Classical Conditioning Part 3

Unconditioned Stimulus (US): Intensity

• Study (Paul & Chair, 1984)
  • Preparation: cat hind-limb flexion.
  • CS = light flash.
  • US = leg shock at 1, 2, 3, 4 mA.
  • CR & UR identical: flexion magnitude.
  • Results
  • Higher US intensity →
    • Faster acquisition (steeper learning curve across 6 sessions).
    • Higher asymptotic CR magnitude.
    • Diminishing returns: performance eventually plateaus for all groups.
  • Interpretation
  • Animals encode not only that shock follows the light but how strong it will be.
  • Greater US intensity supplies more salient prediction error → faster associative strength increment (∆V in Rescorla–Wagner terms).
  • Supports idea that quantitative properties of US are represented in memory.

Unconditioned Stimulus (US): Quality

• Auto-shaping in pigeons (Jenkins & Moore, 1973)
  • CS = key light.
  • US1 = grain; US2 = water.
  • High-speed film revealed two distinct CR topographies:
  • Grain → wide-open, ballistic peck as if swallowing food.
  • Water → closed-beak, gentle “nuzzling” as if sipping liquid.
  • Demonstrates stimulus substitution plus activation of innate feeding vs. drinking motor patterns.
• Innateness test (Woodruff & Starr, 1978)
  • Pigeons hand-reared without ever contacting solid grain or open water.
  • As adults they still produced food-like vs. water-like pecks under the appropriate US, proving the topographies are inherited (nativism).
• Conceptual links
  • Classical debate: empiricism (learned CS–US pairing) + nativism (hard-wired effector systems).
  • Quality of US selects which inherited response system is recruited.

Conditioned Stimulus (CS): Intensity

• Conditioned Emotional Response (CER) paradigm with rats
  • CS = panel light of four intensities: 1.5, 2.6, 8.7, 33 lux.
  • Reference points: 1.5 lux≈full tropical moon; 33 lux≈living-room lighting.
  • House light left on → CS must stand out against background.
  • US = foot-shock (constant across groups).
  • Dependent variable: CER suppression ratio $SR = \frac{R<em>{CS}}{R</em>{CS}+R_{pre}}$ (0 = total suppression, 0.5 = none).
  • Findings
  • Brighter CS → faster decline of SR (quicker learning).
  • Asymptote identical once CS clearly detected.
  • Key insight
  • Effect is perceptual/attentional (figure–ground), not encoding of magnitude (US intensity did that).
  • When background house light is OFF, performance equalises—no detection cost.

Conditioned Stimulus (CS): Quality & Belongingness

Garcia & Koelling (1966) – “Garcia Effect”

• Apparatus: lick spout delivering compound CS (taste + light + noise).
• Two US conditions during training:
  1. Shock (pain UR).
  2. Lithium Cl $\text{LiCl}$ (illness UR).
• Test CS (between-subjects):
  • “Tasty” water only.
  • “Bright-Noisy” water (light & noise only).
• Results (mean licks/min)
  • Illness → strong aversion to taste, no aversion to audiovisual cues.
  • Shock → strong aversion to bright-noisy cues, little aversion to taste.
• Statistical pattern: no main effects of US or CS type, but a perfect US × CS interaction.
• Interpretation
  • Evolutionary preparedness: gustatory cues linked to gastrointestinal consequences; exteroceptive cues linked to external threats.

Species-specific extension (Wilcoxon 1971)

• Training: blue-sweet water + LiCl.
• Test choices
  • Rats (nocturnal, olfactory specialists): avoid sweet; indifferent to blue.
  • Quail (diurnal, visual specialists): avoid blue; indifferent to sweet.
• Confirms that “belongingness” is modulated by each species’ ecological sensory biases.

Temporal Parameters

CS–US Interval (Inter-Stimulus Interval, ISI)

Auto-shaping in pigeons (Lucas 1981)

• ISI = 0, 4, 12, 36, 120 s; ITI long and constant.
• CR rate (pecks/min) declined monotonically with longer ISI.
• Effective conditioning requires contiguity within a few seconds.

Conditioned Taste Aversion (Callet & Rozin 1971)

• ISI = 0.5, 1, 1.5, 3, 6, 24 h.
• Preference score: lower = stronger aversion.
  • 30 min delay already produces robust CTA.
  • Aversion weakens as delay approaches 24 h, yet still detectable at several hours.
• Biological rationale: toxins manifest after digestion; system tuned for long delays.

Trace Conditioning & Contextual Control (Lubow-type shadow box study)

• Apparatus: two-compartment light/dark alley; rats prefer dark side.
• Phase 1 (training on dark side)
  • Groups: 0-s trace, 10-s trace, 30-s trace, No-CS (shock only).
  • CS = light; US = shock.
• Phase 2 (split sub-groups)
  1. Lick-suppression test to CS.
  2. Place-preference test (time in light compartment).
• Outcomes
  • Lick suppression
  • 0-s trace → maximal suppression (strong CS learning).
  • Suppression weakens as trace lengthens; minimal in No-CS group.
  • Place preference
  • 0-s trace → avoid light, prefer dark (context irrelevant).
  • Longer trace → increasing avoidance of dark side = context acquires control.
  • No-CS group shows greatest time in light side (strong context fear, no CS learning).
• Take-home
  • When CS and US are separated in time, associative strength shifts from discrete CS to background context.
  • Demonstrates comparator hypothesis: context competes with CS for predictive value.

Conceptual & Theoretical Connections

• Rescorla–Wagner: higher US intensity raises $\lambda$; faster approach to asymptote.
• Pearce–Hall/Mackintosh attentional models: salient CS (high lux) grabs attention → faster α.
• Preparedness (Seligman): innate CS–US predispositions (taste-illness, audiovisual-shock).
• Stimulus-substitution vs. behavior-systems approach: CR form reflects US-specific motivational/behavioral system.
• Comparator hypothesis / context conditioning: relative validity of cues shifts with temporal gap.

Practical / Ethical / Real-world Implications

• Taste-aversion learning informs chemotherapy nausea management & wildlife baiting (e.g.
  conditioned beef aversion in coyotes).
• Understanding US intensity guides humane shock parameters in animal research.
• Appreciating belongingness prevents ineffective pairings in behavior therapy (e.g.
  nausea-based alcohol aversion uses flavor, not lights).
• Trace/context conditioning findings shape PTSD models where context, not discrete cue, evokes fear.

Key Numerical References & Formulae

• US intensities: 1–4 mA shocks (cats);
  CS intensities: 1.5–33 lux lights (rats).
• CER suppression ratio: $SR = \frac{R<em>{CS}}{R</em>{CS}+R_{pre}}$.
• ISI ranges: 0–120 s (auto-shaping); 0.5–24 h (CTA).
• Trace intervals: 0, 10, 30 s → shift from CS to context control.

Summary Points

• Increasing US intensity speeds and magnifies conditioning; encodes quantitative expectancy.
• US quality steers which innate motor program appears in the CR.
• Increasing CS intensity mainly aids detection/attention; asymptotic performance unchanged.
• Certain CS–US pairings are “biologically prepared”; learning is asymmetric (Garcia effect) and species-specific.
• Effective ISI window is paradigm-specific: seconds for exteroceptive CRs, hours for CTA.
• When CS–US gap widens, background context competes, often overtakes the discrete CS as predictor of the US.