You eat a new dish, and six hours later you're violently sick. You know — intellectually — that it was almost certainly a stomach bug, not the food. The food didn't make you sick. And yet, for years afterward, the smell of that dish turns your stomach. You can't talk yourself out of it.
That reaction breaks every rule of conditioning you learned last lesson. The "neutral stimulus" and the "sickness" were six hours apart, not half a second. It happened once, not over dozens of trials. And it locked in despite your conscious knowledge that the food was innocent.
This is taste aversion, and when John Garcia documented it in the 1960s, it nearly broke behaviorism. The reigning assumption was that any stimulus could be linked to any response if you paired them closely enough. Garcia's rats proved that wrong. Your brain isn't a blank slate that learns whatever it's fed — it comes pre-tuned by evolution to learn certain associations fast and to resist others entirely. This lesson is where biology crashes into learning theory, and where Unit 3 finally clicks together.
For the last three lessons, learning looked mechanical. Pair a tone with food enough times (Pavlov), reinforce a lever-press (Skinner), and associations form on schedule. The implied rule — call it equipotentiality — was that any neutral stimulus is as good as any other for conditioning, and any response can be attached to any cue. This lesson is the long list of exceptions that overturned that rule. Biology and cognition both put limits on what conditioning can do.
Biological preparedness is the principle that organisms are biologically predisposed to learn some associations more easily than others, because those associations mattered for survival. Evolution didn't hand you a general-purpose learning machine; it handed you one with strong priors. You learn snake-fear fast and electrical-outlet-fear slowly, even though outlets are the real modern threat.
The flagship example is taste aversion (also called the Garcia effect): a learned avoidance of a food whose taste or smell was followed by nausea. John Garcia and Robert Koelling (1966) made rats sick (using radiation or a drug) hours after they drank flavored water. The rats developed a powerful aversion to that flavor. Two features make this impossible under standard classical conditioning:
Why does the brain break its own rules here? Because for an animal that eats something toxic, the poison's effects naturally arrive long after the meal, and you rarely get a second chance to learn. A brain pre-tuned to link "novel taste" with "later sickness" — fast, and across a long gap — survives. So evolution built that specific shortcut.
Garcia found something even sharper: the association is selective. Rats readily linked taste with nausea but could not easily link a light or sound with nausea. Flip it: they linked light/sound with shock easily, but couldn't link taste with shock. The brain pairs internal-gut consequences with internal cues (taste) and external-pain consequences with external cues (sights, sounds). This is belongingness — some CS–UCS pairings "belong" together biologically and others don't. Equipotentiality is dead: not all stimuli are interchangeable.
Try This. Survey five friends about a food they can't stand because it once made them sick. Ask: (1) How many bad experiences did it take? (2) How long after eating did the sickness hit? (3) Do they believe the food was actually the cause? You'll almost always hear "just once," "hours later," and "honestly, probably not the food." That gap between belief and reaction is biological preparedness overriding conscious reasoning.
Constraints aren't just a classical-conditioning story. Keller and Marian Breland, two of Skinner's former students, went into business training animals for ads and shows using operant conditioning. It worked — until it didn't. A raccoon trained to drop coins into a piggy bank for food started rubbing the coins together instead, exactly as it would rub and "wash" food. Pigs trained to carry coins began rooting them along the ground.
The Brelands called this instinctive drift: the tendency of a conditioned animal to revert to innate, instinctive behaviors that interfere with the trained response, especially when food is involved. Reinforcement had shaped a behavior, but the animal's species-specific instincts (food-handling behaviors) kept "drifting" back in and overriding the training. The lesson echoes Garcia's: operant conditioning is powerful, but it works with and against an animal's biological predispositions, not on a blank slate. You can't reinforce an animal into doing just anything.
Why do phobias cluster so tightly around a small set of objects — snakes, spiders, heights, deep water, the dark, enclosed spaces — and almost never around cars, guns, or electrical outlets, which injure far more people today? Martin Seligman's answer is preparedness: we are evolutionarily prepared to acquire fears of stimuli that threatened our ancestors. Snake-fear is learned with one bad experience and resists extinction; outlet-fear barely forms at all. The threats that shaped our nervous system are millions of years old, so our fear-learning is tuned to them, not to modern dangers. This is the same principle as taste aversion, pointed at fear instead of nausea: biology decides which associations are easy to learn.
Biology isn't the only thing complicating "pairing builds associations." Cognition — what the organism thinks, expects, and represents — matters too.
Robert Rescorla's work is the key revision to classical conditioning. Pavlov implied that conditioning is about contiguity — CS and UCS occurring together. Rescorla showed it's really about contingency and predictiveness: a CS becomes a conditioned stimulus only if it reliably predicts the UCS — if it provides genuine information. In his studies, a tone paired with shock 100% of the time produced strong conditioning; but a tone that was also paired with shock the same number of times, yet where shocks also occurred without the tone, produced weak conditioning — because now the tone predicts nothing new. The animal isn't a passive association-counter; it's tracking expectancy — "does this cue tell me the UCS is coming?" Conditioning is the brain learning the predictive relationships among events. (This cognitive view is why some psychologists describe even Pavlovian conditioning as the organism building an internal model of the world.)
The cognitive thread runs back through the whole unit. Edward Tolman's rats (recall from Lesson 20) ran a maze without reinforcement and showed latent learning — learning that occurs but isn't demonstrated until there's a reason to perform it. Tolman argued the rats had built a cognitive map, a mental representation of the maze's layout, rather than a chain of stimulus–response habits. That, too, says learning is more than mechanical pairing: organisms acquire knowledge, and they deploy it when it pays.
Put it all together and the simple behaviorist picture gets revised from two sides at once:
The takeaway for the whole unit: conditioning is real and powerful, but it operates inside biological limits and through cognitive processes. Learning is the interaction of three things — what the environment provides (the conditioning), what evolution prepared (biology), and what the organism represents (cognition). That's the synthesis the AP exam wants you to be able to argue.
Garcia & Koelling, "Relation of cue to consequence in avoidance learning" (1966).
Who & when: John Garcia and Robert Koelling, 1966.
Method: Rats drank flavored water that was also paired with a "bright-noisy" stimulus (a light and clicking sound) when they licked the spout — so each lick produced three cues at once: a taste, a light, and a noise. Different groups were then made to experience one of two aversive consequences: nausea (induced by radiation or a drug) or pain (an electric shock to the feet). Researchers then tested which cue each group had learned to avoid.
Findings: The pairing was selective. Rats made nauseous avoided the taste but not the light/noise. Rats given shock avoided the light/noise but not the taste. The brain linked internal illness to an internal cue (taste) and external pain to external cues (sight/sound) — and refused the reverse pairings. Taste aversion also formed in a single trial despite a long delay between taste and sickness.
Significance: This is the study that killed equipotentiality. Not all stimuli are equally conditionable; animals are biologically prepared to form survival-relevant associations and resist others (belongingness). It forced learning theory to make room for evolution and remains the cornerstone example of biological constraints on learning. For the exam: Garcia = taste aversion, one-trial, long-delay, biological preparedness, "cue must belong to consequence."
Scenario 1. After a chemotherapy session, a patient eats a turkey sandwich. The chemo (not the sandwich) causes severe nausea hours later. For months afterward, the patient feels queasy at the mere smell of turkey, even though she knows the chemo was responsible.
Which concepts explain this, and how? This is taste aversion / the Garcia effect, an instance of biological preparedness. The tells are diagnostic: it took one trial, the nausea came hours later (violating contiguity), and it persists despite the patient's conscious knowledge that the food was innocent — showing the association is biologically prepared, not reasoned. (This is a real and well-documented problem in oncology, sometimes managed with a "scapegoat" novel food.)
Scenario 2. A trainer uses food rewards to teach a pig to pick up a wooden disc and deposit it in a box. At first the pig complies, but over weeks it increasingly drops the disc and roots it along the ground with its snout instead of carrying it, slowing down despite the reward.
Which concept explains this? Instinctive drift (Breland & Breland). The pig is reverting to an innate, species-typical food-handling behavior (rooting) that interferes with the trained operant response. Note what it is not: it isn't extinction (the reinforcement is still being delivered) and it isn't poor training — it's biology overriding conditioning.
Scenario 3. Two groups of dogs hear a tone before a puff of air to the eye. For Group A, the tone always precedes the puff and the puff never comes without the tone. For Group B, the tone precedes the puff equally often, but the puff also arrives randomly with no tone. Group A develops a strong blink to the tone; Group B barely responds.
Which concept explains the difference? Rescorla's contingency/predictiveness principle. Both groups received the same number of tone–puff pairings (equal contiguity), so contiguity can't explain the gap. What differs is whether the tone reliably predicts the puff. Only for Group A does the tone carry information, so only Group A conditions — evidence that classical conditioning depends on expectancy, a cognitive factor, not mere pairing.
Taste aversion vs. ordinary classical conditioning. Both are associative learning, but taste aversion breaks two rules: it forms in one trial and across a long delay (hours), whereas ordinary classical conditioning needs repeated pairings with near-simultaneous CS and UCS. Test tip: if you see "once" and "hours later," it's taste aversion / biological preparedness — not generic conditioning. The long delay is the giveaway.
Instinctive drift vs. extinction. Both look like a trained behavior "breaking down," so students swap them. Extinction = the behavior weakens because reinforcement stops. Instinctive drift = the behavior is displaced by an innate behavior even though reinforcement continues. Ask: is reward still coming? If yes and the animal reverts to instinct anyway, it's drift, not extinction.
Preparedness vs. equipotentiality. Equipotentiality (the old behaviorist assumption) = any stimulus can be conditioned to any response equally well. Preparedness (Seligman/Garcia) = evolution makes some associations easy and others nearly impossible. Garcia's study disproved equipotentiality and demonstrated preparedness. Don't credit the disproven idea to the researcher who killed it.
Biological constraints vs. cognitive factors. Both revise simple conditioning, but from different directions. Biological constraints come from evolution/instinct (preparedness, taste aversion, instinctive drift). Cognitive factors come from mental processing (Rescorla's expectancy/contingency, Tolman's cognitive maps, latent learning). If the explanation invokes survival/instinct → biological; if it invokes prediction/expectation/mental maps → cognitive.
Four-choice MCQs in current AP format. Answers and explanations in section (h).
EBQ format. You will read three summarized peer-reviewed sources on a shared topic, then write a single response that (1) states a defensible claim, (2) supports it with specific evidence from at least TWO sources, and (3) explains your reasoning by applying course concepts. Scoring: Claim 0–1 + Evidence (2+ sources) 0–3 + Reasoning & Application 0–2 = 7 points.
Prompt. Using the three sources below, develop an argument about the following question: To what extent do biological factors limit what associations can be learned through conditioning? Write a response that makes a defensible claim and supports it with evidence from at least two of the sources, applying psychological reasoning.
Source A — Garcia & Koelling (1966), "Relation of cue to consequence."
Rats licked flavored water from a spout; each lick simultaneously produced a distinctive taste, a bright light, and a clicking noise. Some rats were then made nauseous (via radiation/drug); others received foot-shock. Rats made nauseous later avoided the taste but not the light/noise; rats given shock avoided the light/noise but not the taste. Aversions formed even when illness followed the cue by a delay of more than an hour, and often after a single pairing. The authors concluded that the cue and the consequence must "belong" together for conditioning to occur, contradicting the assumption that any stimulus is equally conditionable to any consequence.
Source B — Breland & Breland (1961), "The misbehavior of organisms."
Two animal trainers used standard operant conditioning (food reinforcement) to teach commercial tricks to dozens of species. Repeatedly, trained behaviors deteriorated as animals reverted to innate, food-related action patterns: a raccoon "washed" and rubbed coins instead of depositing them; pigs rooted coins along the ground rather than carrying them. These reversions grew stronger over time and persisted even though they delayed or prevented reinforcement. The authors argued that an animal's instinctive behaviors impose limits on what operant conditioning can produce, a phenomenon they named instinctive drift.
Source C — Öhman & Mineka (2001), fear-conditioning review.
In laboratory studies, both humans and lab-reared monkeys acquire fear responses to "fear-relevant" stimuli (snakes, spiders) more rapidly and persistently than to "fear-irrelevant" stimuli (flowers, mushrooms, modern objects), even when pairings with an aversive event are matched. Monkeys with no prior snake experience readily learned snake-fear by watching another monkey react fearfully, but did not learn flower-fear the same way. The authors proposed an evolved "fear module" that is selectively tuned to threats recurrent in ancestral environments and is relatively resistant to conscious control.
Biological factors place substantial limits on what conditioning can produce: organisms are evolutionarily prepared to form certain survival-relevant associations easily while resisting others, so conditioning is not equally effective for every stimulus–response pairing. [CLAIM — 1 pt]
This claim is supported by all three sources. In Source A, Garcia and Koelling found that nauseous rats learned to avoid the taste but not the accompanying light and noise, while shocked rats learned to avoid the light and noise but not the taste; the aversion also formed across a delay of over an hour and often in a single trial. [EVIDENCE — Source A] Source B shows the same kind of limit in operant conditioning: despite continued food reinforcement, raccoons reverted to rubbing coins and pigs to rooting them, instinctive food-handling behaviors that grew stronger over time even though they cost the animals reinforcement. [EVIDENCE — Source B] Source C extends the pattern to fear: humans and monkeys acquired fear of snakes and spiders far more readily and persistently than fear of flowers or modern objects, even when the pairings were matched, and monkeys learned snake-fear but not flower-fear through observation. [EVIDENCE — Source C — 3 pts for specific evidence from 2+ sources]
These findings revise the behaviorist assumption of equipotentiality — that any neutral stimulus can be conditioned to any response with enough pairings. Garcia's selective, one-trial, long-delay learning demonstrates biological preparedness and belongingness: the brain links internal illness to an internal cue (taste) and external pain to external cues (light/noise) because those pairings were adaptive, and it resists the reverse pairings no matter how often they occur. The Brelands' instinctive drift shows the same constraint operating against operant conditioning — innate, species-specific behaviors override reinforced ones, so reinforcement cannot shape just any behavior. Öhman and Mineka's evolved "fear module" is preparedness applied to phobias: fears of ancestral threats are learned fast and resist extinction and conscious control, which is why phobias cluster around snakes and heights rather than the statistically more dangerous cars and outlets. Across classical conditioning (A), operant conditioning (B), and fear learning (C), the same principle holds: evolution pre-tunes the organism, so biology determines which associations conditioning can readily build. [REASONING & APPLICATION — 2 pts: applies preparedness, belongingness, equipotentiality, instinctive drift]
1. (B). Taste aversion's signature challenges to classical conditioning are one-trial learning and long-delay learning. (A) describes ordinary conditioning, which taste aversion violates; (C) is false — nausea is the UCS; (D) is the contiguity rule that taste aversion breaks.
2. (C) Biological preparedness (belongingness). The selective pairing — taste-with-nausea, sight/sound-with-shock — shows some associations "belong" together biologically. (A) equipotentiality is the idea Garcia disproved; (B) and (D) are unrelated conditioning phenomena.
3. (B) Instinctive drift. Reverting to innate food behaviors that disrupt a trained response is the Brelands' instinctive drift. (A) extinction requires reinforcement to stop; (C) is Tolman; (D) is a generalization of a CR to similar stimuli.
4. (B). Seligman's preparedness: evolution readies us to fear ancestral threats. (A) ignores how readily and uniformly these fears form; (C) is too strong (modern fears can form, just less easily); (D) is factually false — it's the opposite (cars injure more), which is the whole puzzle preparedness solves.
5. (C). Rescorla showed predictiveness/contingency — whether the CS reliably forecasts the UCS — is the key, not mere pairing. (A) and (B) are the contiguity/frequency view Rescorla revised; (D) concerns response strength, not the conditioning mechanism.
6. (B). One trial, sickness hours later, persisting despite knowing the food was innocent = taste aversion / biological preparedness. (A) is wrong — it did not take repeated trials; (C) is operant/instinct, not relevant; (D) requires observing a model, which didn't occur.
7. (B). Equipotentiality says any stimulus is equally conditionable; rats learning taste–illness far more easily than taste–shock directly contradicts that. (A) is consistent with equipotentiality; (C) and (D) are operant phenomena unrelated to the principle.
8. (B) Instinctive drift. Treats are still being given, yet the dog reverts to an innate digging/food-seeking pattern that disrupts the trick — drift, not extinction. (A) is wrong because reinforcement continues; (C) latent learning involves unrewarded learning shown later; (D) negative reinforcement is unrelated.
9. (C) Latent learning and the cognitive map. Learning occurred without reinforcement and was demonstrated only once reward appeared, implying a stored mental representation. (A), (B), (D) are different phenomena.
10. (B). Belongingness = certain CS–UCS pairings are learned more readily because they fit together biologically. (A) overstates; (C) describes a reinforcement timing rule; (D) is nonsense pairing of unrelated terms.
11. (B). Equal pairings (40 each) but different outcomes prove that pairing frequency isn't the driver; the difference is whether the tone predicts the puff (Group X: puff never without tone; Group Y: puff often without tone). This is Rescorla's contingency. (A) is refuted by the equal pairings; (C) and (D) aren't manipulated here.
12. (B) One-trial taste aversion (the Garcia effect). A single poisoned meal produces lasting avoidance of the prey — the applied use of taste-aversion learning ("conditioned taste aversion" in wildlife management). (A) is operant/instinct; (C) involves unrewarded learning; (D) chains one CS to another.
13. (C). Biology constrains which associations form easily (preparedness/drift) and cognition means organisms learn predictive relationships (Rescorla) and representations (Tolman). (A) is an overreach; (B) is the disproven equipotentiality; (D) wrongly excludes cognitive factors.
14. (B). Taste aversion is a biologically prepared special case of associative learning — it refines, not refutes, learning theory. (A) overstates; (C) is false (it differs in trials and delay); (D) is wrong — it's a classical (associative) process, just a constrained one.
15. (B). Failure to condition light–nausea but easy conditioning of light–shock, regardless of trial count, shows belongingness — some pairings belong and others don't. (A) is the opposite (this disproves equipotentiality); (C) Rescorla's model is about predictiveness, not which cues belong with which consequences; (D) is unrelated.
| Component | Points | Criteria |
|---|---|---|
| Claim | 0–1 | 1 pt for a defensible, clearly stated position answering the extent of biological limits on conditioning. |
| Evidence | 0–3 | Up to 3 pts for specific, accurate evidence drawn from at least two of the three sources. Vague or single-source support earns fewer points; evidence must be concrete (e.g., the selective taste-vs-light pairing; coins-rubbing reversion under continued reward; matched snake-vs-flower fear acquisition). |
| Reasoning & Application | 0–2 | Up to 2 pts for explaining why the evidence supports the claim by applying named course concepts (preparedness, belongingness, equipotentiality, instinctive drift, contingency). Must go beyond restating evidence. |
| Total | 7 | |
PsyIQ · Lesson 21 of 30 · Unit 3: Development and Learning. This lesson's FRQ practice is an Evidence-Based Question (EBQ) — 3 summarized sources, scored Claim (1) + Evidence from 2+ sources (3) + Reasoning & Application (2) = 7. MCQ and EBQ practice modeled on the redesigned (2025+) AP Psychology exam. Not affiliated with the College Board. AP is a registered trademark of the College Board. Content pending external psychology QC.