Learning and Conditioning

Classical Conditioning

Classical conditioning (Pavlov) is associative learning where a neutral stimulus acquires the capacity to elicit a response through pairing with a stimulus that naturally elicits that response. US (unconditioned stimulus) naturally elicits UR (unconditioned response). A neutral stimulus becomes CS (conditioned stimulus) after pairing with the US and elicits CR (conditioned response), typically similar to but weaker than the UR. Key processes: acquisition (CS-US pairing established), extinction (CR weakens when CS presented without US), spontaneous recovery (extinguished CR reappears after rest), stimulus generalization (responding to similar stimuli), stimulus discrimination (differentiating between stimuli). Higher-order conditioning pairs a new neutral stimulus with an established CS.

Operant Conditioning

Operant conditioning (Skinner) shapes behavior through consequences. Reinforcement increases behavior; punishment decreases it. Positive reinforcement adds a desirable stimulus (praise after studying); negative reinforcement removes an aversive stimulus (aspirin to remove headache). Positive punishment adds an aversive stimulus (scolding); negative punishment removes a desirable stimulus (taking away phone). The negative-reinforcement-versus-punishment distinction is heavily tested: negative reinforcement strengthens behavior by removing something unpleasant; punishment weakens behavior. Shaping reinforces successive approximations of a target behavior. Chaining links sequenced behaviors where each response cues the next.

Reinforcement Schedules

Continuous reinforcement produces rapid acquisition but rapid extinction. Partial (intermittent) reinforcement produces slower acquisition but greater resistance to extinction — the partial reinforcement extinction effect. Fixed-ratio (FR): reinforcement after a set number of responses; high rate with post-reinforcement pause. Variable-ratio (VR): reinforcement after an unpredictable number of responses; highest, steadiest rate with greatest extinction resistance — slot machines exemplify this. Fixed-interval (FI): reinforcement after a fixed time; scalloped response pattern with increased responding near reinforcement time. Variable-interval (VI): reinforcement after variable time; moderate, steady response rate — pop quizzes use this schedule.

Observational Learning

Bandura's social learning theory holds that organisms learn by watching others. The Bobo doll experiments showed children imitate aggressive modeled behavior, especially when the model is rewarded — vicarious reinforcement. Four processes: attention (notice the model), retention (remember the behavior), reproduction (capable of performing it), and motivation (reason to perform). Mirror neurons in the premotor cortex fire during both action execution and observation, providing a neural substrate for imitation and empathy.

Habituation and Sensitization

These are non-associative forms of learning — no linking of stimuli or stimulus-response. Habituation: decreased responsiveness to a repeated harmless stimulus (ignoring train noise near your home). Sensitization: increased responsiveness after a strong or noxious stimulus (heightened startle after a painful event). Dishabituation — restoring a habituated response via a novel stimulus — proves habituation is active CNS learning, not sensory fatigue. Kandel's Aplysia model revealed the synaptic basis: habituation involves decreased neurotransmitter release from sensory neurons; sensitization involves serotonin-mediated facilitation that enhances release.

Cognitive and Biological Factors in Learning

Tolman's latent learning showed rats formed cognitive maps without reinforcement, demonstrating learning only when motivation appeared — contradicting strict behaviorism. Seligman's learned helplessness occurs when uncontrollable aversive events teach an organism that actions are futile, producing passivity even when escape becomes possible. Garcia and Koelling's taste-aversion studies demonstrated biological preparedness: rats easily associate taste with illness but not taste with shock, and associate audiovisual cues with shock but not illness — challenging equipotentiality (the assumption that any stimulus can associate with any response equally well). The Brelands' instinctive drift showed biologically based instinctive behaviors can override learned operant responses.

How it works

Learning operates through synaptic strengthening and weakening. In fear conditioning, CS-US convergence in the lateral amygdala triggers long-term potentiation (LTP); the CS alone eventually drives fear responses via the central amygdala. Operant conditioning engages the mesolimbic dopamine pathway (VTA to nucleus accumbens and prefrontal cortex), encoding reward prediction error: dopamine neurons fire more for unexpected rewards, less when expected rewards are omitted, and shift firing to reward-predicting cues. This dopamine signal is the teaching signal for reinforcement learning. Addictive drugs hijack this circuit. Clinically, phobias are acquired through classical conditioning, maintained through operant conditioning (avoidance negatively reinforced by anxiety reduction), and treated through extinction therapies engaging the ventromedial prefrontal cortex to inhibit amygdala fear.

How it works

Learning operates through synaptic strengthening and weakening. In fear conditioning, CS-US convergence in the lateral amygdala triggers long-term potentiation (LTP); the CS alone eventually drives fear responses via the central amygdala. Operant conditioning engages the mesolimbic dopamine pathway (VTA to nucleus accumbens and prefrontal cortex), encoding reward prediction error: dopamine neurons fire more for unexpected rewards, less when expected rewards are omitted, and shift firing to reward-predicting cues. This dopamine signal is the teaching signal for reinforcement learning. Addictive drugs hijack this circuit. Clinically, phobias are acquired through classical conditioning, maintained through operant conditioning (avoidance negatively reinforced by anxiety reduction), and treated through extinction therapies engaging the ventromedial prefrontal cortex to inhibit amygdala fear.

Comparisons

• Nervous system (Biology): LTP and synaptic plasticity in hippocampus and amygdala are the cellular mechanisms of learning. Expect questions linking AMPA receptor insertion and NMDA-receptor-dependent calcium influx to behavioral outcomes.
• Biochemistry: Dopamine signaling through D1/D2 GPCRs in the mesolimbic pathway is central to reward learning. Drug addiction questions trace from dopamine release to cAMP to CREB-mediated gene expression.
• Sociology: Observational learning and modeling underpin socialization theories including gender role acquisition. Bandura's work bridges psychology and sociology.
• Behavioral medicine: Systematic desensitization and flooding are extinction applications. Exposure therapy with response prevention for OCD leverages habituation and extinction.

Common confusions

• "Negative reinforcement is punishment." Negative reinforcement removes an aversive stimulus to increase behavior. Punishment decreases behavior. Aspirin for a headache is negative reinforcement; a scolding is positive punishment.
• "Extinction erases the original association." Extinction is new inhibitory learning, not erasure. The CS-US association persists — evidenced by spontaneous recovery, renewal, and reinstatement. The prefrontal cortex actively inhibits amygdala fear responses.
• "Variable-ratio and variable-interval are the same." VR is response-based (highest rates, slot machines); VI is time-based (moderate steady rates, checking email).
• "Observational learning requires direct reinforcement of the observer." Vicarious reinforcement — seeing the model rewarded — is sufficient. This distinguishes Bandura from strict operant conditioning.
• "Habituation equals sensory adaptation." Sensory adaptation is receptor-level (physical). Habituation is CNS-level (learning), reversed by dishabituation.
• "Taste aversion violates classical conditioning." It follows classical conditioning but with single-trial learning, long CS-US delays, and biological selectivity — challenging equipotentiality, not the associative framework.

Quick review

• Classical conditioning: US -> UR, CS+US pairing -> CS -> CR. Acquisition, extinction, spontaneous recovery, generalization, discrimination, higher-order conditioning.
• Operant conditioning: positive reinforcement (add pleasant, increase behavior), negative reinforcement (remove aversive, increase behavior), positive punishment (add aversive, decrease behavior), negative punishment (remove pleasant, decrease behavior).
• Reinforcement schedules: FR (high rate, post-reinforcement pause), VR (highest/steadiest, greatest extinction resistance), FI (scalloped pattern), VI (moderate/steady). Partial reinforcement extinction effect.
• Observational learning: attention, retention, reproduction, motivation. Vicarious reinforcement. Mirror neurons fire during action execution and observation.
• Habituation: decreased response to harmless repeated stimulus (CNS learning, not sensory fatigue). Sensitization: increased response after strong stimulus. Dishabituation restores habituated response.
• Latent learning (Tolman): cognitive maps form without reinforcement. Learned helplessness (Seligman): passivity after uncontrollable aversive events.
• Biological preparedness (Garcia): taste-nausea links form easily; taste-shock does not. Challenges equipotentiality. Instinctive drift (Brelands): instinct overrides learned operant behavior.