ch4 lm

Determinants of Effective Conditioned and Unconditioned Stimuli

Initial Behavioral Responses to Stimuli:
- Pavlovian definitions of stimuli:
  - Conditioned Stimulus ( $CS$ ): A stimulus that does not initially elicit the target response, but comes to do so as a direct result of being associated with an unconditioned stimulus.
  - Unconditioned Stimulus ( $US$ ): A stimulus that reliably and unconditionally elicits the target response from the outset, requiring no special prior training.
- Relativity of stimulus identification:
  - Whether a stimulus functions as a $CS$ or a $US$ is determined by its relationship to the other environmental stimuli in the conditioning arrangement.
  - Example of sucrose-flavored food pellets:
    - Operates as a $CS$ when paired with lithium-induced illness or radiation in a conditioned taste-aversion procedure, leading to an aversion to consuming the pellets.
    - Operates as a $US$ in a sign-tracking experiment (Franklin & Hall, 2011), where insertion of a response lever serves as the $CS$ preceding sucrose delivery, eliciting lever pressing and approach.
Stimulus Novelty and Familiarity Effects:
- Novel stimuli elicit stronger behavioral reactions and enter into classical conditioning associations far more rapidly than familiar stimuli.
- The Latent-Inhibition (CS-Preexposure) Effect:
  - Repeated nonreinforced exposure to a stimulus prior to conditioning impairs that stimulus's subsequent ability to serve as an effective $CS$ (Hall, 1991; Lubow & Weiner, 2010).
  - Experimental procedure involves two phases:
    - Phase 1 (Preexposure phase): The designated $CS$ is repeatedly presented alone without consequences, rendering it familiar, benign, and irrelevant.
    - Phase 2 (Conditioning phase): The preexposed $CS$ is paired with a $US$ using conventional Pavlovian procedures.
  - Behavioral outcome: Participants acquire conditioned responding substantially slower than control participants exposed to a novel $CS$ (De la Casa, Marquez, & Lubow, 2009).
  - Distinction from conditioned inhibition: Conditioned inhibition involves active signaling of the absence of an expected $US$ , whereas latent inhibition reflects an attenuation of attentional processing toward a stimulus with no predictive significance.
  - Evolutionary rationale: Protects organisms from cognitive information overload by suppressing attention to biologically irrelevant stimuli (Lubow, 2011).
  - Historical origin: Discovered in sheep by Lubow and Moore (1959).
  - Relevance to schizophrenia:
    - A key symptom of schizophrenia is the cognitive deficit in filtering out irrelevant ambient stimuli.
    - Patients with schizophrenia show a disruption or absence of the latent-inhibition effect.
    - Latent inhibition and schizophrenia share dopaminergic neurobiological substrates, making latent-inhibition paradigms a primary behavioral assay for evaluating cognitive dysfunctions in schizophrenia.
- The US-Preexposure Effect:
  - Prior familiarization with an unconditioned stimulus in the absence of a predictive conditioned stimulus severely delays subsequent associative conditioning when that $US$ is paired with a novel $CS$ .
  - Empirical study (Franklin & Hall, 2011):
    - Setup: Rats received $80$ preexposure deliveries per day of one food pellet flavor (bacon or chocolate) for $5$ consecutive days, totaling $400$ $US$ preexposure trials in a food cup flanked by two retractable levers.
    - Conditioning: $20$ Pavlovian trials were administered where insertion of one lever for $10\,s$ signaled the delivery of the bacon pellet, and insertion of the other lever signaled the chocolate pellet.
    - Results: Conditioning of the sign-tracking lever contact response proceeded significantly faster for the lever paired with the novel pellet flavor than for the lever paired with the familiar pellet flavor.
  - Generality: Documented extensively in appetitive conditioning, fear conditioning, taste-aversion conditioning, and drug conditioning (Randich & LoLordo, 1979; Hall, 2009).

US-preexposure effect showing faster acquisition of lever contacts for novel food over familiar food

Stimulus Intensity and Salience:
- Conditioning proceeds more rapidly and generates more vigorous conditioned responding when intense, highly salient conditioned and unconditioned stimuli are utilized (Bevins et al., 1997; Kamin, 1965; Scavio & Gormezano, 1974).
- Psychological stimulus salience corresponds to physical noticeability and biological significance (McLaren & Mackintosh, 2000; Pearce & Hall, 1980).
- Methods for manipulating salience:
  - Physical parameter alteration: Escalating stimulus amplitude, brightness, or volume.
  - Biological state manipulation: Inducing nutritional or physiological deficits makes specific stimuli more salient (Krieckhaus & Wolf, 1968). For instance, sodium-deficient rats learn significantly stronger conditioned aversions to salt flavors than sodium-replete controls (Sawa, Nakajima, & Imada, 1999).
  - Naturalistic stimulus configuration: Making conditioned stimuli structurally closer to stimuli encountered in natural ecologies.
    - Sexual conditioning in domesticated quail (Cusato & Domjan, 1998, 2012; Domjan et al., 2004): Male quail receive access to a female bird ( $US$ ) signaled by a $CS$ .
    - An arbitrary $CS$ (a foam block covered in terrycloth) produces slower, less durable conditioning.
    - A naturalistic $CS$ (a terrycloth block incorporating the taxidermically prepared head and neck feathers of a female quail) produces rapid acquisition, engages comprehensive copulatory behavior, and resists associative decay when the $CS$ - $US$ interval is widened.

Arbitrary terrycloth CS versus naturalistic taxidermic female head CS used in quail sexual conditioning

CS–US Relevance and Belongingness:
- Genetic predispositions and evolutionary constraints dictate which stimulus combinations are rapidly associated, violating the assumption of equipotentiality.
- Garcia and Koelling's (1966) "Bright-Noisy-Tasty Water" Experiment:
  - Procedure: Water-deprived laboratory rats drank from a drinking tube filled with flavored water (salty or sweet; taste $CS$ ). Each lick also triggered a brief flash of light and an audible click (audiovisual $CS$ ). Thus, taste and audiovisual cues occurred simultaneously.
  - Conditioning: One group of rats received a brief foot-shock ( $US$ causing peripheral cutaneous pain); a second group received ionizing radiation or a toxic drug injection ( $US$ producing visceral illness).
  - Testing: Cues were separated. Rats were tested for drinking suppression either with flavored water alone (no audiovisual feedback) or with unflavored water paired with the lick-triggered light and click.
  - Results:
    - Rats conditioned with illness suppressed drinking substantially when tested with the taste $CS$ , but drank normally when tested with the audiovisual $CS$ .
    - Rats conditioned with shock suppressed drinking substantially when tested with the audiovisual $CS$ , but drank normally when tested with the taste $CS$ .
  - Conclusions: Learning is determined by the ecological relevance or belongingness between the $CS$ and $US$ . Cutaneous pain selectively associates with external audiovisual cues, whereas internal gut distress selectively associates with gustatory cues.

Diagram of Garcia and Koelling compound taste-audiovisual conditioning procedure

Graph of licks per minute for taste versus audiovisual cues following sickness or shock

Developmental and Evolutionary Generality:
- Innate nature: The selective taste-sickness association is present in laboratory rats $1\,\text{day}$ after birth, demonstrating that individual trial-and-error experience is unnecessary for stimulus-relevance effects (Gemberling & Domjan, 1982).
- Human food aversions: People associate nausea with ingested tastes rather than dining room ambient sounds or lighting (Logue et al., 1981; Pelchat & Rozin, 1982).
- Primate fear acquisition (Öhman & Mineka, 2001; Mineka & Öhman, 2002): Rhesus macaques and humans acquire conditioned fear substantially faster to fear-relevant stimuli (images of snakes) than to fear-irrelevant stimuli (images of flowers or mushrooms). This selective bias operates below conscious awareness (Öhman et al., 2007). Human infants aged $8$ to $14\,\text{months}$ orient significantly faster to snake pictures than to flower pictures (LoBue & DeLoache, 2010).

Classical Conditioning Without an Unconditioned Stimulus

Higher-Order Conditioning:
- Enables learning to occur in the absence of a primary biological reinforcer by using a previously conditioned stimulus to condition a novel stimulus.
- Clinical Case Study (Wolpe, 1990):
  - A patient developed an irrational phobia of movie theaters.
  - Phase 1: Physical crowding ( $CS_1$ ) had previously been paired with physical assault or panic ( $US$ ), establishing a fear response ( $CR$ ) to crowds.
  - Phase 2: To avoid crowds, the individual attended movie theaters only in the middle of the day. On one occasion, the theater unexpectedly filled with students. The movie theater environment ( $CS_2$ ) was thus paired with crowds ( $CS_1$ ).
  - Outcome: The individual acquired severe conditioned fear to empty movie theaters ( $CS_2$ ), despite never experiencing physical injury ( $US$ ) within a theater.
- Procedural structure:
  - First-order conditioning: $CS_1 \rightarrow US$ , establishing a $CR$ to $CS_1$ .
  - Second-order conditioning: $CS_2 \rightarrow CS_1$ , establishing a $CR$ to $CS_2$ .
  - Third-order conditioning: $CS_3 \rightarrow CS_2$ , transferring associative control to a third level.

Procedure diagram for higher-order conditioning

* Factors dictating excitatory second-order conditioning versus conditioned inhibition:
    * Number of nonreinforced trials: Few presentations of  $CS_1$  and  $CS_2$  without the  $US$  yield excitatory second-order conditioning. Extensive nonreinforced training leads to conditioned inhibition (Yin, Barnet, & Miller, 1994).
    * Temporal arrangement: Sequential presentation ( $CS_2 \rightarrow CS_1$ ) facilitates excitatory second-order conditioning, whereas simultaneous compound presentation ( $CS_1 + CS_2$ ) promotes conditioned inhibition to  $CS_2$  (Stout, Escobar, & Miller, 2004; Wheeler, Sherwood, & Holland, 2008).
* Real-world applications:
    * Secondary reinforcement of money: Currency acts as a potent  $CS_1$  through historical associations with food, candy, and entertainment ( $US$ s). An uncle ( $CS_2$ ) who gifts money becomes positively conditioned via second-order conditioning.
    * Advertising: Commercial products ( $CS_2$ ) are paired with culturally established positive celebrities or imagery ( $CS_1$ ) to condition consumer preference (Schachtman, Walker, & Fowler, 2011).

Sensory Preconditioning:
- Establishes associations between two neutral conditioned stimuli ( $CS_1$ and $CS_2$ ) before either stimulus is associated with an unconditioned stimulus.
- Procedural structure:
  - Phase 1: Two neutral stimuli are paired together repeatedly ( $CS_2 \rightarrow CS_1$ ) in the absence of a $US$ . No overt conditioned response is displayed.
  - Phase 2: $CS_1$ is paired with a primary $US$ ( $CS_1 \rightarrow US$ ) until a robust $CR$ develops.
  - Test Phase: $CS_2$ is presented alone. $CS_2$ elicits the $CR$ despite having never been directly paired with the $US$ .

Procedure diagram for sensory preconditioning

Naturalistic example: Pastries often contain both vanilla ( $CS_2$ ) and cinnamon ( $CS_1$ ) flavors together. If food poisoning is later induced by consuming cinnamon alone in another food item, the individual will subsequently reject pure vanilla products due to the prior sensory association established between vanilla and cinnamon.
Empirical demonstrations: Documented by Dunsmoor, White, & LaBar (2011), Leising, Sawa, & Blaisdell (2007), and Rovee-Collier & Giles (2010).

Determinants of the Conditioned Response

The Unconditioned Stimulus as a Determinant:
- The nature of the unconditioned stimulus determines the general biological category of the conditioned response (e.g., food produces salivation; corneal air puffs produce eyeblinks).
- Fine-grained morphological fidelity (Jenkins & Moore, 1973):
  - Pigeons were conditioned with an illuminated response key paired with either solid grain or liquid water as the $US$ .
  - Grain $US$ : Pigeons directed pecks toward the key with high velocity, forceful impacts, and beaks open at the point of contact, mirroring food ingestion.
  - Water $US$ : Pigeons directed pecks toward the key that were slower, made with the beak closed, and accompanied by rhythmic swallowing movements, mirroring water drinking (Klein, LaMon, & Zeigler, 1983).
  - Corroborated in rats using solid food pellets versus liquid milk (Davey & Cleland, 1982; Davey, Phillips, & Cleland, 1981).
- Pavlov's Stimulus Substitution Model:
  - Classical conditioning constructs direct functional neural connections between the sensory input center of the $CS$ and the unconditioned response pathway of the $US$ .
  - The $CS$ becomes a neural surrogate for the $US$ , activating response pathways previously engaged solely by the $US$ .

Pavlov stimulus substitution model showing newly established neural connections

* Theoretical failure of strict stimulus substitution: In many preparations, the  $CR$  diverges completely from the  $UR$ . Foot shock ( $US$ ) reliably produces vigorous jumping, vocalization, and motor running, whereas a tone  $CS$  predictive of foot shock elicits behavioral freezing and motor arrest.

The Conditioned Stimulus as a Determinant:
- The physical properties, visual configurations, and ecological affordances of the $CS$ shape the topography of the conditioned response.
- Timberlake and Grant (1975) Rat Social Conditioning Study:
  - Setup: Food-deprived rats were trained with a live, gently restrained conspecific rat mounted on a movable platform that was introduced into the chamber via a flap door for $10\,s$ as the $CS$ immediately preceding food pellet delivery.

Apparatus for Timberlake and Grant live rat CS social conditioning experiment

    * Stimulus-substitution prediction: The experimental rat was predicted to direct feeding-related behaviors (gnawing, biting) toward the live stimulus rat  $CS$ .
    * Actual result: As conditioning progressed, the experimental rat directed social affiliative responses (orientation, approach, close sniffing, pawing) toward the stimulus rat. Unpaired control rats did not display these social behaviors.
    * Mechanistic implication: Social affiliative behavior was determined by the physical identity of the  $CS$  as a conspecific. When a retractable response lever acts as the  $CS$  for food, rats gnaw the lever or track the food cup, showing no social affiliative responding.
* Quail sexual conditioning: A three-dimensional taxidermic quail model elicits copulatory mounting, whereas a flat illuminated light cue elicits visual tracking and localized spatial approach.

The CS–US Interval as a Determinant:
- The temporal duration separating the $CS$ and the $US$ determines the temporal urgency of the response mode.
- Short $CS$ - $US$ intervals elicit immediate, proximate defensive or focal search actions. Long $CS$ - $US$ intervals elicit distant, preparatory, or exploratory behaviors.
- Empirical demonstration in quail sexual conditioning (Akins, 2000):
  - Protocols compared male quail conditioned with a $1\,\text{min}$ versus a $20\,\text{min}$ $CS$ - $US$ interval preceding access to a female.
  - Results:
    - $1\,\text{min}$ interval: Conditioned quail showed robust localized approach to the $CS$ ( $\approx 70\%$ of time spent near the $CS$ ) and negligible general locomotion ( $<1$ crossing).
    - $20\,\text{min}$ interval: Conditioned quail showed virtually no $CS$ approach ( $<5\%$ of time), but displayed high general locomotor behavior ( $>10$ crossings).

Effects of CS-US interval on CS approach versus general locomotor behavior in quail

Behavior Systems Theory:
- Developed by Timberlake (2001), Domjan (1997), and Rau & Fanselow (2007) to explain how $US$ , $CS$ , and temporal parameters interact.
- Biological systems are organized into sequential, hierarchical response modes:
  - General search behavior: Nondirected exploratory locomotion engaged when the animal has not localized a resource.
  - Focal search behavior: Spatially targeted orientation and approach triggered when resources or indicators are detected.
  - Consummatory behavior: Stereotyped species-specific manipulation (handling, chewing, copulatory clasping) upon physical contact with the goal object.
- Integration mechanism: Presentation of a $US$ activates its domain-specific behavior system (e.g., predatory/feeding, sexual, or defensive). The $CS$ becomes embedded within this behavioral sequence at a position dictated by the $CS$ - $US$ interval:
  - Short intervals embed the $CS$ within focal search or consummatory stages.
  - Long intervals embed the $CS$ within the earlier general search stage.

Behavior system response sequence illustrating CS insertion based on the CS-US interval

* Explanatory breadth: Accounts for food-directed conditioning (Silva & Timberlake, 1997) and fear/anxiety conditioning (Waddell, Morris, & Bouton, 2006). Anticipation is not represented by any single behavioral readout (Balsam et al., 2009).

Associative Structure: S–R Versus S–S Learning:
- Theoretical distinction:
  - Stimulus-Response ( $S$ - $R$ ) Learning: Conditioning establishes a direct automatic pathway between the $CS$ and the motor response ( $CR$ ). The $CS$ triggers the response directly without activating a representation of the $US$ .
  - Stimulus-Stimulus ( $S$ - $S$ ) Learning: Conditioning establishes an association between the $CS$ and a central cognitive representation of the $US$ . Responding to the $CS$ is mediated by the current motivational status of the $US$ representation.
- The US Devaluation Technique:
  - Phase 1: Mildly food-deprived rats receive conditioning where a tone $CS$ is paired with food pellets ( $US$ ), establishing a conditioned activity response.
  - Phase 2 (Devaluation): The experimental group receives ad libitum free food until fully satiated, reducing the current biological value of the food $US$ representation. The control group remains food-deprived.
  - Test Phase: The tone $CS$ is tested alone under nonreinforced conditions.
  - Predictions:
    - If learning is $S$ - $R$ , the $CR$ should persist unchanged in the experimental group because the direct $CS$ - $CR$ link is unaffected by hunger state.
    - If learning is $S$ - $S$ , the $CR$ should decline in the experimental group because the $CS$ activates an internal representation of food that is currently devalued.
  - Outcome (Holland & Rescorla, 1975): Satiated experimental rats showed significantly less conditioned responding to the $CS$ than hungry control rats, confirming $S$ - $S$ associative mediation.

Diagram of the US devaluation experimental strategy

Ubiquity: $S$ - $S$ learning mechanisms have been demonstrated across fear conditioning, appetitive conditioning, and sexual conditioning (Colwill & Motzkin, 1994; Delamater et al., 2006; Storsve, McNally, & Richardson, 2012).

Modification of the Unconditioned Response

Conditioned Diminution of the UR:
- Classical conditioning modifies the behavioral and physiological reaction to the unconditioned stimulus itself.
- Methodologies for evaluating conditioned modification of the unconditioned response:
  - Common testing design: Experimental group receives paired training ( $A \rightarrow US$ ); control group receives unpaired training ( $A / US$ ). Both groups are tested with $A \rightarrow US$ , and the response to the $US$ is recorded.
  - Common training design: All subjects receive paired training with one stimulus ( $A \rightarrow US$ ) and unpaired training with a second stimulus ( $B / US$ ). Responses to the $US$ are tested and compared following presentations of $A$ versus $B$ .

* Eyeblink attenuation (Kimble & Ost, 1961): Presentation of a conditioned light  $CS$  prior to a corneal air-puff  $US$  significantly attenuates the vigor and amplitude of the unconditioned blink response.

Conditioned Analgesia:
- Physical injury or painful foot shock activates the endogenous opiate system (endorphins and enkephalins), blunting nociception to permit active escape and defense.
- When cues predictive of an aversive shock are presented, the $CS$ triggers an anticipatory release of endogenous opiates, causing conditioned analgesia.
- Consequently, foot-shock becomes progressively less painful over successive trials as conditioned diminution of the $UR$ develops (Zelikowsky & Fanselow, 2011).
Conditioned Drug Tolerance:
- Pharmacological tolerance refers to the reduction in drug effectiveness following repeated administrations, necessitating higher doses to produce equivalent effects.
- Pavlovian conditioning model of drug tolerance (Siegel, 1999, 2005, 2008):
  - $CS$ : Environmental administration context, drug paraphernalia, odors, and ritual cues.
  - $US$ : Direct pharmacological and physiological impact of the drug on bodily systems.
  - Opponent-Process mechanism: A drug generates a primary unconditioned disturbance ( $a\text{ process}$ ) that triggers an opposing homeostatic compensatory adjustment ( $b\text{ process}$ ).
  - Conditioned Compensatory Responses: Through repeated pairings, drug administration cues ( $CS$ ) elicit the compensatory $b\text{ process}$ in advance of the drug's delivery. When the drug is consumed, this conditioned compensatory response cancels out the drug's effect, producing a net attenuated response (tolerance).

Conditioning model of drug tolerance showing primary reaction, conditioned compensatory response, and net attenuated response

Clinical Implications for Drug Overdose (Box 4.1):
- An individual who routinely injects high doses of heroin in a familiar environment with regular associates is protected by localized conditioned compensatory responses.
- If the individual injects that identical high dose in a novel geographical location or with unfamiliar acquaintances, the predictive drug $CS$ cues are absent.
- Without anticipatory conditioned compensatory responses, the

Determinants of Effective Conditioned and Unconditioned Stimuli

Initial Behavioral Responses to Stimuli:
- Pavlovian definitions of stimuli:
  - Conditioned Stimulus ( $CS$ ): A stimulus that does not initially elicit the target response, but comes to do so as a direct result of being associated with an unconditioned stimulus.
  - Unconditioned Stimulus ( $US$ ): A stimulus that reliably and unconditionally elicits the target response from the outset, requiring no special prior training.
- Relativity of stimulus identification:
  - Whether a stimulus functions as a $CS$ or a $US$ is determined by its relationship to the other environmental stimuli in the conditioning arrangement.
  - Example of sucrose-flavored food pellets:
    - Operates as a $CS$ when paired with lithium-induced illness or radiation in a conditioned taste-aversion procedure, leading to an aversion to consuming the pellets.
    - Operates as a $US$ in a sign-tracking experiment (Franklin & Hall, 2011), where insertion of a response lever serves as the $CS$ preceding sucrose delivery, eliciting lever pressing and approach.
Stimulus Novelty and Familiarity Effects:
- Novel stimuli elicit stronger behavioral reactions and enter into classical conditioning associations far more rapidly than familiar stimuli.
- The Latent-Inhibition (CS-Preexposure) Effect:
  - Repeated nonreinforced exposure to a stimulus prior to conditioning impairs that stimulus's subsequent ability to serve as an effective $CS$ (Hall, 1991; Lubow & Weiner, 2010).
  - Experimental procedure involves two phases:
    - Phase 1 (Preexposure phase): The designated $CS$ is repeatedly presented alone without consequences, rendering it familiar, benign, and irrelevant.
    - Phase 2 (Conditioning phase): The preexposed $CS$ is paired with a $US$ using conventional Pavlovian procedures.
  - Behavioral outcome: Participants acquire conditioned responding substantially slower than control participants exposed to a novel $CS$ (De la Casa, Marquez, & Lubow, 2009).
  - Distinction from conditioned inhibition: Conditioned inhibition involves active signaling of the absence of an expected $US$ , whereas latent inhibition reflects an attenuation of attentional processing toward a stimulus with no predictive significance.
  - Evolutionary rationale: Protects organisms from cognitive information overload by suppressing attention to biologically irrelevant stimuli (Lubow, 2011).
  - Historical origin: Discovered in sheep by Lubow and Moore (1959).
  - Relevance to schizophrenia:
    - A key symptom of schizophrenia is the cognitive deficit in filtering out irrelevant ambient stimuli.
    - Patients with schizophrenia show a disruption or absence of the latent-inhibition effect.
    - Latent inhibition and schizophrenia share dopaminergic neurobiological substrates, making latent-inhibition paradigms a primary behavioral assay for evaluating cognitive dysfunctions in schizophrenia.
- The US-Preexposure Effect:
  - Prior familiarization with an unconditioned stimulus in the absence of a predictive conditioned stimulus severely delays subsequent associative conditioning when that $US$ is paired with a novel $CS$ .
  - Empirical study (Franklin & Hall, 2011):
    - Setup: Rats received $80$ preexposure deliveries per day of one food pellet flavor (bacon or chocolate) for $5$ consecutive days, totaling $400$ $US$ preexposure trials in a food cup flanked by two retractable levers.
    - Conditioning: $20$ Pavlovian trials were administered where insertion of one lever for $10\,s$ signaled the delivery of the bacon pellet, and insertion of the other lever signaled the chocolate pellet.
    - Results: Conditioning of the sign-tracking lever contact response proceeded significantly faster for the lever paired with the novel pellet flavor than for the lever paired with the familiar pellet flavor.
  - Generality: Documented extensively in appetitive conditioning, fear conditioning, taste-aversion conditioning, and drug conditioning (Randich & LoLordo, 1979; Hall, 2009).

Stimulus Intensity and Salience:
- Conditioning proceeds more rapidly and generates more vigorous conditioned responding when intense, highly salient conditioned and unconditioned stimuli are utilized (Bevins et al., 1997; Kamin, 1965; Scavio & Gormezano, 1974).
- Psychological stimulus salience corresponds to physical noticeability and biological significance (McLaren & Mackintosh, 2000; Pearce & Hall, 1980).
- Methods for manipulating salience:
  - Physical parameter alteration: Escalating stimulus amplitude, brightness, or volume.
  - Biological state manipulation: Inducing nutritional or physiological deficits makes specific stimuli more salient (Krieckhaus & Wolf, 1968). For instance, sodium-deficient rats learn significantly stronger conditioned aversions to salt flavors than sodium-replete controls (Sawa, Nakajima, & Imada, 1999).
  - Naturalistic stimulus configuration: Making conditioned stimuli structurally closer to stimuli encountered in natural ecologies.
    - Sexual conditioning in domesticated quail (Cusato & Domjan, 1998, 2012; Domjan et al., 2004): Male quail receive access to a female bird ( $US$ ) signaled by a $CS$ .
    - An arbitrary $CS$ (a foam block covered in terrycloth) produces slower, less durable conditioning.
    - A naturalistic $CS$ (a terrycloth block incorporating the taxidermically prepared head and neck feathers of a female quail) produces rapid acquisition, engages comprehensive copulatory behavior, and resists associative decay when the $CS$ - $US$ interval is widened.

CS–US Relevance and Belongingness:
- Genetic predispositions and evolutionary constraints dictate which stimulus combinations are rapidly associated, violating the assumption of equipotentiality.
- Garcia and Koelling's (1966) "Bright-Noisy-Tasty Water" Experiment:
  - Procedure: Water-deprived laboratory rats drank from a drinking tube filled with flavored water (salty or sweet; taste $CS$ ). Each lick also triggered a brief flash of light and an audible click (audiovisual $CS$ ). Thus, taste and audiovisual cues occurred simultaneously.
  - Conditioning: One group of rats received a brief foot-shock ( $US$ causing peripheral cutaneous pain); a second group received ionizing radiation or a toxic drug injection ( $US$ producing visceral illness).
  - Testing: Cues were separated. Rats were tested for drinking suppression either with flavored water alone (no audiovisual feedback) or with unflavored water paired with the lick-triggered light and click.
  - Results:
    - Rats conditioned with illness suppressed drinking substantially when tested with the taste $CS$ , but drank normally when tested with the audiovisual $CS$ .
    - Rats conditioned with shock suppressed drinking substantially when tested with the audiovisual $CS$ , but drank normally when tested with the taste $CS$ .
  - Conclusions: Learning is determined by the ecological relevance or belongingness between the $CS$ and $US$ . Cutaneous pain selectively associates with external audiovisual cues, whereas internal gut distress selectively associates with gustatory cues.

Developmental and Evolutionary Generality:
- Innate nature: The selective taste-sickness association is present in laboratory rats $1\,\text{day}$ after birth, demonstrating that individual trial-and-error experience is unnecessary for stimulus-relevance effects (Gemberling & Domjan, 1982).
- Human food aversions: People associate nausea with ingested tastes rather than dining room ambient sounds or lighting (Logue et al., 1981; Pelchat & Rozin, 1982).
- Primate fear acquisition (Öhman & Mineka, 2001; Mineka & Öhman, 2002): Rhesus macaques and humans acquire conditioned fear substantially faster to fear-relevant stimuli (images of snakes) than to fear-irrelevant stimuli (images of flowers or mushrooms). This selective bias operates below conscious awareness (Öhman et al., 2007). Human infants aged $8$ to $14\,\text{months}$ orient significantly faster to snake pictures than to flower pictures (LoBue & DeLoache, 2010).

Classical Conditioning Without an Unconditioned Stimulus

Higher-Order Conditioning:
- Enables learning to occur in the absence of a primary biological reinforcer by using a previously conditioned stimulus to condition a novel stimulus.
- Clinical Case Study (Wolpe, 1990):
  - A patient developed an irrational phobia of movie theaters.
  - Phase 1: Physical crowding ( $CS_1$ ) had previously been paired with physical assault or panic ( $US$ ), establishing a fear response ( $CR$ ) to crowds.
  - Phase 2: To avoid crowds, the individual attended movie theaters only in the middle of the day. On one occasion, the theater unexpectedly filled with students. The movie theater environment ( $CS_2$ ) was thus paired with crowds ( $CS_1$ ).
  - Outcome: The individual acquired severe conditioned fear to empty movie theaters ( $CS_2$ ), despite never experiencing physical injury ( $US$ ) within a theater.
- Procedural structure:
  - First-order conditioning: $CS_1 \rightarrow US$ , establishing a $CR$ to $CS_1$ .
  - Second-order conditioning: $CS_2 \rightarrow CS_1$ , establishing a $CR$ to $CS_2$ .
  - Third-order conditioning: $CS_3 \rightarrow CS_2$ , transferring associative control to a third level.

- Factors dictating excitatory second-order conditioning versus conditioned inhibition:

    - Number of nonreinforced trials: Few presentations of  $CS_1$  and  $CS_2$  without the  $US$  yield excitatory second-order conditioning. Extensive nonreinforced training leads to conditioned inhibition (Yin, Barnet, & Miller, 1994).

    - Temporal arrangement: Sequential presentation ( $CS_2 \rightarrow CS_1$ ) facilitates excitatory second-order conditioning, whereas simultaneous compound presentation ( $CS_1 + CS_2$ ) promotes conditioned inhibition to  $CS_2$  (Stout, Escobar, & Miller, 2004; Wheeler, Sherwood, & Holland, 2008).

- Real-world applications:

    - Secondary reinforcement of money: Currency acts as a potent  $CS_1$  through historical associations with food, candy, and entertainment ( $US$ s). An uncle ( $CS_2$ ) who gifts money becomes positively conditioned via second-order conditioning.

    - Advertising: Commercial products ( $CS_2$ ) are paired with culturally established positive celebrities or imagery ( $CS_1$ ) to condition consumer preference (Schachtman, Walker, & Fowler, 2011).

Sensory Preconditioning:
- Establishes associations between two neutral conditioned stimuli ( $CS_1$ and $CS_2$ ) before either stimulus is associated with an unconditioned stimulus.
- Procedural structure:
  - Phase 1: Two neutral stimuli are paired together repeatedly ( $CS_2 \rightarrow CS_1$ ) in the absence of a $US$ . No overt conditioned response is displayed.
  - Phase 2: $CS_1$ is paired with a primary $US$ ( $CS_1 \rightarrow US$ ) until a robust $CR$ develops.
  - Test Phase: $CS_2$ is presented alone. $CS_2$ elicits the $CR$ despite having never been directly paired with the $US$ .

Naturalistic example: Pastries often contain both vanilla ( $CS_2$ ) and cinnamon ( $CS_1$ ) flavors together. If food poisoning is later induced by consuming cinnamon alone in another food item, the individual will subsequently reject pure vanilla products due to the prior sensory association established between vanilla and cinnamon.
Empirical demonstrations: Documented by Dunsmoor, White, & LaBar (2011), Leising, Sawa, & Blaisdell (2007), and Rovee-Collier & Giles (2010).

Determinants of the Conditioned Response

The Unconditioned Stimulus as a Determinant:
- The nature of the unconditioned stimulus determines the general biological category of the conditioned response (e.g., food produces salivation; corneal air puffs produce eyeblinks).
- Fine-grained morphological fidelity (Jenkins & Moore, 1973):
  - Pigeons were conditioned with an illuminated response key paired with either solid grain or liquid water as the $US$ .
  - Grain $US$ : Pigeons directed pecks toward the key with high velocity, forceful impacts, and beaks open at the point of contact, mirroring food ingestion.
  - Water $US$ : Pigeons directed pecks toward the key that were slower, made with the beak closed, and accompanied by rhythmic swallowing movements, mirroring water drinking (Klein, LaMon, & Zeigler, 1983).
  - Corroborated in rats using solid food pellets versus liquid milk (Davey & Cleland, 1982; Davey, Phillips, & Cleland, 1981).
- Pavlov's Stimulus Substitution Model:
  - Classical conditioning constructs direct functional neural connections between the sensory input center of the $CS$ and the unconditioned response pathway of the $US$ .
  - The $CS$ becomes a neural surrogate for the $US$ , activating response pathways previously engaged solely by the $US$ .

- Theoretical failure of strict stimulus substitution: In many preparations, the  $CR$  diverges completely from the  $UR$ . Foot shock ( $US$ ) reliably produces vigorous jumping, vocalization, and motor running, whereas a tone  $CS$  predictive of foot shock elicits behavioral freezing and motor arrest.

The Conditioned Stimulus as a Determinant:
- The physical properties, visual configurations, and ecological affordances of the $CS$ shape the topography of the conditioned response.
- Timberlake and Grant (1975) Rat Social Conditioning Study:
  - Setup: Food-deprived rats were trained with a live, gently restrained conspecific rat mounted on a movable platform that was introduced into the chamber via a flap door for $10\,s$ as the $CS$ immediately preceding food pellet delivery.

    - Stimulus-substitution prediction: The experimental rat was predicted to direct feeding-related behaviors (gnawing, biting) toward the live stimulus rat  $CS$ .

    - Actual result: As conditioning progressed, the experimental rat directed social affiliative responses (orientation, approach, close sniffing, pawing) toward the stimulus rat. Unpaired control rats did not display these social behaviors.

    - Mechanistic implication: Social affiliative behavior was determined by the physical identity of the  $CS$  as a conspecific. When a retractable response lever acts as the  $CS$  for food, rats gnaw the lever or track the food cup, showing no social affiliative responding.

- Quail sexual conditioning: A three-dimensional taxidermic quail model elicits copulatory mounting, whereas a flat illuminated light cue elicits visual tracking and localized spatial approach.

The CS–US Interval as a Determinant:
- The temporal duration separating the $CS$ and the $US$ determines the temporal urgency of the response mode.
- Short $CS$ - $US$ intervals elicit immediate, proximate defensive or focal search actions. Long $CS$ - $US$ intervals elicit distant, preparatory, or exploratory behaviors.
- Empirical demonstration in quail sexual conditioning (Akins, 2000):
  - Protocols compared male quail conditioned with a $1\,\text{min}$ versus a $20\,\text{min}$ $CS$ - $US$ interval preceding access to a female.
  - Results:
    - $1\,\text{min}$ interval: Conditioned quail showed robust localized approach to the $CS$ ( $\approx 70\%$ of time spent near the $CS$ ) and negligible general locomotion ( $<1$ crossing).
    - $20\,\text{min}$ interval: Conditioned quail showed virtually no $CS$ approach ( $<5\%$ of time), but displayed high general locomotor behavior ( $>10$ crossings).

Behavior Systems Theory:
- Developed by Timberlake (2001), Domjan (1997), and Rau & Fanselow (2007) to explain how $US$ , $CS$ , and temporal parameters interact.
- Biological systems are organized into sequential, hierarchical response modes:
  - General search behavior: Nondirected exploratory locomotion engaged when the animal has not localized a resource.
  - Focal search behavior: Spatially targeted orientation and approach triggered when resources or indicators are detected.
  - Consummatory behavior: Stereotyped species-specific manipulation (handling, chewing, copulatory clasping) upon physical contact with the goal object.
- Integration mechanism: Presentation of a $US$ activates its domain-specific behavior system (e.g., predatory/feeding, sexual, or defensive). The $CS$ becomes embedded within this behavioral sequence at a position dictated by the $CS$ - $US$ interval:
  - Short intervals embed the $CS$ within focal search or consummatory stages.
  - Long intervals embed the $CS$ within the earlier general search stage.

- Explanatory breadth: Accounts for food-directed conditioning (Silva & Timberlake, 1997) and fear/anxiety conditioning (Waddell, Morris, & Bouton, 2006). Anticipation is not represented by any single behavioral readout (Balsam et al., 2009).

Associative Structure: S–R Versus S–S Learning:
- Theoretical distinction:
  - Stimulus-Response ( $S$ - $R$ ) Learning: Conditioning establishes a direct automatic pathway between the $CS$ and the motor response ( $CR$ ). The $CS$ triggers the response directly without activating a representation of the $US$ .
  - Stimulus-Stimulus ( $S$ - $S$ ) Learning: Conditioning establishes an association between the $CS$ and a central cognitive representation of the $US$ . Responding to the $CS$ is mediated by the current motivational status of the $US$ representation.
- The US Devaluation Technique:
  - Phase 1: Mildly food-deprived rats receive conditioning where a tone $CS$ is paired with food pellets ( $US$ ), establishing a conditioned activity response.
  - Phase 2 (Devaluation): The experimental group receives ad libitum free food until fully satiated, reducing the current biological value of the food $US$ representation. The control group remains food-deprived.
  - Test Phase: The tone $CS$ is tested alone under nonreinforced conditions.
  - Predictions:
    - If learning is $S$ - $R$ , the $CR$ should persist unchanged in the experimental group because the direct $CS$ - $CR$ link is unaffected by hunger state.
    - If learning is $S$ - $S$ , the $CR$ should decline in the experimental group because the $CS$ activates an internal representation of food that is currently devalued.
  - Outcome (Holland & Rescorla, 1975): Satiated experimental rats showed significantly less conditioned responding to the $CS$ than hungry control rats, confirming $S$ - $S$ associative mediation.

Ubiquity: $S$ - $S$ learning mechanisms have been demonstrated across fear conditioning, appetitive conditioning, and sexual conditioning (Colwill & Motzkin, 1994; Delamater et al., 2006; Storsve, McNally, & Richardson, 2012).

Modification of the Unconditioned Response

Conditioned Diminution of the UR:
- Classical conditioning modifies the behavioral and physiological reaction to the unconditioned stimulus itself.
- Methodologies for evaluating conditioned modification of the unconditioned response:
  - Common testing design: Experimental group receives paired training ( $A \rightarrow US$ ); control group receives unpaired training ( $A / US$ ). Both groups are tested with $A \rightarrow US$ , and the response to the $US$ is recorded.
  - Common training design: All subjects receive paired training with one stimulus ( $A \rightarrow US$ ) and unpaired training with a second stimulus ( $B / US$ ). Responses to the $US$ are tested and compared following presentations of $A$ versus $B$ .

- Eyeblink attenuation (Kimble & Ost, 1961): Presentation of a conditioned light  $CS$  prior to a corneal air-puff  $US$  significantly attenuates the vigor and amplitude of the unconditioned blink response.

Conditioned Analgesia:
- Physical injury or painful foot shock activates the endogenous opiate system (endorphins and enkephalins), blunting nociception to permit active escape and defense.
- When cues predictive of an aversive shock are presented, the $CS$ triggers an anticipatory release of endogenous opiates, causing conditioned analgesia.
- Consequently, foot-shock becomes progressively less painful over successive trials as conditioned diminution of the $UR$ develops (Zelikowsky & Fanselow, 2011).
Conditioned Drug Tolerance:
- Pharmacological tolerance refers to the reduction in drug effectiveness following repeated administrations, necessitating higher doses to produce equivalent effects.
- Pavlovian conditioning model of drug tolerance (Siegel, 1999, 2005, 2008):
  - $CS$ : Environmental administration context, drug paraphernalia, odors, and ritual cues.
  - $US$ : Direct pharmacological and physiological impact of the drug on bodily systems.
  - Opponent-Process mechanism: A drug generates a primary unconditioned disturbance ( $a\text{ process}$ ) that triggers an opposing homeostatic compensatory adjustment ( $b\text{ process}$ ).
  - Conditioned Compensatory Responses: Through repeated pairings, drug administration cues ( $CS$ ) elicit the compensatory $b\text{ process}$ in advance of the drug's delivery. When the drug is consumed, this conditioned compensatory response cancels out the drug's effect, producing a net attenuated response (tolerance).

Clinical Implications for Drug Overdose (Box 4.1):
- An individual who routinely injects high doses of heroin in a familiar environment with regular associates is protected by localized conditioned compensatory responses.
- If the individual injects that identical high dose in a novel geographical location or with unfamiliar acquaintances, the predictive drug $CS$ cues are absent.
- Without anticipatory conditioned compensatory responses, the full pharmacological impact of the drug is experienced without homeostatic counteraction, leading to potential overdose or death from a previously tolerated dose (Siegel et al., 2000).
Conditioned Reproduction and Fertility:
- Pavlovian conditioning alters physiological and behavioral responses to appetitive unconditioned stimuli, conferring substantial reproductive advantages.
- Blue Gourami Fish Study (Hollis et al., 1997):
  - Setup: Male fish received Pavlovian pairings where a visual light $CS$ ( $10\,s$ ) signaled visual access to a female behind a clear barrier ( $US$ ). Unpaired controls received the stimuli separately.
  - Test: The physical barrier was removed for $6\,\text{days}$ , allowing unrestrained mating interaction.
  - Results: Males receiving the paired $CS$ exhibited significantly less aggression toward the female, spent markedly more time building bubble nests, achieved more copulatory clasps, and produced far greater numbers of viable offspring.
- Domesticated Quail Studies (Domjan & Akins, 2011; Domjan, Mahometa, & Matthews, 2012):
  - Exposure to a sexually conditioned $CS$ increases female receptivity to copulation.
  - Males demonstrate accelerated copulatory execution, elevated cloacal gland foam secretion, greater sperm transfer, and higher egg fertilization rates compared to unpaired controls.

How Conditioned and Unconditioned Stimuli Become Associated

The Blocking Effect:
- Discovery and Definition (Kamin, 1968, 1969):
  - Demonstrates that temporal contiguity between a $CS$ and a $US$ is insufficient for associative learning to occur.
  - Phenomenon wherein prior conditioning of stimulus $A$ with a $US$ prevents or impairs the conditioning of an added stimulus $B$ when the compound $AB$ is subsequently paired with the same $US$ .
- Experimental Design:
  - Phase 1: Experimental group receives extensive training with $A \rightarrow US$ until conditioned responding reaches asymptote. The control group receives $A$ and the $US$ in an unpaired fashion.
  - Phase 2: Both groups receive compound conditioning trials where $A$ and $B$ are presented simultaneously and paired with the $US$ ( $AB \rightarrow US$ ).
  - Test Phase: Stimulus $B$ is presented alone. The experimental group exhibits negligible conditioned responding to $B$ , whereas the control group shows robust conditioned responding.

- Empirical Replications:

    - Human geometric cue learning (Prados, 2011): Students learned to locate target corners in geometric floor plans. Prior conditioning with triangle  $A$  blocked acquisition of spatial information regarding triangle  $B$  when presented as a compound ( $20\%$  correct responses in the blocking group versus  $79\%$  and  $92\%$  in control conditions).

    - Reading acquisition in children (Singh & Solman, 1990; Box 4.2): Presenting known pictorial prompts simultaneously with unfamiliar printed words produced a blocking effect, impeding the acquisition of word-reading skills compared to text-only presentation.

- Kamin's Surprise Hypothesis: Conditioning occurs only when an unconditioned stimulus is surprising or unexpected. When a  $US$  is fully predicted by a previously conditioned stimulus ( $A$ ), it activates minimal cognitive processing or "mental effort," blocking associative attachment to the redundant cue ( $B$ ).

The Rescorla–Wagner Model:
- Conceptual Foundation (Rescorla & Wagner, 1972):
  - Formal mathematical theory of conditioning based on the principle that the effectiveness of a $US$ depends on how surprising it is.
  - An event is surprising if it deviates from expectation: an unexpectedly large $US$ drives excitatory conditioning, while an unexpectedly small or absent $US$ drives inhibitory conditioning.
  - Conditioned expectation is represented by the summed associative strength of all cues preceding the $US$ .
- The Delta Rule Equation:

$\Delta V = k(\lambda - V)$

    -  $\Delta V$ : The change in the associative value of a stimulus on a given trial.

    -  $k$ : A fractional constant representing the combined salience of the  $CS$  and  $US$ .

    -  $\lambda$ : The asymptote of conditioning supported by the actual  $US$  delivered.

    -  $V$ : The total associative value of all predictive cues present on the trial ( $V = \sum V_i$ ).

    -  $(\lambda - V)$ : The prediction error or degree of surprise.

    - Learning operates as an error-correction mechanism: as  $V$  approaches  $\lambda$ , prediction error shrinks to zero and associative learning ceases.

Growth of associative value during conditioning until reaching asymptote

- Explanations of Core Phenomena:

    - Mathematical account of blocking: In Phase 1, stimulus  $A$  is conditioned to asymptote ( $V_A = \lambda$ ). In Phase 2, compound presentation yields total expectation  $V = V_A + V_B = \lambda + 0 = \lambda$ . The change in associative strength for stimulus  $B$  is therefore:

$\Delta V_B = k(\lambda - \lambda) = k(0) = 0$

        - Stimulus  $B$  acquires zero associative value because the  $US$  is completely predicted by stimulus  $A$ .

    - Loss of associative value despite reinforcement (Overexpectation):

        - Phase 1: Stimuli  $A$  and  $B$  are trained independently to asymptote with a single-pellet  $US$  ( $V_A = \lambda$ ,  $V_B = \lambda$ ).

        - Phase 2: Stimuli are presented simultaneously ( $AB$ ) and paired with the same single-pellet  $US$ .

        - Expected outcome is double the actual delivery ( $V_{AB} = V_A + V_B = 2\lambda$ ).

        - The prediction error becomes negative ( $\lambda - 2\lambda = -\lambda$ ), causing both stimuli to lose associative strength despite continual pairings with the  $US$ .

    - Conditioned inhibition: On nonreinforced compound trials ( $CS^+ + CS^- \rightarrow \text{no } US$ ), the presence of  $CS^+$  creates an expectation of the  $US$ . Because  $\lambda = 0$ , achieving zero net error requires that the associative value of  $CS^-$  becomes negative ( $V_{CS-} < 0$ ), formalizing a conditioned inhibitor as a negative associative cue.

    - Extinction: Repeated nonreinforced presentation of a conditioned excitor produces a negative prediction error, systematically driving its positive associative strength back toward zero.

- Empirical Problems and Theoretical Failures:

    - Erroneous prediction of inhibitor extinction: The model predicts that repeated isolated presentations of a conditioned inhibitor ( $CS^-$ ) should extinguish its inhibition toward zero. Empirically, presenting  $CS^-$  alone does not diminish its inhibitory properties.

    - Extinction of excitation unmasks inhibition loss: Extinguishing the excitatory  $CS^+$  with which the inhibitor was trained diminishes conditioned inhibition, contrary to the model's formulation.

    - Extinction is not unlearning: The model treats extinction as the erasure of associative value back to zero, whereas empirical evidence indicates extinction represents new inhibitory learning that leaves the original memory trace intact.

Neurobiological Mechanisms of Prediction Error (Box 4.3):
- Neural Circuitry of Fear Conditioning:
  - Basolateral Amygdala (BLA; lateral and basal nuclei): Serves as the primary convergence site where sensory signals from the $CS$ (relayed via the sensory thalamus and cortex) converge with noxious sensory information from the $US$ .
  - Plasticity mechanisms: Synaptic strengthening within the basolateral amygdala requires the activation of $NMDA$ receptor gated channels triggered by strong $US$ inputs.
  - Central Nucleus of the Amygdala ( $CeA$ ): Receives processed information from the BLA and serves as the major output coordinator, projecting to midbrain and brainstem effector sites to drive defensive behavioral responses (freezing, acoustic startle enhancement, cardiovascular adjustments).
  - Periaqueductal Gray ( $PAG$ ):
    - Ventral PAG ( $vlPAG$ ): Mediates conditioned behavioral freezing and immobility.
    - Dorsolateral PAG ( $dlPAG$ ): Coordinates circa-strike active defensive reactions (fight-or-flight behaviors).

Neural components of fear conditioning and prediction error computation in the PAG

- Computation of the Error Signal:

    - The  $vlPAG$  acts as a comparative comparator hub for error signaling.

    - It receives ascending sensory nociceptive signals ( $\lambda$ ) from the spinal cord and descending expectancy signals ( $V$ ) from the central amygdala.

    - Conditioned expectation triggers the release of endogenous opioids in the  $vlPAG$ , muting incoming pain signals and reducing subsequent learning in the amygdala.

    - Microinjecting opioid antagonists into the  $vlPAG$  blocks this analgesia and unblocks fear conditioning to an otherwise fully expected shock  $US$ , confirming that the descending output reflects prediction error ( $\lambda - V$ ).

- Distinct Appetitive Amygdalar Functions: Basolateral lesions impair an appetitive  $CS$  from functioning as a secondary reinforcer in second-order conditioning, while central amygdalar lesions eliminate conditioned orienting and surprise-induced attentional modulation.

Attentional Models of Conditioning:
- Conceptual Contrast: Whereas the Rescorla–Wagner model focuses on trial-by-trial variations in $US$ effectiveness, attentional theories postulate that learning is modulated by the degree of cognitive attention allocated to the $CS$ .
- Single-Mechanism Formulations:
  - Mackintosh (1975): Organisms attend selectively to stimuli that are reliable, informative predictors of biologically significant outcomes, ignoring non-predictive cues.
  - Pearce and Hall (1980): Attention is allocated to stimuli whose consequences are uncertain. If a $US$ is surprising on trial $n$ , attention to the preceding $CS$ increases on trial $n+1$ ("looking for learning"). Conversely, fully predicted outcomes attenuate attention on subsequent trials.
- Multi-Category Attentional Framework (Hogarth, Dickinson, & Duka, 2011):
  - Looking for action: Directs attention toward reliable predictive cues to guide automatic, low-effort behavioral responses.
  - Looking for learning: Directs attention toward ambiguous or novel cues with unresolved predictive validity to facilitate new associative learning.
  - Looking for liking: Allocates attentional resources toward cues on the basis of their acquired emotional valence.
- Theoretical Limitation: Attentional theories postulate proactive attentional shifts across successive trials and therefore fail to account for one-trial blocking, where learning about a novel stimulus is blocked on its very first compound trial.
Timing and Information Theory Models:
- Temporal Coding Hypothesis (Balsam, Gallistel, & Miller):
  - Classical conditioning involves encoding precise temporal intervals between events rather than merely linking stimuli associatively.
  - Animals learn not only that a $US$ will occur, but precisely when it will arrive relative to $CS$ onset.
- The Relative-Waiting-Time Hypothesis and the I/T Ratio:
  - Conditioned responding depends critically on the ratio between the intertrial interval ( $I$ ) and the trial duration or $CS$ - $US$ interval ( $T$ ).
  - Empirical findings (Holland, 2000): In rat appetitive conditioning, goal-tracking responding to an auditory $CS$ is a direct linear function of the $I/T$ ratio, irrespective of whether the absolute $CS$ length is $10\,s$ or $20\,s$ .

Conditioned responding as a function of the I/T ratio

    - Informational interpretation: High  $I/T$  ratios mean waiting for the  $US$  during the  $CS$  is substantially shorter than waiting during the intertrial baseline, rendering the  $CS$  highly informative. When the ratio is low, the  $CS$  provides negligible temporal information relative to background context.

    - Limitation: Primarily applies to multi-trial learning schedules; fails to provide a compelling model for single-trial learning paradigms (e.g., one-trial taste aversion).

The Comparator Hypothesis:
- Core Premise (Miller and colleagues):
  - Conditioned responding is governed by retrieval and performance processes rather than variations in associative learning.
  - All co-occurring stimuli form excitatory links during training; performance deficits (such as blocking) reflect a failure to express learning rather than a failure of acquisition.
- Three-Link Associative Architecture:
  - Link 1: Direct excitatory association between the target conditioned stimulus ( $X$ ) and the unconditioned stimulus ( $US$ ).
  - Link 2: Excitatory within-compound association between the target conditioned stimulus ( $X$ ) and companion comparator cues ( $C$ ).
  - Link 3: Excitatory association between the comparator cues ( $C$ ) and the unconditioned stimulus ( $US$ ).

Associative three-link structure of the comparator hypothesis

Performance Comparator Balance:
- Upon testing target $CS$ ( $X$ ), the representation of the $US$ is activated directly via Link 1 and indirectly via Links 2 and 3.
- Conditioned excitation occurs if the direct Link 1 activation exceeds the indirect activation through comparator cues.
- Conditioned inhibition occurs if indirect activation via comparator cues surpasses the direct Link 1 activation.

Comparator balance model determining excitatory versus inhibitory output

Post-Training Revaluation Effects:
- Unmasking of blocked responding: Presenting the blocking stimulus ( $CS_A$ ) alone without reinforcement following compound conditioning extinguishes Link 3, unmasking robust conditioned responding to the target $CS_B$ .
- Reduction of conditioned inhibition: Extinguishing the companion excitatory comparator ( $CS^+$ ) reduces indirect activation, significantly diminishing the inhibitory potency of the target $CS^-$ .