* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Classical Conditioning
Learning theory (education) wikipedia , lookup
Neuroeconomics wikipedia , lookup
Psychophysics wikipedia , lookup
Behaviorism wikipedia , lookup
Psychoneuroimmunology wikipedia , lookup
Psychological behaviorism wikipedia , lookup
Operant conditioning wikipedia , lookup
BrainSoft.ir Classical Conditioning ROBERT E. CLARK University of California, San Diego I. Introduction unconditioned responses do not show much habituation. In other words, repeated presentations of the unconditioned stimulus will continue to induce unconditioned responses. Alpha responses quickly habituate, which means that after a few presentations of the conditioned stimulus, they no longer occur. Ideally, conditioned stimuli are chosen because they are neutral (i.e., do not initially elicit a response). However, sometimes it is preferable to use a particular conditioned stimulus even if it temporarily results in alpha responses. II. Brief History III. The Distinction between Classical and Instrumental/ Operant Conditioning IV. Origin of the Term Classical Conditioning V. Types of Classical Conditioning VI. Acquisition of the CR amplitude The magnitude of the conditioned response or unconditioned response. For example, in eyeblink conditioning the amplitude is expressed as the distance the eyelid moves in millimeters. Accordingly, a conditioned response of 2 mm has a greater amplitude than a conditioned response of only 1 mm. VII. Eyeblink Classical Conditioning in Humans VIII. Problems for Studies of Human Eyeblink Classical Conditioning IX. The Emergence of Animal Studies of Classical Conditioning conditioned response A learned response that is elicited by the conditioned stimulus. X. Eyeblink Classical Condition as a Tool to Study Brain Function conditioned stimulus A stimulus that signals the unconditioned stimulus. Initially the conditioned stimulus does not cause a response, but eventually it elicits a conditioned response. XI. The Search for the Engram conditioned stimulus-alone test trial A conditioned stimulusalone test trial is a trial in which the unconditioned stimulus is omitted, leaving only the conditioned stimulus presentation. Conditioned stimulus-alone test trials are sometimes used because the presentation of the unconditioned stimulus and subsequent unconditioned response can obscure or mask the conditioned response. XII. The Brain Substrates of the Classically Conditioned NM Response in the Rabbit XIII. The Hippocampus Is Required for Trace Classical NM Conditioning XIV. The Amygdala Is Essential for the Acquisition and Retention of the Classically Conditioned Fear Response extinction training Extinction training can be presented following the acquisition of the conditioned response. Conditioned responses are formed by pairing a conditioned stimulus and an unconditioned stimulus. During extinction training only the conditioned stimulus is presented (i.e., the unconditioned stimulus is omitted). Extinction training will gradually result in the disappearance of the conditioned response. This is referred to as extinction ‘‘training’’ because the conditioned response is not forgotten but inhibited. In other words, extinction is an active process. Evidence that extinction training is different from forgetting comes from paired conditioned stimulus and unconditioned stimulus trials. This results in the rapid reinstatement of the conditioned response. The reinstatement occurs much more rapidly than the original acquisition of the conditioned XV. Brain Structures Involved in Human Classical Conditioning GLOSSARY alpha responses Nonassociative responses that follow the presentation of the conditioned stimulus. They are nonassociative in the same way that unconditioned responses are nonassociative in that they are innate reflexes to the presentation of a stimulus. However, Encyclopedia of the Human Brain Volume 1 813 Copyright 2002, Elsevier Science (USA). All rights reserved. 814 CLASSICAL CONDITIONING response. Therefore, the conditioned response is not forgotten but, rather, inhibited. interstimulus interval Sometimes called the conditioned stimulus–unconditioned stimulus interval, the interstimulus interval is the amount of time between the onset of the conditioned stimulus and the onset of the unconditioned stimulus. intertrial interval A classical conditioning trial begins with the onset of the conditioned stimulus and ends with the offset of the unconditioned stimulus. The intertrial interval is the amount of time between each individual trial and is usually an average amount of time. The exact time between trials is usually varied within a restricted range in order to prevent ‘‘time’’ from becoming a conditioned stimulus as it is in temporal conditioning. latency The amount of time between one event and another event. For example, the latency of the conditioned response is the amount of time between the onset of the conditioned stimulus and the onset of the conditioned response. The latency of the unconditioned response is the time between the onset of the unconditioned stimulus and the onset of the unconditioned response. pseudoconditioning Classical conditioning results in conditioned responses when an association forms between the conditioned stimulus and the unconditioned stimulus. However, sometimes responses that appear to be conditioned responses, in that they follow the presentation of a conditioned stimulus, result from experience with the unconditioned stimulus only and not because of an association between the conditioned stimulus and the unconditioned stimulus. This is known as pseudoconditioning. spontaneous recovery Spontaneous recovery is related to the process of extinction. Extinction training will result in the disappearance of the conditioned response. However, when the subject is given further extinction training following a break, the conditioned response will initially reemerge or spontaneously recover. Continued extinction training will more quickly result in the disappearance of the conditioned response until spontaneous recovery no longer occurs. unconditioned response An innate response that is elicited by the unconditioned stimulus. unconditioned stimulus A stimulus that is signaled by the conditioned stimulus. The unconditioned stimulus always elicits an unconditioned response. unconditioned stimulus-alone test trial In unconditioned stimulus-alone test trials, the conditioned stimulus is omitted, leaving only the unconditioned stimulus presentation. In many classical conditioning experiments, the amplitude of the unconditioned response is an important measure (e.g., as a test to evaluate if a manipulation such as a brain lesion affects the ability of the subject to make a completely normal response). However, in a subject that is emitting conditioned responses, it is not possible to obtain an accurate measure of the unconditioned response amplitude because of contamination from the presence of conditioned responses. Unconditioned stimulus-alone test trials allow an uncontaminated measure of unconditioned response amplitude in subjects that are emitting conditioned responses. Classical conditioning is a basic form of associative learning in which the organism learns something about the causal fabric of the environment or, in an experimental setting, the relationship of stimuli. Stimuli can be arranged so that one stimulus provides the organism with information concerning the occurrence of another stimulus. This type of associative learning is most commonly referred to as classical conditioning, but it has also been termed Pavlovian conditioning, respondent conditioning, and conditioned reflex type I conditioning, or type S conditioning. I. INTRODUCTION In the most basic form of classical conditioning, the stimulus that predicts the occurrence of another stimulus is termed the conditioned stimulus (CS). The predicted stimulus is termed the unconditioned stimulus (US). The CS is a relatively neutral stimulus that can be detected by the organism but does not initially induce a reliable behavioral response. The US is a stimulus that can reliably induce a measurable response from the first presentation. The response that is elicited by the presentation of the US is termed the unconditioned response (UR). The term ‘‘unconditioned’’ is used to indicate that the response is ‘‘not learned’’ but, rather, it is an innate or reflexive response to the US. With repeated presentations of the CS followed by US (referred to as paired training) the CS begins to elicit a conditioned response (CR). Here, the term ‘‘conditioned’’ is used to indicate that the response is ‘‘learned.’’ The most well-known example of classical conditioning comes from the pioneering work of Ivan Pavlov (1849–1936) and his dogs. In this prototypical example, a bell (in reality the CS was usually a metronome or buzzer) is rung just before meat powder is placed on the dog’s tongue. The meat powder causes the dog to salivate. Therefore, the meat powder acts as the US, and the salivation caused by the meat powder is the UR. Initially, the ringing bell, which serves as the CS, does not cause any salivation. With repeated pairings of the ringing bell CS and the meat powder US, the ringing bell CS causes the salivation to occur before the presentation of the meat powder US or even if the meat powder not presented. The salivation in response to the presentation of the ringing bell CS is the learned or conditioned response. II. BRIEF HISTORY In 1904, Ivan Pavlov was awarded the Nobel prize in medicine for his pioneering work on the physiology of CLASSICAL CONDITIONING digestion. Pavlov primarily studied the process of digestion in dogs. This early research set the stage for the discovery of a learning process that would come to be known as classical conditioning. As early as 1880, Pavlov observed that sham feedings, in which food was eaten but failed to reach the stomach (being lost through a surgically implanted esophageal fistula), produced gastric secretions, just like real food did. He immediately understood that this phenomenon must involve the central nervous system and began a program to thoroughly evaluate the parametric features of this learned response. Pavlov modified his preparation in order to simplify the forthcoming studies. Rather than measure gastric secretions, he began measuring salivation. The first of these studies involved showing the dog a piece of bread as the CS. Pavlov then noticed that in dogs with extensive training, even the act of an experimenter walking into a room could elicit salivation. This finding led to the discovery that a variety of stimuli could induce salivation if paired with meat powder. Accordingly, in a further simplification of his experimental procedures he began using the sound of a bell presented with the meat powder to elicit salivation. It is this preparation that has become synonymous with Pavlov and historical examples of classical conditioning. In fact, Pavlov has become synonymous with classical conditioning because the term Pavlovian conditioning can be used interchangeably with classical conditioning. Initially, Pavlov referred to the conditioned response as a ‘‘psychic secretion’’ to distinguish this type of response from the unlearned physiological secretions that would later come to be known as the unconditioned response. In 1903, a student of Pavlov’s published a paper that changed the term psychic secretion to conditioned reflex. The terms conditioned reflex and unconditioned reflex were used during the first two decades of the 20th century, during which time this type of learning was often referred to as ‘‘reflexology.’’ Although Pavlov is correctly credited with the discovery of classical conditioning, and with identifying and describing almost all the basic phenomena associated with this form of conditioning, it is worth noting that the phenomenon of classical conditioning was independently discovered by an American graduate student in 1902. Edwin B. Twitmyer made this discovery while finishing his dissertation work on the ‘‘knee-jerk’’ reflex. When the Patellar tendon is lightly tapped with a doctor’s hammer, it results in the wellknown knee-jerk reflex. Twitmyer’s work required 815 many tap-induced reflexes for each subject. Twitmyer, BrainSoft.ir like Pavlov, noticed that eventually the mere sight of the doctor’s hammer (the CS) could produce a kneejerk reflex (the CR). This largely forgotten report was the first example of classical conditioning of a muscle reflex. The potential significance of this finding was not apparent to Twitmyer, and the work was never extended or cast in a theoretical framework as Pavlov had done. Pavlov’s work on classical conditioning was essentially unknown in the United States until 1906 when his lecture ‘‘The Scientific Investigation of the Psychical Faculties or Processes in the Higher Animals’’ was published in the journal Science. In 1909, Robert Yerkes, who would later become president of the American Psychological Association, and Sergius Morgulis published a thorough review of the methods and results obtained by Pavlov. Although these reports provided a true flavor of the potential value of classical conditioning, the method was not immediately embraced by psychologists. This changed when John B. Watson, who is widely regarded as the founder of a branch of psychology known as behaviorism, championed the use of classical conditioning as a research tool for psychological investigations. Watson’s presidential address delivered in 1915 to the American Psychological Association was titled ‘‘The Place of the Conditioned Reflex in Psychology.’’ Watson was highly influential in the rapid incorporation of classical conditioning, as well as other forms of conditioning, into American psychology. In 1920, his work with classical conditioning culminated in the now infamous case of ‘‘little Albert.’’ Albert B. was an 11-month-old boy who had no natural fear of white rats. Watson and Rosalie Rayner used the white rat as a CS. The US was a loud noise that always upset the child. By pairing the white rat and the loud noise, Albert began to cry and show fear of the white ratFa CR. With successive training sessions over the course of several months, Watson and Rayner were able to demonstrate that this fear of white rats generalized to other furry objects. The plan had been to then systemically remove this fear using methods that Pavlov had shown would eliminate or extinguish the CRFin this case, fear of furry white objects. Unfortunately, little Albert, as he as historically come to be known, was removed from the study by his mother on the day these procedures were to begin. There is no known reliable account of how this experiment on classical conditioning of fear ultimately affected Albert B. Nevertheless, this example of classical conditioning may be the 816 CLASSICAL CONDITIONING most famous single case in the literature on classical conditioning. In 1921, a popular textbook on conditioning changed the terms of conditioned and unconditioned reflex to the current terms of conditioned and unconditioned response. This broadened the concept of conditioning to include other behaviors that were not merely automatic reflexes. In 1927, the Anrep (a former student of Pavlov’s) translation of Pavlov’s Conditioned Reflexes was published, thus making all of his work available in English for the first time. The availability of 25 years worth of Pavlov’s research, in vivid detail, led to increased interest in the experimental examination of classical conditioningFan interest that has continued to this day. III. THE DISTINCTION BETWEEN CLASSICAL AND INSTRUMENTAL/OPERANT CONDITIONING In classical conditioning, no contingency exists between the CR and the presentation or omission of the US. In other words, it does not matter if a CR is made or not; the presentation of the US still occurs and is experienced and processed by the subject. This has been termed stimulus-contingent reinforcement or the law of contiguity. In instrumental conditioning, in contrast, a contingent relationship is arranged between the subject’s response and the presentation or nonpresentation of a reinforcer. In other words, instrumental conditioning involves response-contingent reinforcement or the law of effect. The simplest forms of learning are known as nonassociative learning. Examples of nonassociative learning include habituation (a decrease in a response to repeated presentations of a stimulus) and sensitization (an increase in a response to repeated presentation of a stimulus). Associative learning is a more complex form of learning. Classical conditioning and instrumental conditioning are both examples of associative learning. In classical conditioning, an association is made between two stimuli, the CS and US. This association is manifested by the occurrence of a conditioned response. In instrumental conditioning, an association is made between a stimulus and the outcome of a response. In other words, the organism learns what responses are reinforced given a particular stimulus. Although classical conditioning and instrumental conditioning are both examples of associative learning, classical conditioning is generally viewed as the simpler form of learning. There are many reasons for considering classical condition to be a simpler form of learning. For example, classical conditioning has been demonstrated in organisms that are very low on the phylogenetic scale, such as planaria, slugs, and leeches. Second, the ability to be classically conditioned appears earlier ontogenetically than the ability to be instrumental conditioned. Successful classical conditioning has been reported for chick embryos, neonatal monkeys, goats, and dogs as well as human fetuses in utero. IV. ORIGIN OF THE TERM CLASSICAL CONDITIONING Often, it is possible to precisely identify the time and reason a new scientific term is introduced. Generally, this is possible because the author of the new term describes the reasons for introducing the term and justifies why the particular term was chosen. For example, the term instrumental conditioning was coined by Clark L. Hull to describe the type of learning in which the subject is ‘‘instrumental’’ in obtaining reinforcement. That is, the animal is the instrument, as in maze learning. Operant conditioning was chosen by B. F. Skinner because the subject must perform a behavioral ‘‘operation’’ to obtain reinforcement. In other words, the animal operates on a manipulandum in the environment, such as pressing a bar to obtain food. However, unlike instrumental and operant conditioning, there does not appear to be an instance in which a single individual coined the term classical conditioning. Therefore, only an inference can be made concerning the origin of the term classical conditioning. Only in the most rare cases do authors comment on the origin of the term classical conditioning. Often in these cases, the term is inaccurately attributed to John B. Watson. This is understandable because Watson was largely responsible for publicizing the classical conditioning method and outlining how it could be used as a method of scientific investigation. However, Watson never actually used the term classical conditioning, instead referring to it as simply the ‘‘conditioned reflex.’’ His 1930 edition of Behaviorism was his last significant scientific publication, and although the process of classical conditioning is referenced extensively, the term was never used. Before the 1940s, the process of classical conditioning was referred to most often as the conditioned reflex, Pavlovian conditioning, or simply conditioning. 817 CLASSICAL CONDITIONING However, in the 1930s, scientists began to understand that the laws governing learning, in paradigms in which reinforcement was contingent on the organism’s behavior, appeared to be fundamentally different from the laws governing the conditioned reflex. The latter type of learning would come be known as instrumental or operant conditioning. This created a need to distinguish these different forms of conditioning. It is my contention that the term classical conditioning developed as a contraction of the descriptive phase ‘‘classical Pavlovian conditioning’’ that was used to denote the ‘‘well-known’’ (i.e., classical) type of conditioning used by Pavlov (i.e., Pavlovian). V. TYPES OF CLASSICAL CONDITIONING Classical conditioning is a generic term that can refer to a variety of different types of classical conditioning procedures and paradigms. Two different parameters can distinguish the type of classical conditioning: (i) the temporal arrangement and spacing of the CS and the US and (ii) the type of response that is measured and conditioned. Some general comments regarding the influence that different types of conditioning have on behavior can be made, but it is impossible to make any specific comments regarding conditioning. The specifics depend on the response system being measured, the nature of the stimuli being used, and the species being conditioned. A. Temporal Arrangement and Spacing of the CS and the US The relationship of the CS and the US can be arranged in different ways to produce different conditioning paradigms. These arrangements can greatly influence how the CR develops. Figure 1 illustrates the major classical conditioning paradigms. The upward movement of a line represents the onset of a stimulus, and the downward movement of a line represents the offset of a stimulus. 1. Simultaneous Conditioning In simultaneous conditioning the CS and the US are presented simultaneously. In this case, because the US always elicits a UR and the CS and US are presented together, it is not possible to determine if CRs are present. In order to determine if a CR has developed, a BrainSoft.ir Figure 1 Temporal arrangements of the CS and the US (top trace) used in five classical conditioning paradigms. The upward movement of a line represents the onset of a stimulus, and the downward movement of a line represents the offset of a stimulus. A, simultaneous conditioning; B, delay conditioning; C, trace conditioning; D, backward conditioning; E, temporal conditioning. In the case of temporal conditioning, there is no discrete CS. The interval of time between the USs serves as the CS. CS-alone trial must be presented (i.e., the US is omitted). If the CS elicits a response on the CS-alone trial, this response is a CR. This form of conditioning is not commonly used, and it generally yields only weak conditioning, if any at all. 2. Delay Conditioning In delay conditioning, the CS onset precedes the US onset. The termination of the CS occurs with the US onset, during the US, at the termination of the US, or at some point after the US. This paradigm is called delay conditioning because the onset of the US is delayed relative to the onset of the CS. Generally, responses that develop to the CS and occur before the onset of the US are CRs. This is the most common conditioning paradigm and generally results in the most robust and rapid conditioning. 3. Trace Conditioning In trace conditioning, the CS is presented and terminated before the onset of the US. The interval separating the CS offset and the US onset is called the trace interval. This paradigm was named trace conditioning by Pavlov because in order for conditioning to occur, the subject (i.e., the subject’s brain) must maintain a memory ‘‘trace’’ of the CS. Responses that develop in response to the CS and occur before the onset of the US are CRs. This form of conditioning is 818 CLASSICAL CONDITIONING also common and yields good conditioning but generally not as readily as delay conditioning. 4. Backward Conditioning In the backward conditioning paradigm, the US is presented and terminated prior to the CS. A CR is a response that follows the presentation of the CS. This form of conditioning is not commonly used and in most circumstances does not result in conditioning. However, in some situations backward conditioning can occur. 5. Temporal Conditioning In temporal conditioning there are no discrete CSs. Instead, the US is presented at regular intervals, and over time the CR will be exhibited just prior to the onset of the US. In this case, the CS is the time interval. Temporal conditioning is possible in some experimental paradigms, but in most classical conditioning paradigms it does not result in conditioning. 6. Differential Conditioning unpaired training is used to describe a situation in which the CS and the US never coincide. Random unpaired training is used to describe a situation in which the CS and the US rarely coincide but occasionally (i.e., by chance) do so. Random unpaired training is generally thought to be superior to explicitly unpaired training because during explicitly unpaired training the animal learns that the CS signals a safety period (i.e., the US will not occur). Because the CS and the US are never paired, an association between them cannot be made. If responses nevertheless follow the presentation of the CS, it can be concluded that pseudoconditioning has occurred and not classical conditioning. If unpaired presentations do not result in responses to the CS, then any responses that subsequently develop in response to the CS, after paired training is begun, can be considered true CRs. B. Types of Classical Conditioning Based on the Measured Response In differential conditioning, two CSs are used. One of the CSs always precedes and predicts the US. This CS is termed the positive CS or the CS+. The other CS is not predictive of the US and occurs alone. This CS is termed the negative CS or the CS. Differential conditioning is indicated by a greater number of CRs in response to the CS+ than to the CS. In addition to the conditioning paradigms discussed previously, different types of classical conditioning can be characterized by the measured response. For example, in Pavlov’s experiments, he measured salivation, whereas in Twitmyer’s experiments, he measured the knee-jerk response. Although both of these basic experiments are examples of classical conditioning, they involve different response systems. 7. Controls for Pseudoconditioning 1. Two Fundamental Response Classes Classical conditioning results in CRs when an association forms between the CS and the US. However, sometimes responses that appear to be CRs, in that they follow the presentation of a CS, result from experience with US only and not because of an association between the CS and the US. This is known as pseudoconditioning. For example, if a very intense US is presented (e.g., a strong shock), the organism might respond to any subsequent stimulus presentation. The response does not occur because of an association between the CS and the US but, rather, because the US sensitizes the subject, making it more likely to respond to any stimulus presentation. To test for the possibility of pseudoconditioning, the CS and the US can be arranged so that the CS does not predict the US. This is known as unpaired training. There are two basic forms of unpaired training. Explicitly The nervous system can be divided into the central nervous system (CNS) and the peripheral nervous system (PNS). The CNS consists of all the neural tissue that is encased in bone (i.e., the brain and spinal cord). The CNS is discussed later. Because this section focuses on responses, the PNS will be discussed. In order for a response to occur, the PNS must be engaged (excluding responses of neurons that can be recorded from the CNS). The PNS can be divided into the autonomic nervous system and the somatic nervous system. The autonomic nervous system controls the viscera, and the somatic nervous system controls muscles. a. Autonomic Classical Conditioning Autonomic classical conditioning refers to any classical conditioning paradigm in which the measured response is under CLASSICAL CONDITIONING the control of the autonomic nervous system. The autonomic nervous system consists of two divisions, the sympathetic nervous system and the parasympathetic nervous systems. These two systems work together, in opposing directions, to control bodily functions such as heart rate, breathing, dilation and constriction of the pupil, and the control of sweat glands. These systems are not generally under voluntary control but can be modified by classical conditioning. Autonomic conditioning has been used extensively because in addition to involving primarily involuntary responses, autonomic responses can also be used as an index of changes in emotion. Autonomic conditioning is sometimes referred to as conditioned emotional responses because changes in emotion are accompanied by changes in these autonomic measures. For example, many of these measures are also used in polygraph/lie-detection work. The following is a brief overview of the most common types of autonomic conditioning studies. i. Galvanic Skin Response/Skin Conductance Response Conditioning This response is measured by the change in skin resistance to an electrical current. Research using this response measure dates back to the 1880s. For most of this time, the response was termed the Galvanic skin response (GSR) in honor of Luigi Galvani (1737–1798), an early pioneer in research describing the electrical nature of the body. Today, descriptive terms are commonly used, such as the electrodermal response or the skin conductance response (SCR). This response is measured by passing a small amount of electrical current between two electrodes pasted to the skin. The conductance (the reciprocal of resistance) between the two electrodes is measured. Many different stimuli can induce a SCR. For example, a mild shock, or any new stimulus of sufficient intensity as to attract the attention of the subject, can cause an autonomic response in which sweat glands pump extra sweat into sweat ducts located in the skin. The net effect of this action is an increase in skin conductance (i.e., a skin conductance response). In conditioning experiments in which, for example, a tone CS is paired with a shock or loud noise US, an association is made between the CS and the US and the CS begins to elicit a classically conditioned SCR. ii. Heart Rate Conditioning Another common autonomic conditioning paradigm is the classically conditioned heart rate response. In this paradigm the change in heart rate (i.e., heartbeats per minute) is 819 measured with electrodes pasted on the chest. For BrainSoft.ir example, a tone CS can be paired with a mild shock US. Initially, the CS does not cause a change in the heart rate, but the shock will increase the heart rate. With continued pairing of the CS and the US, the CS will elicit a CRFa change in heart rate. Interestingly, the CR in this case is change in heart rate to the CS, but the direction of change (i.e., slower or faster) depends on the animal species being testing. For example, the CR with human subjects is an increase in heart rate, which is called conditioned tachycardia (heart rate increasing). However, the CR with, for example, rabbits, is a decrease in heart rate, which is called conditioned bradycardia (heart rate slowing). iii. Other Examples Another example of autonomic classical conditioning is the conditioned pupillary response, which is a conditioned change in the size of the pupil. This response was first conditioned in 1922, but today it is rarely used because it is a difficult response to condition and to measure, and it is subject to a great deal of noise (i.e., spontaneous responses that are unrelated to the CS or the US). As noted previously, salivary conditioning was used by Pavlov in the initial demonstration of classical conditioning and this response was adapted for work with humans by Karl Lashley (1890–1958). Today, this paradigm is almost never used because salivation is a relatively slow response, difficult to measure, and difficult to condition in humans. b. Somatic Classical Conditioning Somatic classical conditioning refers to any classical conditioning paradigm in which the measured response is under the control of the somatic nervous system. The somatic nervous system as it relates to classical conditioning controls striate or skeletal muscles. Accordingly, any response that requires a motor movement must be controlled by striate muscles and the somatic nervous system. The following is a brief overview of the most common types of somatic conditioning paradigms. i. Eyeblink Classical Conditioning Eyeblink classical conditioning is by far the most common form of experimental conditioning paradigm with both humans and experimental animals. As early as 1922, eyeblink classical conditioning paradigms were being used. In eyeblink conditioning a CS (typically a tone or light) is paired with a US (e.g., a mild shock or puff of air to the eye). Eyeblink studies have been carried out in a variety of species, including humans, monkeys, rabbits, cats, rats, and mice. The response can be 820 CLASSICAL CONDITIONING measured with a minitorque potentiometer. A string is attached to the eyelid and any eyelid movement changes the resistance in the potentiometer, which can then be recorded. Electromyographic (EMG) measures can also be used to record the activity of the muscles controlling the eyeblink. Today, the most common method of measuring eyeblink responses in humans is the use of an infrared reflective sensor. In this method, an infrared beam is directed at the eye (usually mounted in goggles) and the amount of light reflected by the eyelid is recorded. Eyeblinks change the amount of light that is reflected and subsequently detected by an infrared sensor. ii. Nictitating Membrane Classical Conditioning Closely related to eyeblink conditioning, conditioning of the nictitating membrane (NM) is the most frequently used response measure in rabbits, which are the most frequently used experimental subjects for conditioning studies. The NM is often called the ‘‘third eyelid’’ and it consists of a sheet of cartilage located behind the inner canthus of the eye. When the eyeball is stimulated, the eyeball retracts into the eye socket. This causes the NM to passively sweep across a portion of the eyeball. The NM response is popular because it is simple to measure (usually with a minitorque potentiometer) and because the NM cannot completely cover the eye. Thus, it prevents the subject from avoiding the airpuff US (the eyelids are prevented from closing by the experimenter during NM conditioning). iii. Leg Flexion Classical Conditioning In the typical leg flexion experiment, an animal is usually restrained so that the legs hang freely, although it has also been used in freely moving subjects. A CS is paired with a mild shock US to the leg, which causes the leg to flex. With pairing of the CS and the US, a CR develops where the leg flexes in response to the CS. The response can be measured by a minitorque potentiometer or EMG. iv. Fear Classical Conditioning Classical conditioning has been increasingly used to study the learning of fear. This paradigm can be considered a hybrid of autonomic and somatic classical conditioning because fear causes numerous autonomic changes, which could be measured as the CR. However, in the rat, the most common subject for studies of this type, fear can also be measured with the somatic response of freezing. In the typical paradigm, a tone CS is paired with a shock US. The shock US is delivered to the rat through an electrified floor grid. With pairing of the CS and the US, a fear CR develops in response to the CS. In this case, the fear CR is freezing (the rat holds completely still). VI. ACQUISITION OF THE CR As noted previously, the specifics regarding any aspect of classical conditioning must be reserved for the particular type of classical conditioning paradigm, the response that is measured, and the species that is used. However, some general features can be noted regarding the acquisition of the CR. The most fundamental element of classical conditioning is the association of the CS and the US that results in the acquisition of the CR. During the initial presentation of the CS and the US, no response to the CS is observed. With continued pairing of the CS and the US (i.e., presentation of additional training trials), CRs begin to develop in response to the CS. This development of the CR is referred to as the acquisition of the CR. It is often referred to as an increasing probability of CRs. In other words, with continued training, the probability of a CR on a given trial increases. In some conditioning paradigms (such as classical conditioning of the NM response in rabbits) this probability can approach 0.99 (i.e., CRs on almost every trial). During the early phases of CR acquisition, CRs are generally smallamplitude responses that gradually (over the course of training) grow larger until they become as large as the UR. The latency of the CR also tends to change during CR acquisition. Initially, the latency of the CR onset begins just before the onset of the US. With continued training the latency decreases so that the onset of the CR begins well before the onset of the US. Finally, the CR is acquired to the extent that the maximum amplitude of the CR (the CR peak) occurs at the time of the US onset. This is sometimes referred to as a ‘‘well-timed’’ CR. A. Factors That Can Influence the Acquisition of the CR 1. Interstimulus Interval Every type of conditioning paradigm has an optimal interstimulus interval (ISI). The optimal ISI depends on the particular paradigm, but it is clear that ISIs that are very short or very long and result in little or no CR acquisition. For example, in the rabbit classically conditioned NM response paradigm, the optimal ISI CLASSICAL CONDITIONING for delay conditioning is 250 msec. An ISI of 100 msec or less results in poor conditioning, as does increasing the ISI to longer than 250 msec. In this paradigm, an ISI of 2000 msec (i.e., 2 sec) will not result in any CR acquisition. 2. Intertrial Interval It is a general finding from many conditioning paradigms that increasing the ITI will tend to produce more rapid CR acquisition, although this influence is not robust in most cases. Reducing the ITI to less than several seconds can drastically impair acquisition of the CR. The typical ITI ranges from 30 to 60 sec. 3. Temporal Arrangement and Spacing of the CS and the US As noted previously, delay conditioning generally results in the most rapid rate of CR acquisition followed by trace conditioning. Simultaneous conditioning generally results in much poorer CR acquisition. In many cases simultaneous conditioning (ISI¼0) does not produce CRs. Backward conditioning results in CRs in only a few paradigms and generally does not produce any CR acquisition. 4. CS Intensity It was initially believed that the CS intensity did not influence that rate of CR acquisition, but recent studies indicate that more intense CSs tend to result in slightly more rapid CR acquisition. 5. US Intensity An increase in the acquisition rate of the CR has been a consistent finding with increases in the intensity of the US. VII. EYEBLINK CLASSICAL CONDITIONING IN HUMANS It was previously noted that different types of classical conditioning paradigms can be distinguished on the basis of the measures response. By the early 1930s, successful classical conditioning had been reported in 23 different response systems (e.g., eyeblink response, skin conductance response, pupillary response, leg flexion response, and salivary response). Of these 821 numerous preparations, the majority were explored BrainSoft.ir for only short periods before being, for the most part, abandoned for one reason or another. For example, the salivary preparation in humans or other experimental animals never flourished because of many methodological difficulties. Classical conditioning of the skin conductance response has had a long history dating back to the late 19th century. However, skin conductance conditioning failed to flourish in the 1930s–1950s because the physiological basis of the response was poorly understood and difficult to measure accurately with the equipment of the day. A renewed interest occurred in the 1960s due to a better understanding of the physiology of the response and better and more readily available equipment for measuring and quantifying the response. Today, this paradigm still enjoys some popularity with researchers studying cognitive factors of classical conditioning and the neural mechanism underlying conditioned emotional responses. By the 1940s the human eyeblink classical conditioning paradigm had surpassed all other conditioning paradigms in terms of number of articles published primarily because it was methodologically superior to all other classical conditioning paradigms. From the 1940s through the 1960s, studies of human eyeblink conditioning remained the dominant paradigm for studying the processes and variables that related to classical conditioning. During this time, methodological improvements that included precise measurement of the eyeblink response allowed investigators to explore the effects of such variables as the intertrial interval, interstimulus interval, CS intensity, US intensity, and variable reinforcement schedules. Additionally, cognitive factors were also extensively explored. For example, subjects were asked about the information they acquired during the conditioning session. This information was then compared to how well the subjects acquired the CR. Other studies explored how instruction sets affected CR acquisition rates: that is, how different verbal instructions given to subjects before conditioning affected CR acquisition rates. VIII. PROBLEMS FOR STUDIES OF HUMAN EYEBLINK CLASSICAL CONDITIONING In all forms of classical conditioning, the responses are considered primarily reflexive in nature. For example, the UR is an innate, automatic, ‘‘reflexive’’ reaction to 822 CLASSICAL CONDITIONING the puff of air that is delivered to the eye. With repeated pairing of a CS and a US a learned, conditioned response develops. This response is also thought to be reflexive and, as such, should not require cognitive involvement. Nevertheless, although an eyeblink response can be involuntary, clearly the eyeblink response can also be brought under voluntary control. If individuals become aware of the fact that the CS predicts the US, they are in a position to voluntarily blink their eyes to avoid the airpuff (i.e., blink on purpose). This circumstance has caused concern among experimentalists because it has been argued, at least since the 1940s, that the processes and characteristics of responses that are voluntary are very different from those that are involuntary or reflexive. That is, classical conditioning measures involuntary, automatic responses, not purposeful behavior. Therefore, some means of identifying voluntary responses needed to be developed so that they could either be discarded as contaminants or analyzed separately. Two criteria emerged for identifying voluntary responses. The first was based on the slope of the response. Voluntary responses were believed to be more rapid and thus to have a characteristically steeper slope than a true CR. The second was based on response latency. Voluntary responses were believed to have a short onset latency (which also involved a steep slope), and eye closure was maintained until the onset of the US. Despite these improvements, there is no consensus about the most appropriate methods for detecting voluntary responses. In fact, the only largescale study to test the validity of these two criteria (slope and latency) found that neither was fully satisfactory for discriminating voluntary responses from conditioned responses. By the 1960s, studies of human eyeblink classical conditioning began to wane. This gradual decrease in human eyeblink classical conditioning research was likely due to a combination of at least two factors. First, researchers were never completely successful in identifying and dealing with the issue of voluntary vs conditioned responding. Second, researchers never reached a consensus on what the ‘‘standard’’ conditions should be for conditioning studies. Various laboratories used different CS and US intensities, different ISIs and ITIs, and different criterion for determining if responses were CRs. Consequently, there were persistent problems with obtaining reproducible results between different laboratories, which essentially prevented progress in exploring interesting variables. IX. THE EMERGENCE OF ANIMAL STUDIES OF CLASSICAL CONDITIONING Although human eyeblink classical conditioning studies began a steep decline in the 1960s, classical conditioning studies using animals showed rapid growth. This was in part likely due to at least two factors. First, some experimentalists moved from human conditioning research to working with animals, which served to stimulate the field of animal work. Second, although problematic, the human research on classical conditioning nevertheless revealed the potential of classical conditioning to serve as a paradigm for the systematic and thorough analysis of associative learning if it could be appropriately exploited. What was needed was the development of a ‘‘model’’ paradigm of classical conditioning in which all the basic features (e.g., CS and US types and intensities, measurement of the responses, and ISIs and ITIs) would be held consistent across laboratories. A model paradigm must result in robust acquisition of the CR, which is reliable across laboratories. This would allow theoretical questions about learning and memory to be addressed. Additionally, these methods must be economical and relatively easy to implement, and the characteristics of the learned response must not be unique to the experimental circumstances or to the species being tested. In the early 1960s, Isidore Gormezano and colleagues developed a paradigm in which classical conditioning of the NM response was used with rabbits. The NM is vestigial in humans but is quite pronounced in the rabbit. It consists of a sheet of cartilage located behind the inner canthus of the eye. When the eyeball is stimulated, it retracts into the eye socket. This causes the NM to passively sweep across a portion of the eyeball. This movement is the measured response. The NM response was chosen because it is simple to measure (usually with a minitorque potentiometer) and because the NM cannot completely cover the eye. In this preparation, the rabbit’s eyelids are held open with clips to prevent the subject from completely closing its eyelids and avoiding an airpuff US. Because the rabbit cannot avoid the airpuff US, the response cannot be an instrumental response but remains squarely within the domain of classical conditioning. This paradigm remains the model classical conditioning paradigm. It has endured because it has simply proven to be an ideal paradigm. The result is that the processes and factors that influence the acquisition of the CR are now understood in detail. These details CLASSICAL CONDITIONING allow learning theories to be constructed. This model system then allows the hypotheses that are derived from these theories to be rigorously tested and interpreted against an immense background of empirical data. In this respect, the importance of classical conditioning to modern learning theory cannot be overestimated. X. EYEBLINK CLASSICAL CONDITION AS A TOOL TO STUDY BRAIN FUNCTION In his preface to the Russian edition of his 1927 book Conditioned Reflexes, Ivan Pavlov wrote ‘‘At [this] time [I am] am convinced that this method of research [classical conditioning] is destined, in the hands of other workers, and with new modifications in the mode of experimentation, to play a yet more considerable part in the study of the physiology of the nervous system.’’ Pavlov’s statement can be viewed not only as prophetic but also as in fact quite understated. Until the past 25 years, Pavlov’s enthusiasm for classical conditioning as a method of studying brain function had not been shared by physiologists. As a research procedure, classical conditioning had been almost the exclusive property of psychologists. However, beginning in the mid-1970s, classical conditioning became one of the most important tools for relating learning to specific brain structures. XI. THE SEARCH FOR THE ENGRAM The term engram is a hypothetical construct used to represent the physical processes and changes that constitute memory in the brain, and the search for the engram is the attempt to locate and identify that memory. Karl Lashley was perhaps the first person to clearly conceptualize the issue in a framework that would lend itself to experimental analysis. Lashley’s primary interest was in localizing the engram to a specific region of the brainFan endeavor that would prove to be extraordinarily difficult and ultimately end in failure for Lashley. A likely reason why Lashley and others have historically had such great difficulty in locating an engram is because the behavior they typically used was an operant form of conditioningFa relatively complex form of associative learning. To make localization more tractable a simpler behavior was needed. Classical conditioning of the NM response in rabbits proved to be the solution. 823 For many reasons, Classical conditioning is a valuable tool forBrainSoft.ir studying the physical processes of the brain that may be important for forming and storing memory. The presentation of the CS and the US is determined by the experimenter, not by the subject’s behavior. This has important implications for the analysis of stimulus selection, which can be precisely manipulated by the experimenter. However, of greater importance is the fact that the CR is time locked to the CS. This allows a temporal analysis of neural events and then correlation of electrophysiological recordings with changes in the CR. The conditioning procedures provide a more adequate control for nonspecific effects of training on biological processes than do operant procedures. For example, the same kind of density of stimulation and number of unconditioned responses can be produced in both experimental and control conditions. Perhaps the greatest advantage is that the effects of experimental manipulations on ‘‘learning’’ rather than ‘‘performance’’ can be easily evaluated. This problem of learning versus performance has plagued the study of brain substrates of learning from the beginning. For example, does a brain lesion impair a learned behavior because it damages the memory trace or because it impairs the animal’s ability to respond? This problem can be circumvented by simply comparing the amplitude of the learned or conditioned response to the reflex or unconditioned response amplitude. If the lesion abolishes the CR (the learned response) but leaves the UR (the ability to respond) intact, then it can be concluded that the lesion has impaired learning rather than performance. The use of the classically conditioned NM response in rabbits offers additional advantages for studying the neural correlates of the learned behavior. Rabbits are docile animals that do not find full-body restraint aversive. They will usually sit quietly for more than 2 hr with little or no struggling. Rabbits also have a very low spontaneous blink rate, which reduces the amount of contamination caused by blinks that are unrelated to the CS and the US. The parametric features of the CR have been well characterized. The behavioral NM response is robust and discrete and the exact amplitude–time course of the response is easily measurable. The CR is acquired, at least to a significant degree, in a single training session, but not all at once. This allows for an analysis of brain substrates over a period of time as learning is proceeding (as opposed to one-trial learning or some operant tasks, which show a rapidly accelerating acquisition curve). The presentation of the CS and the US does not yield sensitization or 824 CLASSICAL CONDITIONING pseudoconditioning. That is, unpaired presentations of the CS and the US do not increase the ability of the CS to elicit a response; they also do not increase the spontaneous blink rate. These are the primary reasons that led Richard F. Thompson and associates to choose this paradigm to continue the search for the elusive engram. Conceptually, to search for an engram, a neural circuit must be identified. No single neurobiological method or technique is sufficient in and of itself to define the characteristics of a neural circuit, including the essential site of plasticity. Therefore, numerous methods must be used to determine the roles that different brain structures play in any learned behavior. To date, the most fruitful methods have been electrophysiological recordings, permanent and reversible lesions, and electrical microstimulation of various nuclei and fiber tracts. These methods are used in combination with classical conditioning and with the assumption that there are neural pathways connecting the CS, the US, and the UR pathways. It is also assumed that it is possible to localize a specific brain region where some essential modification takes place to drive the conditioned response. All these methods have been used successfully in defining the circuit for classical eyeblink conditioning. XII. THE BRAIN SUBSTRATES OF THE CLASSICALLY CONDITIONED NM RESPONSE IN THE RABBIT Since the early 1970s, Thompson and many others have used a variety of methods to search for the classically conditioned NM response engram. A neuronal circuit diagram has been systematically constructed during the course of 20 years to represent all the brain structures that are essential for the acquisition and retention of the classically conditioned NM response for the delay conditioning paradigm. This circuit is the most thoroughly investigated and completely understood learning and memory circuit known for the mammalian brain. The essential site of plasticity (i.e., the location of the engram) appears to be the interpositus nucleus of the cerebellum, although the cerebellar cortex is also important. A. Components of the Circuit Figure 2 is a neuronal circuit diagram that represents all the brain components that are essential for the acquisition and retention of the CR. Figure 2 The cerebellar neural circuit for delay eyeblink/nictitating membrane (NM) classical conditioning. The focus of this diagram is on the relationships between the conditioned stimulus (tone CS), unconditioned stimulus (airpuff US), conditioned response (CR), unconditioned response (UR), and their connecting fiber pathways. The cerebellar cortex in the diagram consists of granule cells, parallel fibers, and Purkinje cells. Cerebellar structures in the diagram include the cerebellar cortex and interpositus nucleus. All other structures depicted are located in the brain stem. Pairing of the CS and US results in an essential plastic change in the interpositus nucleus that results in the generation of the CR. CLASSICAL CONDITIONING 1. The CS Pathway The tone CS is detected by auditory hair cells, which stimulate spiral ganglion cells. These cells project to the cochlear nucleus, which then projects to the lateral pontine nuclei. The lateral pontine nuclei send mossy fibers to both interpositus nucleus and granule cells in the cerebellar cortex. 2. The US Pathway The airpuff US is detected by similunar ganglion cells that send somatosensory information to the spinal trigeminal nucleus. This nucleus projects to the dorsal accessory olive. The olive sends climbing fibers to both the interpositus and the cerebellar cortex. 3. The UR Pathway The sensory trigeminal projects to the abducens, accessory abducens, and facial motor nucleus, which are the motor nuclei that drive the eyeblink and NM response. 4. The CR Pathway The interpositus nucleus, which receives convergent information concerning the CS and the US, projects to the red nucleus. The red nucleus then projects to the same motor nuclei that drive the eyeblink and NM response (i.e., the abducens, accessory abducens, and facial motor nucleus). 5. Circuit Philosophy and Function Figure 2 focuses on the relationships of the CS, US, CR, and UR pathways rather than emphasizing the physical position of the nuclei and tracts within the brain. Initially, in a naive subject, the CS does not cause a NM response. The US stimulates the trigeminal nucleus, which then projects to motor nuclei to cause the reflexive UR from the first trial on. During training the interpositus nucleus receives convergent CS and US information. Eventually, this convergent information causes a change (i.e., learning) in the interpositus nucleus and cerebellar cortex. This change enables the CS information to activate the interpositus, which can then activate the red nucleus. The red nucleus then activates the motor nuclei, which then drive the NM response. This response occurs before the delivery of the airpuff US. This is the conditioned response. 825 B. Brain Structures That Are Critical for Classical BrainSoft.ir Conditioning of the NM Response 1. Interpositus Nucleus The interpositus nucleus receives convergent CS and US information from the pontine nuclei (CS information) and the inferior olive (US information). The plastic changes that occur in the interpositus as a result of this convergent information are responsible for the CR. The interpositus is considered to be the prime location of the classically conditioned NM response engram. Damage to the interpositus nucleus completely and permanently abolishes the CR without affecting the UR. 2. Cerebellar Cortex The cerebellar cortex also receives convergent CS and US information. The output of the cerebellar cortex goes through the interpositus nucleus. It has been suggested that the cerebellar cortex might form and store the essential plasticity for the engram and that interpositus lesions only block the expression of the CR. However, large lesions of the cerebellar cortex do not abolish the CR or prevent the acquisition of the CR. The cerebellar cortex likely plays a critical modulatory role in the acquisition and expression of the CR. 3. Pontine Nuclei The pontine nuclei receive information about the tone CS and relay this information to the interpositus nucleus and cerebellar cortex. Rabbits can be classically conditioning by electrically stimulating the pontine nuclei in place of a tone CS and pairing the stimulation with an airpuff US. 4. Inferior Olive The inferior olive receives US information from the trigeminal nucleus located in the brain stem. The inferior olive then relays this information to the interpositus and cerebellar cortex. Rabbits can be classically conditioned by electrically stimulating the inferior olive in place of an airpuff US. Pairing a tone CS with inferior olive stimulation as the US results in the development of CRs. In fact, rabbits can be conditioned without tones or airpuffs by stimulating the pontine nuclei as the CS and the inferior olive as the US. In well-trained rabbits (i.e., rabbits showing CRs), 826 CLASSICAL CONDITIONING lesions of the inferior olive result in a gradual extinction of the CR as if the US has been removed, even though the US is still being presented. This indicates that the inferior olive sends critical US information to the interpositus and cerebellar cortex. 5. Red Nucleus The red nucleus is a critical structure in the CR pathway. It receives input from the interpositus, which is outputting a neural signal for driving the CR. The red nucleus then relays this signal to motor nuclei for driving the learned NM response. Lesions of the red nucleus abolish the conditioned response. It is known, however, that the red nucleus is not responsible for forming the CR because reversibly inactivating the red nucleus during training prevents the expression of the CR but does not prevent the acquisition of the CR. When the lesion is reversed (i.e., restored to normal function), CRs rapidly appear. In contrast, when the interpositus is inactivated, there is no acquisition of the CR. 6. Trigeminal Nucleus The trigeminal nucleus receives information regarding the occurrence of the US. It then sends this information to the interpositus and cerebellar cortex (via the inferior olive) so that an association between the CS and the US can be accomplished. It also sends the US signal to the motor nuclei for driving the UR. 7. Accessory Abducens The accessory abducens is the motor nucleus that drives the NM response. It receives information from the trigeminal nucleus for driving the UR and information from the red nucleus for driving the CR. XIII. THE HIPPOCAMPUS IS REQUIRED FOR TRACE CLASSICAL NM CONDITIONING The circuit just described includes all the brain structures required for the successful acquisition and retention of the CR in the delay classical conditioning paradigm. For example, in a well-trained animal, all the neural tissue above the level of the midbrain can be completely removed without affecting the retention and expression of the CR. In other words, removing the entire neocortex, thalamus, basal ganglia, and limbic system, including the hippocampus, has no effect on the CR. For trace classical conditioning, these structures are also essential; however, the hippocampus is also required. If an interval of at least 500 msec separates the offset of the CS and the onset of the US, hippocampal lesions prevent CR acquisition and retention. XIV. THE AMYGDALA IS ESSENTIAL FOR THE ACQUISITION AND RETENTION OF THE CLASSICALLY CONDITIONED FEAR RESPONSE Although amygdala lesions do not affect the acquisition or retention of eyeblink/NM classical conditioning, the amygdala is essential for the acquisition and retention of the classical conditioned fear response. For example, in rats, when a tone CS is paired with a shock US several times, a conditioned fear response develops. In other words, after pairing the CS will cause a fear response of freezing. The rat that has been conditioned will hold completely still when the CS is presented. Amygdala lesions completely prevent rats from learning or retaining this fear response. Thus, classical condition has revealed that the amygdala is a critical brain structure for emotional fear learning. XV. BRAIN STRUCTURES INVOLVED IN HUMAN CLASSICAL CONDITIONING Although studies of classical conditioning in humans began to wane in the 1960s, particularly for eyeblink classical conditioning, within the past 10 years there has been a resurgence of experimental work that can largely be attributed to the success of classical conditioning as a tool to study brain function in the experimental animal. Currently, our understanding of how different human brain structures contribute to classical condition lags far behind what is known in the animal and will likely never approach the level of precision that is possible with animal studies. Nevertheless, there has recently been much progress with relating classical conditioning to brain function in humans. Recent findings from human studies have been remarkably consistent with previous work with animal studies of conditioning. For example, work in rats has convincingly demonstrated the importance of the amygdala in fear classical conditioning. Studies using 827 CLASSICAL CONDITIONING brain imagining methods such as positron emission tomography (PET) and functional magnetic resonance image (fMRI) have shown that fear classical conditioning activates the amygdala. Additionally, humans with bilateral degeneration of the amygdala, as a result of Urbach–Wiethe disease, show an impaired ability to acquire fear classical conditioning. Results from human studies of eyeblink classical conditioning have also been remarkably consistent with those of studies in rabbits. For example, work with rabbits clearly indicates that the cerebellum is critical for acquisition of delay classical conditioning. Humans with cerebellar damage are also severely impaired on eyeblink classical conditioning. Humans with bilateral hippocampal damage due to anoxia are normal at acquiring CRs in the delay classical conditioning paradigm, but they are impaired when tested on trace classical conditioning. These results are entirely consistent with the animal work. Imagining studies using PET and fMRI have also consistently identified the cerebellum and hippocampus as being activated during classical conditioning of the eyeblink response. In addition to supporting previous work in animals, human eyeblink conditioning studies have also extended our understanding of brain function in several interesting ways. For example, it has been reported that the knowledge that humans sometime acquire about the stimulus contingencies of the conditioning experiment (e.g., the CS predicts the US) is an important variable for trace conditioning but irrelevant or superfluous for delay eyeblink conditioning. Humans with hippocampal damage fail to acquire this knowledge and accordingly fail to acquire trace conditioning while being unaffected on delay conditioning. It has repeatedly been reported that subjects with Alzheimer’s disease and those with probable Alzheimer’s disease are impaired on delay classical conditioning of the eyeblink response. At first, this finding did not appear to be congruent with the animal work because Alzheimer’s disease does not affect the cerebellum or other brain stem structures that are critical for delay conditioning. However, work in rabbits has shown that although the hippocampus can BrainSoft.ir be removed without affecting acquisition in the delay paradigm, disrupting the hippocampus by inactivating the cholinergic input to the hippocampus can disrupt delay conditioning. Alzheimer’s disease disrupts the septohippocampal cholinergic system and it is this disruption that likely causes the impairment in delay eyeblink conditioning. If fact, this classical conditioning paradigm is so sensitive to the early effects of Alzheimer’s disease that, it has been proposed as a simple neuropsychological test for Alzheimer’s. Finally, humans with autism, a developmental disorder characterized by severe impairments in communication and social relating and by ritualistic and repetitive patterns of behavior, also show abnormalities in the cerebellum. Subjects with autism also show abnormal acquisition and extinction of the CR. This finding further supports the involvement of the cerebellum in classical eyeblink conditioning. See Also the Following Articles BEHAVIORAL NEUROGENETICS d COGNITIVE PSYCHOLOGY, OVERVIEW d INTELLIGENCE d REINFORCEMENT, REWARD, AND PUNISHMENT Suggested Reading Gormezano, I., Prokasy, W. F., and Thompson, R. F. (Eds.) (1987). Classical Conditioning. Erlbaum, Hillsdale, NJ. Green, J. T., and Woodruff-Pak, D. S. (2000). Eyeblink classical conditioning: Hippocampal formation is for neutral stimulus associations as cerebellum is for association-response. Psychol. Bull. 126, 138–158. Kim, J. J., and Thompson, R. F. (1997). Cerebellar circuits and synaptic mechanisms involved in classical eyeblink conditioning. Trends Neurosci. 20, 177–181. Pavlov, I. P. (1927). Conditioned Reflexes (G. V. Anrep, Trans.). Oxford Univ. Press, London. Woodruff-Pak, D. S., and Steinmetz, J. E. (Eds.) (2000a). Eyeblink Classical Conditioning: Volume IFApplications in Humans. Kluwer, Boston. Woodruff-Pak, D. S., and Steinmetz, J. E. (Eds.) (2000b). Eyeblink Classical Conditioning: Volume IIFAnimal Models. Kluwer, Boston.