different context.
Key Principle — Extinction Is NOT Unlearning
The original CS-US association remains latent. Spontaneous recovery after rest, and context-dependent return of the CR, both prove the original memory was never destroyed. This is why recovered addicts remain vulnerable to relapse — conditioned cravings are suppressed by new inhibitory learning, not erased. A change of environment (context) can bring them flooding back.
Section 3
Compound Conditioning: Blocking & Overshadowing
Compound conditioning: two or more cues presented together as a compound CS (e.g., tone + light).
Overshadowing: the more salient cue in the compound captures a disproportionately large associative weight, reducing learning about the weaker cue. Can occur on the very first compound trial — no pre-training required.
Blocking (Kamin, 1969): Pre-training one element of the compound prevents the other element from acquiring associative strength, even when both are paired with the US.
Kamin Blocking Paradigm
Phase 1: Light → Shock W_Light = 100Phase 2: Light + Tone → Shock PE = 0 (Light already predicts fully)
Phase 3: Tone alone → no CR (Tone was never informative)
Prediction error (PE): actual US − expected US.
| PE Value | Interpretation | Outcome |
|---|---|---|
| Positive (actual > expected) | US was surprising | Associative weight increases — learning occurs |
| Zero (actual = expected) | US fully predicted | No change — asymptote reached |
| Negative (actual < expected) | Predicted US absent | Weight decreases — extinction |
Clinical Application — Rachel's Popsicles
A novel cucumber-flavored popsicle given before chemotherapy becomes a strongly conditioned CS for nausea, blocking the conditioning of preferred everyday foods to nausea. The novel cue absorbs the associative weight, protecting the patient's normal food preferences.
Section 4
The Rescorla-Wagner Model
Rescorla-Wagner Update Rule
ΔW(CS) = α × (actual US − expected US)where expected US = Σ W of all CSs present on the trial
| Parameter | Meaning |
|---|---|
| W | Associative weight of the CS (scale 0–100); how strongly CS predicts US. |
| α (alpha) | Learning rate (0–1). Example: α=0.2 means 20% of prediction error is absorbed per trial. |
| Expected US | Sum of all W values for CSs present on this trial. |
| US Modulation | Learning is governed by how surprising the US is — the R-W framework. |
Three Core Situations
- Novel CS + unexpected US: PE positive → weight increases → organism learns CS predicts US.
- Well-trained CS + expected US: PE = 0 → no change → learning has reached asymptote.
- CS predicts US but US absent: PE negative → weight decreases → extinction.
Extinction calculation example: W_Tone = 100, α = 0.2. First extinction trial (tone presented, no shock): actual = 0, expected = 100, PE = 0 − 100 = −100. ΔW = 0.2 × (−100) = −20. New W = 80. Learning declines gradually, producing the extinction curve.
CS modulation theory (Mackintosh alternative): Learning rate changes based on attention to the CS, not US surprise. Explains phenomena R-W cannot.
R-W Limitation — Latent Inhibition
Pre-exposing the CS without the US: actual = 0, expected = 0, PE = 0. R-W predicts nothing is learned. But empirically, pre-exposure significantly retards later conditioning — something meaningful was learned (CS predicts nothing). This requires a CS modulation account where attention to the CS itself changes.
Section 5
Latent Inhibition & CS Modulation
Latent inhibition: prior exposure to the CS without the US impairs later CS-US conditioning. The CS becomes "familiar as irrelevant." Mackintosh argued that organisms learn to ignore stimuli that do not reliably predict important events — attention to the CS decreases.
| Theory | What drives learning? | Explains blocking? | Explains latent inhibition? |
|---|---|---|---|
| R-W (US Modulation) | How surprising the US is (PE = actual − expected) | Yes — light fully predicts US, PE = 0 for tone | No — PE = 0 during pre-exposure, predicts no effect |
| Mackintosh (CS Modulation) | Attention allocated to the CS | Yes — tone is redundant so attention shifts away | Yes — organism learns to ignore the pre-exposed CS |
Current View
Both US modulation and CS modulation operate simultaneously, mediated by distinct neural systems: the cerebellum implements error-correction (R-W-like); the hippocampus mediates CS modulation and latent inhibition.
Section 6
Brain Substrates: Cerebellum & Hippocampus
Eyeblink conditioning has been mapped to a highly specific cerebellar circuit. The hippocampus plays a separate, modulatory role.
| Structure | Role in Eyeblink Conditioning |
|---|---|
| Cerebellar Cortex / Purkinje Cells | CS pathway: mossy fibers → granule cells → parallel fibers → Purkinje cells. Purkinje cells INHIBIT the interpositus nucleus. Their firing decreases as the CR is learned, releasing inhibition of CR output. |
| Interpositus Nucleus | Sole output pathway for the CR. Lesion permanently abolishes CRs while leaving URs intact. Critical bottleneck. |
| Inferior Olive | Carries the US signal via climbing fibers to cerebellar cortex. Encodes the prediction error (actual − expected US). Activity starts high early in training and decreases as CR is acquired — exactly mirroring R-W. |
| Pontine Nuclei | CS brainstem relay. Different subregions carry different CSs (tone vs. light) to the cerebellum via mossy fibers. |
| Hippocampus | NOT required for basic CR acquisition. Essential for CS modulation / latent inhibition. Hippocampal lesions eliminate latent inhibition in rabbit eyeblink conditioning. |
Inhibitory feedback loop: The interpositus nucleus sends an inhibitory signal back to the inferior olive, implementing the R-W prediction error computation. As the CR is acquired, this feedback reduces inferior olive firing — signaling "US is now expected." Disrupting this feedback loop impairs blocking.
Exam Tip — Double Dissociation
Interpositus lesion = no CR (but UR intact) — cerebellar circuit required for CR output.Hippocampal lesion = no latent inhibition (but basic CR intact) — hippocampus required for CS modulation.
Know both lesion patterns and what each tells us about the neural division of labor.
Section 7
Invertebrate Model: Aplysia
Classical conditioning in Aplysia requires the CS (siphon touch) to precede the US (tail shock) by approximately 0.5 seconds — this temporal window reflects the duration of an intracellular priming state.
Activity-dependent enhancement: The CS activates the sensory neuron, transiently priming it. When the US pathway releases serotonin shortly after, the primed neuron produces a larger glutamate release than sensitization alone would cause. The CS must come first; the US pathway finds a primed target.
| Feature | Short-Term Conditioning | Long-Term Conditioning |
|---|---|---|
| Level of change | Intracellular; vesicle numbers | Structural; new synapse growth |
| Gene expression required? | No | Yes — requires CREB-1 activation |
| CREB-2 role | Not relevant | Opposes CREB-1; must be suppressed |
| Duration | Minutes to hours | Days to weeks |
Section 8
Clinical Applications: Drug Tolerance & Addiction
Drug tolerance as conditioned compensatory response: Environmental cues (CS) associated with repeated drug use trigger a homeostatic CR that opposes the drug's physiological effects. The body anticipates the drug and partially counteracts it.
When a drug is taken in a novel context (CS absent), no compensatory CR fires. The same dose now produces an exaggerated effect. This explains why most heroin overdoses occur in experienced users who are in an unusual setting — the familiar cues that normally trigger tolerance are absent.
Bouton's Principles for Extinction-Based Therapy
- Conduct cue-exposure therapy across many different contexts, not just a clinical office.
- Spread therapy over time — a single session of extinction is insufficient for lasting suppression.
- When possible, use the same contexts where the habit originally formed.
- Consider including small drug doses during cue exposure — the drug itself is part of the CS context.
Why Multiple Contexts Matter
Extinction is highly context-specific. Extinguishing a CR only in a clinical office does not generalize well to real-world environments where the original habit formed. Conditioning across multiple contexts and time periods creates broader, more robust inhibitory learning.
Key Terms — 25 Flashcards
Click any card to reveal its definition. Use the filters to focus on a category.
Behavioral
CS (Conditioned Stimulus)
Click to reveal →
Behavioral
Formerly neutral stimulus that, after repeated pairing with the US, comes to predict the US and elicit a conditioned response.
Behavioral
US (Unconditioned Stimulus)
Click to reveal →
Behavioral
Stimulus that naturally evokes a response without prior training. Examples: food, airpuff, electric shock, drug.
Behavioral
CR (Conditioned Response)
Click to reveal →
Behavioral
Learned anticipatory response to the CS that occurs before the US arrives. Protective — prepares the organism for the expected US.
Behavioral
UR (Unconditioned Response)
Click to reveal →
Behavioral
Automatic, reflexive response to the US. Requires no prior learning. Distinguished from CR by its eliciting stimulus and timing.
Behavioral
Extinction
Click to reveal →
Behavioral
CS presented without the US → CR declines. Creates new inhibitory learning that competes with the original excitatory memory — does NOT erase it.
Behavioral
Spontaneous Recovery
Click to reveal →
Behavioral
Return of the extinguished CR after a rest period without additional CS-US pairings. Proves extinction did not destroy the original memory.
Behavioral
Appetitive Conditioning
Click to reveal →
Behavioral
Classical conditioning with a positive US (food, sex). The CR typically reflects approach behavior toward the CS.
Behavioral
Aversive Conditioning
Click to reveal →
Behavioral
Classical conditioning with an unpleasant US (shock, airpuff). CR reflects defensive preparation — withdrawal, freezing, eyeblink.
Behavioral
Conditioned Compensatory Response
Click to reveal →
Behavioral
Homeostatic CR opposing the drug's effects, triggered by environmental CS cues. Underlies tolerance; its absence in novel contexts explains overdose.
Behavioral
Latent Inhibition
Click to reveal →
Behavioral
Prior CS-alone exposure retards later CS-US conditioning. CS becomes "familiar as irrelevant." Cannot be explained by Rescorla-Wagner.
Theory
Rescorla-Wagner Model
Click to reveal →
Theory
ΔW(CS) = α × (actual US − expected US). A US modulation theory — learning is driven by how surprising the US is (prediction error).
Theory
Prediction Error
Click to reveal →
Theory
actual US − expected US. Positive → weight increases. Zero → no change. Negative → weight decreases (extinction). Core drive of R-W learning.
Theory
Blocking / Kamin Effect
Click to reveal →
Theory
A pre-trained CS blocks a new CS in a compound from acquiring associative weight, because PE = 0 — the US is already fully predicted.
Theory
Overshadowing
Click to reveal →
Theory
In a compound CS, the more salient cue captures disproportionately more associative weight. Unlike blocking, no pre-training is required — can occur on the first trial.
Theory
CS Modulation Theory / Mackintosh
Click to reveal →
Theory
Learning rate (α) changes based on attention to the CS. Explains latent inhibition: organisms learn to ignore irrelevant CSs. Contrasts with R-W's US modulation.
Theory
US Modulation Theory
Click to reveal →
Theory
Learning is governed by how unexpected the US is. More surprising US = more learning. Implemented by the Rescorla-Wagner model.
Neural
Interpositus Nucleus
Click to reveal →
Neural
Deep cerebellar nucleus; the sole output pathway for the conditioned eyeblink CR. Lesion permanently abolishes CRs but leaves URs intact.
Neural
Inferior Olive
Click to reveal →
Neural
Carries US information via climbing fibers to the cerebellar cortex. Encodes the prediction error (actual − expected US). Activity decreases as CR is learned.
Neural
Purkinje Cells
Click to reveal →
Neural
Cerebellar cortex output neurons that inhibit the interpositus nucleus. Their firing decreases as the CR is acquired, releasing the interpositus to produce the CR.
Neural
Pontine Nuclei
Click to reveal →
Neural
Brainstem relay for the CS pathway to the cerebellum via mossy fibers. Different subregions carry different CSs (tone vs. light).
Neural
Activity-Dependent Enhancement
Click to reveal →
Neural
Aplysia mechanism: CS primes the sensory neuron so that subsequent US serotonin produces larger glutamate release than sensitization alone. Explains why CS must precede US.
Clinical
Drug Tolerance
Click to reveal →
Clinical
Reduced drug effect with repeated use, partly due to a conditioned compensatory CR triggered by environmental CS cues. Same dose in novel context (no CSs) = exaggerated effect = overdose risk.
Clinical
Extinction-Based Therapy / Bouton's Principles
Click to reveal →
Clinical
Addiction treatment via cue exposure: (1) many different contexts, (2) spaced over time, (3) same contexts as original habit, (4) consider small drug doses as part of the CS context.
Molecular
CREB-1
Click to reveal →
Molecular
Transcription factor required for long-term conditioning in Aplysia. Activates genes driving new synapse growth. CREB-1 inactivation blocks long-term but not short-term memory.
Molecular
CREB-2
Click to reveal →
Molecular
Transcription factor that opposes CREB-1. Must be suppressed for long-term conditioning to occur. Acts as a gating mechanism for memory consolidation.
Practice Multiple Choice — 25 Questions
Click an option to check your answer. Explanations appear automatically.
35
Questions
0
Correct
0
Incorrect
—
Score
Big Picture Synthesis
How this week's concepts connect across levels of analysis and to the course arc.
The Unifying Principle
Classical conditioning is the brain's fundamental mechanism for predicting biologically significant events. The organism learns WHEN to prepare (CS), for WHAT (US), enabling adaptive anticipatory responses — from eyeblinks to drug tolerance to emotional fear.
Levels of Analysis
Behavior
CR emerges — acquisition
CR declines — extinction
CR returns — spontaneous recovery
Circuit
CS pathway: pontine → cerebellar cortex
US pathway: inferior olive → interpositus
Hippocampus: CS modulation / LI
Synapse
Activity-dependent enhancement (Aplysia)
LTP at parallel fiber–Purkinje synapses
Inhibitory feedback: interpositus → olive
Structure
New synapse growth (long-term Aplysia)
CREB-1 gene expression required
Cerebellar interpositus as CR gate
Likely Exam Themes
- CR vs. UR distinction — same behavior, different timing and eliciting stimulus
- Blocking calculation — trace through phases, show PE = 0 in Phase 2
- R-W math — compute ΔW for acquisition and extinction trials
- Latent inhibition as R-W failure — explain why PE = 0 doesn't mean nothing learned
- Interpositus vs. hippocampus — double dissociation lesion predictions
- Short vs. long-term Aplysia memory — vesicles vs. new synapses, CREB-1 role
- Drug overdose scenario — apply conditioned compensatory response logic
Cross-Course Connections
- R-W prediction error mirrors dopamine RPE signals (Week 8+)
- Inferior olive = first neural PE implementation in the course
- Extinction-based therapy = real-world use of conditioning principles
- Latent inhibition → CS modulation bridges to Week 9 generalization
- Week 6 link: sensitization (heterosynaptic) vs. classical conditioning (activity-dependent enhancement) in Aplysia
- Interpositus inhibitory feedback previews RL error-correction architectures