Jump to Content Jump to Main Navigation

Psychology Neuroscience of Associative Learning
by
Gabrielle Weidemann, Gavan McNally

Introduction

Associative learning involves the encoding of relationships between events, for example, between two stimuli or between a stimulus and a response. Associative learning is distinguished from nonassociative learning, which involves only a single stimulus. In the narrowest definition of associative learning, it is restricted to the learning that occurs during classical conditioning and instrumental conditioning. However, associative learning can also be used more broadly to encompass all memory for the relationship between events and as such includes other forms of short-term and long-term memory. During the 20th century and since 1980 in particular, vast gains were made in understanding the neural mechanisms of associative learning, particularly the neuroscience of classical and instrumental conditioning and of associative memories. Much of this newfound knowledge has come from the study of a relatively small number of model systems: preparations in animals in which the essential neurocircuitry, as well as the cellular and molecular mechanisms underlying associative learning, can be isolated and analyzed. While there are some reference works that summarize the commonalities in associative learning processes across these different model systems, much of literature is devoted to describing the neurobiology of a particular model system. The first section of this bibliography introduces some articles and reference works that review the neuroscience of associative learning and memory across a variety of different model preparations and synthesize the commonalities across these different preparations. It also recommends journals that publish high-quality empirical and review articles on the neuroscience of associative learning across the various different model preparations. The bibliography’s remaining sections examine the neuroscience of associative learning in particular model systems.

General Overviews

Although most of the literature on the neuroscience of associative learning describes the neurobiology of a particular model system, there are a number of articles that attempt to summarize the neuroscience in each of the different model systems and other articles that describe the commonalities of mechanism between these different model systems. Steinmetz, et al. 2003 provides a summary of current opinion of the biological substrates of learning in the most commonly studied Pavlovian and instrumental conditioning preparations. Kandel 2001 provides a review of the cellular mechanisms that allow for the neuronal plasticity that underlies different forms of associative learning, with a particular emphasis on the commonalities in molecular mechanisms among different species. Fanselow and Poulos 2005 and Medina, et al. 2002 both review the commonalties of mechanism in Pavlovian fear conditioning and eyeblink conditioning. Thompson 1986 provides a justification for the model systems approach to studying the neuroscience of associative learning.

  • Fanselow, Michael S., and Andrew M. Poulos. 2005. The neuroscience of mammalian associative learning. Annual Review of Psychology 56:207–234.

    DOI: 10.1146/annurev.psych.56.091103.070213Save Citation »Export Citation »E-mail Citation »

    Reviews the neuroscience of associative learning with a particular focus on the Pavlovian conditioning model systems of fear conditioning and eyeblink conditioning, which are the most extensively studied mammalian associative learning preparations. Available online for purchase or by subscription.

    Find this resource:

  • Kandel, Eric R. 2001. The molecular biology of memory storage: A dialogue between genes and synapses. Science 294:1030–1038.

    DOI: 10.1126/science.1067020Save Citation »Export Citation »E-mail Citation »

    Reviews the seminal work of the author, for which he shared the 2000 Nobel Prize in Physiology or Medicine, on the mechanisms of learning and memory. The focus is on the molecular events that occur inside neurons that underpin memory storage and plasticity in different model preparations, most notably Aplysia. Emphasis is also placed on the evidence that these molecular events are shared between different species. Available online for purchase or by subscription.

    Find this resource:

  • Medina, Javier F., J. Christopher Repa, Michael D. Mauk, and Joseph E. LeDoux. 2002. Parallels between cerebellum- and amygdala-dependent conditioning. Nature Reviews Neuroscience 3.2: 122–131.

    DOI: 10.1038/nrn728Save Citation »Export Citation »E-mail Citation »

    Compares and contrasts the neural circuitry and the cellular mechanisms of associative learning in eyeblink conditioning and fear conditioning, the two most extensively studied forms of associative learning within the mammalian brain. Available online for purchase or by subscription.

    Find this resource:

  • Steinmetz, Joseph E., Jeansok Kim, and Richard F. Thompson. 2003. Biological models of associative learning. In Handbook of psychology. Vol. 3, Biological psychology. Edited by Michela Gallagher, Randy J. Nelson, and Irving B. Weiner, 499–541. New York: Wiley.

    Save Citation »Export Citation »E-mail Citation »

    This chapter summarizes the contributions of various invertebrate and mammalian models of associative learning to our understanding of the neural mechanisms of associative learning. It has a particular emphasis on Pavlovian conditioning but also includes a section on the neural substrates of discrete instrumental responses.

    Find this resource:

  • Thompson, R. F. 1986. The neurobiology of learning and memory. Science 233:941–947.

    DOI: 10.1126/science.3738519Save Citation »Export Citation »E-mail Citation »

    Describes why it is that the model systems approach to studying learning and memory, which involves the selection of an organism that exhibits a particular well-defined form of learning and memory and has a nervous system that is amenable to investigation, has been the most productive research strategy for understanding how the brain codes, stores, and retrieves memories. Available online for purchase or by subscription.

    Find this resource:

Reference Works

Several reference works broadly address the neuroscience of learning and memory. Rudy 2008, Eichenbaum 2002, Macphail 1993, and Kesner and Martinez 2007 are largely targeted at advanced undergraduate and graduate students and represent an attempt to summarize our current knowledge about the molecular and cellular mechanisms of memory in a variety of different model preparations. These references are largely intended as textbooks, but each has a slightly different emphasis. Gormezano and Wasserman 1992 and Steinmetz, et al. 2001 are collections of articles from researchers that summarize the work from different labs that are investigating the biological basis of learning and memory in a variety of different model preparations. As such, they each provide a snapshot of the current thinking and particular challenges in the neuroscience of associative learning at the time they were published.

  • Eichenbaum, Howard. 2002. The cognitive neuroscience of memory. Oxford: Oxford Univ. Press.

    DOI: 10.1093/acprof:oso/9780195141740.001.0001Save Citation »Export Citation »E-mail Citation »

    Targeted at advanced undergraduates and graduate students, this text provides an excellent overview of the neuroscience of memory. The book covers the cellular mechanisms of memory, working memory, and declarative memory, as well as emotional memory. It considers both the neural and psychological underpinnings of learning and memory and provides an excellent introduction to additional topics such as memory consolidation and neuroanatomical compartmentalization of memory systems.

    Find this resource:

  • Gormezano, Isadore, and Edward A. Wasserman. 1992. Learning and memory: The behavioral and biological substrates. Hillsdale, NJ: Lawrence Erlbaum.

    Save Citation »Export Citation »E-mail Citation »

    A collection of articles from researchers invited to participate in a symposium on learning and memory. The text provides a depiction of the challenges and advances in the study of the biological basis of learning and memory in the early 1990s. As such, it illustrates how knowledge has moved forward through applying sophisticated biological techniques to well-defined behavioral paradigms.

    Find this resource:

  • Kesner, Raymond P., and Joe L. Martinez. 2007. Neurobiology of learning and memory. 2d ed. Boston: Butterworth-Heinemann.

    Save Citation »Export Citation »E-mail Citation »

    This book is intended for graduate students interested in behavioral neuroscience. The information is organized into sections on the developmental and genetic contributions to learning and memory, the various different neural systems that have been shown to mediate learning and memory, and the application of the empirical findings to clinical problems.

    Find this resource:

  • Macphail, E. M.. 1993. The neuroscience of animal intelligence: From the seahare to the seahorse. Animal Intelligence. New York: Columbia Univ. Press.

    Save Citation »Export Citation »E-mail Citation »

    This book is intended for advanced undergraduate and graduate psychology students. It explores the behavioral and neural mechanisms of nonassociative learning in invertebrates (e.g., Aplysia) as well as associative learning in invertebrates and vertebrates. The book synthesizes psychological, neurobiological, and computational approaches to the study of learning and memory and is relatively unique because of its strong emphasis on learning theory and attempts to identify commonalities across different model systems.

    Find this resource:

  • Rudy, Jerry W. 2008. The neurobiology of learning and memory. Sunderland, MA: Sinauer.

    Save Citation »Export Citation »E-mail Citation »

    This book is intended for students who have a rudimentary background in psychology and neuroscience. The information is organized in terms of the cellular and molecular basis of learning and memory, how these molecular mechanisms are related to behavioral aspects of memory functioning, and the different neural systems that encode different types of learning and memory.

    Find this resource:

  • Steinmetz, Joseph E., Mark A. Gluck, Paul R. Solomon, and Richard F. Thompson. 2001. Model systems and the neurobiology of associative learning: A festschrift in honor of Richard F. Thompson. Mahwah, NJ: Lawrence Erlbaum.

    Save Citation »Export Citation »E-mail Citation »

    This book is a collection of articles by former students, postdoctoral fellows, and colleagues of Richard Thompson and includes chapters on the mechanisms of associative learning in a wide variety of different model preparations, including Pavlovian and instrumental preparations.

    Find this resource:

Journals

There are many different journals that publish empirical and review articles on the neuroscience of associative learning. The neuroscience specialist journals Nature Neuroscience, The Journal of Neuroscience, Neuron, Nature Reviews Neuroscience, Neuroscience and Biobehavioral Reviews, and Trends in Neuroscience regularly include articles on the neuroscience of associative learning. However, there are also a number of specialist journals such as Learning and Memory and Neurobiology of Learning and Memory that focus almost exclusively on the neuroscience of associative learning. While Behavioral Neuroscience is not purely a learning journal, it places a heavy emphasis on the neuroscience of associative learning.

Pavlovian Conditioning

Pavlovian or classical conditioning is an associative learning process that is the result of the pairing of a neutral conditioned stimulus (CS; e.g., a tone) with a biologically significant unconditioned stimulus (US; e.g., an airpuff to the eye) in a predictive fashion. The US is able to elicit a reflex response, termed the unconditioned response (UR; e.g., a blink in response to the airpuff to the eye) prior to pairings of the CS and the US. As a consequence of repeated pairings of the CS and the US, the CS comes to elicit an anticipatory conditioned response (CR; e.g., a blink to the tone in anticipation of the airpuff to the eye). While there are many different Pavlovian conditioning preparations that have been studied, there are only a few model preparations in which the neurobiology of the associative learning process has been systematically examined. The model preparations that have been the most extensively studied and that we cover in this bibliography are Pavlovian conditional fear, Pavlovian conditioning of the eyeblink response, and conditioning of Pavlovian appetitive responses.

Fear Conditioning

Pavlovian fear conditioning involves the pairings of a CS with an aversive US. Typically, this US is a mild electric shock, but other USs can be equally effective in humans and other animals (e.g., loud noises). As a consequence of these pairings, animals display a diverse but correlated range of conditioned fear responses upon subsequent presentations of the CS, including suppression of ongoing appetively motivated behavior (conditioned suppression), species-typical defense responses (e.g., freezing), enhanced startle reactivity (fear potentiated startle), and changes in heart rate, respiration, and blood pressure, as well as analgesia. The neural circuitry of fear learning is increasingly well understood. This knowledge has come largely from studies of rats measuring the species-typical defense response of freezing or potentiation of the acoustic startle response. The seminal work by Davis, LeDoux, Fanselow, and others (reviewed in Davis 1992 and Davis 2006) has identified key roles for the amygdala, and its distinct subnuclei, in the acquisition of fear learning as well as the formation and storage of fear memories. Two key early studies are Kapp, et al. 1979, which identified the amygdala as a critical locus for Pavlovian fear learning, and Quirk, et al. 1995, which was the first to describe conditioning related plasticity in rat amygdala neurons. Maren and Quirk 2004 reviews the neurophysiology of fear conditioning. Kim and Jung 2006 provides an overview of the neural circuits and neuropharmacology of fear conditioning. Recent work has further delineated how classic psychological learning rules for associative learning may be instantiated in the activity of these circuits. Other studies involving genetically modified mice (e.g., Haubensak, et al. 2010) have provided important insights into the cellular basis of fear learning. Finally, studies using functional neuroimaging in humans, such as Büchel, et al. 1998, have extended these findings to normal as well as clinically anxious populations.

Role of the Hippocampus

The role of hippocampus in Pavlovian fear conditioning is controversial. The hippocampus is important under some conditions for the encoding and consolidation of contextual fear memories. Holland and Bouton 1999 reviews the critical empirical findings up to 1999 on the role of hippocampus in contextual learning. In fear conditioning, the hippocampus is believed to mediate the processes whereby the diverse array of cues comprising the experimental context (e.g., the tactile, visual, olfactory, and auditory features of the conditioning chamber) are integrated into a unified representation. This contextual representation is then stored, temporarily, in the hippocampus during a period of contextual memory consolidation prior to being stored more permanently in the cortex. Manipulations of hippocampal function affect the formation, storage, and consolidation of these contextual representations. These same manipulations typically do not affect the association of that contextual representation with shock (if the representation is already consolidated) or the association of discrete CSs such as discrete auditory CS. There have been important and powerful demonstrations of these dissociations in the literature, such as those reported in Kim and Fanselow 1992 as well as Anagnostaras, et al. 1999, but there have also been failures to detect such dissociations. The difficulties associated with the study of hippocampal function in fear conditioning are reviewed in Maren 2008.

Fear Extinction

Extinction refers to the reduction in a CR when the contingency between the CS and US is broken. Typically, this contingency is broken via presenting the CS alone, in the absence of the US. As a consequence of these CS-alone presentations, fear CRs are reduced. This reduction in the CR usually does not constitute erasure of the original CS–US association. Rather, a wealth of evidence shows that the original CS–US association survives extinction training. Thus, fear extinction involves imposition of a mask on the fear response. The neural mechanisms for extinction learning are increasingly well understood. Falls, et al. 1992 was the first to identify a role for glutamate neurotransmission in the amygdala for fear extinction learning. Davis 2002 reviews the literature to that date on the role of glutamate neurotransmission, as well as the associated intracellular signaling cascades, in fear extinction learning. The amygdala plays an important role in both the expression of fear as well as the expression of fear extinction. Herry, et al. 2008 shows how these functions are served by distinct populations of amygdala neurons, with each population located in a different neural circuitry. The prefrontal cortex is also important for inhibiting the expression of fear after extinction training. Milad and Quirk 2002 has been instrumental in identifying prefrontal contributions to extinction expression and fear inhibition. Human neuroimaging findings reported in Phelps, et al. 2004 have shown that similar brain regions are recruited during fear extinction learning in humans. Moreover, Ressler, et al. 2004 in clinical studies exploited this basic scientific knowledge to develop novel adjuncts to augment the efficacy of exposure-based treatments for human anxiety. The promises, and pitfalls, of this approach to translational research are reviewed in Ressler and Mayberg 2007.

Eyeblink Conditioning

While eyeblink conditioning has been examined in many different mammalian species, the majority of the early studies on the neurobiology of the conditioned eyeblink response was carried out on the rabbit. Thompson 1986 provides a concise summary of early work done in the rabbit identifying the circuits involved for learning the basic conditioned eyeblink response. The eyeblink response in the rabbit involves movement of the external eyelid, retraction of the eyeball, and movement of the nictitating membrane, the translucent third eyelid, and it is usually the movement of the nictitating membrane that is assessed. More recently, researchers have begun to investigate eyeblink conditioning in rats and mice as a result of new techniques developed to measure the eyeblink response in the moving animal, which has allowed for an examination of eyeblink conditioning in mutant and transgenic animals. As a consequence of the large amount of work that has been done on the neural substrates of this form of associative learning, the conditioned eyeblink response is perhaps the most well-delineated form of mammalian associative learning. Christian and Thompson 2003 provides an excellent review of the work identifying the neural circuitry involved in learning the conditioned eyeblink response. Although eyeblink conditioning is a very simple form of associative learning, there are a number of different brain areas that are involved, predominately the motor nuclei, which are involved in the production of the UR and the CR; the cerebellum, which is critically involved in learning the CR; and the hippocampus, which modulates the eyeblink response. The volume on neuroscience of eyeblink conditioning (Woodruff-Pak and Steinmetz 2000) contains chapters on the evidence for the involvement of each of these brain areas in eyeblink conditioning. The subsequent sections examine CS, CR, US, and UR pathways; the critical role of the cerebellum; and the studies of role of the hippocampus in eyeblink conditioning.

CS, US, CR, and UR Pathways

As a result of the work conducted in the rabbit, the essential circuitry for CS, US, UR, and CR pathways in eyeblink conditioning have been well established. With respect to the CS pathway, information about auditory, visual, and somatosensory stimuli is transmitted to the cerebellum via mossy fibers from the pontine nucleus in the pons. It has been shown in Steinmetz, et al. 1986 that electrical stimulation of the mossy fibers in the pontine nucleus and the middle cerebellar peduncle can serve as a CS. Subsequent work in Steinmetz 1990 has shown that when animals are trained with electrical stimulation of the pontine nucleus as CS, in some animals at least, there is a complete transfer of the behavioral CR to a tone CS. Therefore, it appears that mossy fiber activity coming from the pontine nuclei provides the CS-related activity necessary for eyeblink conditioning. With respect to the US pathway, information about the periorbital shock or airpuff to the eye, which is typically used as the US in eyeblink conditioning, is transferred to the cerebellum via climbing fibers that originate in the inferior olivary nucleus in the medulla. Mauk, et al. 1986 showed that electrical stimulation of the inferior olive can serve as a US and lead to the acquisition of the CR. Furthermore McCormick, et al. 1985 showed that lesions of the inferior olive result in gradual extinction of the CR in well-trained animals even when they continue to experience tone-airpuff trials. Therefore, it appears that the climbing fiber activity coming from the inferior olive provides the US-related activity necessary for eyeblink conditioning. With respect to the CR, McCormick and Thompson 1984 demonstrated that regions of the interpositus nucleus ipsilateral to the trained eye show neuronal signals that precede and predict the emergence of the CR, and that lesions of the interpositus nucleus abolish the CR. However, these lesions had no effects on the performance of the UR, indicating that the interpositus nucleus is essential to the CR pathway but not the UR pathway. Conversely, lesion of the accessory abducens nucleus, one of the cranial motor nuclei, has been shown in Disterhoft, et al. 1985 to abolish both CRs and URs. Therefore, it appears that the same motor nuclei are involved in the production of both the CR and the UR.

Critical Involvement of the Cerebellum

From the work referenced on the essential pathways of the CS, US, CR, and UR, it is apparent that the cerebellum, a brain structure that overlies the dorsal surfaces of the brain stem, is critical for the acquisition and expression of the eyeblink CR. The first study to identify the cerebellum as having an important role in eyeblink conditioning was McCormick, et al. 1981 and showed that large aspiration lesions of the cerebellar cortex and nuclei and also electrolytic lesions of the dentate-interpositus nuclear region abolished eyeblink conditioning. Subsequently, McCormick and Thompson 1984 showed that neuronal recording from the cerebellar cortex and cerebellar nuclei showed evidence of activity that preceded and predicted the occurrence of the CR. Although neuronal recording and lesion studies have implicated the cerebellum in eyeblink conditioning, it has been more challenging to determine the relative contribution of the cerebellar cortex and nuclei to the formation of the association. The evidence, which is reviewed in Thompson and Steinmetz 2009, suggests that synaptic/neuronal plasticity in the cerebellar interpositus nucleus is essential for the acquisition and retention of the CR, but that the synaptic/neuronal plasticity that occurs in the cerebellar cortex plays a role in the acquisition of adaptive timing of the CR. The results of clinical studies in humans with brain damage parallel the animal data and show that appropriate cerebellar lesions either prevent or impair eyeblink conditioning. Gerwig, et al. 2007 provides an excellent review of the work done in humans showing that the cerebellum is critically involved in eyeblink conditioning in humans. The mechanisms of synaptic plasticity that occurs in the cerebellar cortex have been studied extensively, and there is good evidence that eyeblink conditioning results in long-term depression at the granule to Purkinje cell synapse, evidence that is reviewed in Linden and Connor 1993. The mechanisms of synaptic/neuronal plasticity in the cerebellar nucleus are less well understood, although there is evidence in the rat in Kleim et al. 2002 to suggest that there is a significant increase in the number of excitatory synapses in the interpositus nucleus as a result of paired presentations of the CS and US.

  • Gerwig, M., F. P. Kolb, and D. Timmann. 2007. The involvement of the human cerebellum in eyeblink conditioning. The Cerebellum 6.1: 38–57.

    DOI: 10.1080/14734220701225904Save Citation »Export Citation »E-mail Citation »

    This article reviews the recent human data, with a particular emphasis on lesion data and functional brain imaging, on the neural correlates of eyeblink conditioning. The authors conclude that the human data is concordant with the animal literature but that the particular limitations of the human data make it difficult to draw conclusions about the relative contributions of the cerebellar cortex and nuclei. Available online for purchase or by subscription.

    Find this resource:

  • Linden, David J., and John A. Connor. 1993. Cellular mechanisms of long-term depression in the cerebellum. Current Opinion in Neurobiology 3.3: 401–406.

    DOI: 10.1016/0959-4388(93)90133-JSave Citation »Export Citation »E-mail Citation »

    Reviews the evidence with respect to the cellular mechanisms involved in long-term depression in the parallel fiber Purkinje neuron synapse that are induced by coactivation of parallel fibers and climbing fibers. Available online for purchase or by subscription.

    Find this resource:

  • Kleim, Jeffrey A., John H. Freeman Jr., Rochelle Bruneau, et al. 2002. Synapse formation is associated with memory storage in the cerebellum. Proceedings of the National Academy of Sciences 99.20: 13228–13231.

    DOI: 10.1073/pnas.202483399Save Citation »Export Citation »E-mail Citation »

    Showed that rats exposed to CS–US pairings showed significantly more synapses per neuron within the cerebellar interpositus than either explicitly unpaired or untrained controls. This is the first article to show synaptogensis as a possible mechanism for the plasticity that occurs in the cerebellar nucleus during eyeblink conditioning.

    Find this resource:

  • McCormick, David A., David G. Lavond, Gregory A. Clark, Ronald E. Kettner, Christina E. Rising, and Richard F. Thompson. 1981. The engram found? Role of the cerebellum in classical conditioning of nictitating membrane and eyelid responses. Bulletin of the Psychonomic Society 18:103–105.

    Save Citation »Export Citation »E-mail Citation »

    This article provided the first clear evidence that the cerebellum is involved in eyelid conditioning.

    Find this resource:

  • McCormick, D. A., and R. F. Thompson. 1984. Neuronal responses of the rabbit cerebellum during acquisition and performance of a classically conditioned nictitating membrane-eyelid response. Journal of Neuroscience 4.11: 2811–2822.

    Save Citation »Export Citation »E-mail Citation »

    Showed that neuronal responses in the cerebellar cortex and the dentate-interpositus nucleus of rabbits display learned changes in their pattern of firing that precede and predict the behavioral response. Also showed that lesions of the dentate-interpositus abolished the eyeblink CR.

    Find this resource:

  • Thompson, R. F., and J. E. Steinmetz. 2009. The role of the cerebellum in classical conditioning of discrete behavioural responses. Neuroscience 162:732–755.

    DOI: 10.1016/j.neuroscience.2009.01.041Save Citation »Export Citation »E-mail Citation »

    Review of the literature on the role of the cerebellar cortex and the cerebellar nucleus in eyeblink conditioning. They argue that the evidence indicates that the interpositus nucleus is essential for eyeblink conditioning, whereas the cerebellar cortex contributes to normal acquisition and adaptive timing of the conditioned response. Available online for purchase or by subscription.

    Find this resource:

Studies of the Hippocampus

Early work on the neural substrates of eyeblink conditioning in Berger, et al. 1976 indicated that cells in the hippocampus show learning-related changes in their pattern of firing that precede and predict the emergence of the CR. Recent work, reviewed in Gruart and Delgado-Gacía 2007, has elucidated the synaptic changes that occur in the hippocampus during eyeblink conditioning in mice. However, Schmaltz and Theios 1972 showed that the hippocampus was not critical for eyeblink conditioning, as rabbits with lesions of the hippocampus could learn and retain the conditioned eyeblink response. Nonetheless, the hippocampus is thought to be important in modulating the rate of acquisition of eyeblink conditioning. Berry and Hoffman 2011 summarizes the findings that indicate that hippocampal theta activity has an important role in modulating cerebellar activity in eyeblink conditioning. Although hippocampal lesions do not affect simple eyeblink conditioning, Moyer, et al. 1990 shows that hippocampal lesions abolish a more complex form of learning, trace eyeblink conditioning. In trace conditioning, the CS is turned on and off before the presentation of the US so that there is a gap, or trace interval, between the presentation of the CS and the US. Woodruff-Pak and Disterhoft 2008 reviews the evidence for the role of the hippocampus as it relates to the cerebellum in trace eyeblink conditioning. Hippocampal lesions also impair other, more complex forms of eyeblink conditioning such as discrimination reversal and conditional discriminations. Somewhat controversially, Clark and Squire 1998 suggests that when the hippocampus is critically involved in eyeblink conditioning, this is associated with awareness of the stimulus on the part of the participant, implying that awareness is required for learning when the hippocampus is required for learning.

Appetitive Conditioning

Pavlovian appetitive conditioning involves the pairings of a CS with an appetitive US—for example, a food pellet or delivery of sucrose. As a consequence of these pairings, animals display conditioned approach behaviors toward the food source (e.g., head entries in a magazine containing the food). Rats and other animals also show distinct orienting responses to the CS itself. These responses may include eye movements (saccades) toward a visual CS signaling food as well as larger movements of the body. The topography of these CS-generated responses depends on the nature of the CS. In rats, for example, visual CSs elicit rearing, whereas auditory CSs elicit jerking of the head toward the source of the CS. Our understanding of the neuroscience of appetitive conditioning has flourished in recent years. The first major development was the use of electrophysiological recordings in primate midbrain dopamine neurons during appetitive conditioning tasks. It is well known that the neurotransmitter dopamine is critical for appetitive learning and reward. Elegant work by Schultz and colleagues (see Schultz 2002 and Schultz 2006), as well as by Matsumoto and Hikosaka (see Matsumoto and Hikosaka 2009), has shown that the firing of midbrain dopamine neurons conforms to the prediction of formal associative learning theory. The canonical findings from recordings in primates during Pavlovian appetitive conditioning, as exemplified in Waelti, et al. 2001, are that midbrain dopamine neurons show increases in firing to unexpected rewards, little change in firing to expected rewards, and inhibited firing to omission of an expected reward. Concurrently, these neurons show no increase in firing to the CS at the start of training but show an increase in firing rates to the CS as it comes to predict the US. Bromberg-Martin, et al. 2010 provides an important analysis of whether similar changes in firing are observed during fear learning.

Brain Areas

The second major development is the use of electrophysiological recordings in rodents, as well as classical lesion and reversible inactivation techniques of behavioral neuroscience, to identify the forebrain and cortical brain regions critical for appetitive learning. The role for dopamine in appetitive learning is well established, and findings regarding the relevant electrophysiological properties of dopamine neurons were described previously here. However, such electrophysiological data provide correlative evidence. Takahashi, et al. 2009 provides causal evidence that midbrain dopamine neurons are involved in signaling prediction error during Pavlovian appetitive conditioning. In addition to midbrain dopamine, the amygdala and the orbitofrontal cortex play important roles in appetitive learning. Critical to the emergence of this field was the work of Gallagher and Holland (see Gallagher and Holland 1994), which demonstrated an important role for the amygdala in Pavlovian appetitive conditioning. This work showed unequivocally that the amygdala is important for distinct aspects of Pavlovian appetitive that were in addition to its previously understood role in Pavlovian aversive or fear conditioning. One important role involves determining learned variations in attention to appetitive CSs as a function of how well the CS predicts its consequences, specifically with the central nucleus of the amygdala important for the upregulation of attention to CSs whose consequences are uncertain. Two empirical demonstrations of this role are provided in Calu, et al. 2010 and Holland and Gallagher 2006. Amygdala interactions with prefrontal cortex are also important for appetitive learning (Holland and Gallagher 2004). The role of the orbitofrontal cortex in Pavlovian appetitive conditioning is reviewed in Schoenbaum, et al. 2009, which considers several lines of evidence that the orbitofrontal cortex is signaling outcome expectancies during Pavlovian appetitive conditioning. Roesch, et al. 2010 provides a theoretical integration of these findings regarding dopamine, amygdala, and prefrontal cortex in Pavlovian appetitive learning.

Instrumental Conditioning

Instrumental or operant conditioning is an associative learning process whereby behaviors are strengthened or weakened as a result of their consequences. The behavior involved is typically instrumental in producing the resultant consequences; that is, the response made by the organism affects the delivery of the reinforcing stimulus, so this type of learning is often referred to as instrumental conditioning. Unlike Pavlovian Conditioning, the behaviors in instrumental conditioning do not automatically occur as a result of any particular stimulus but are voluntary or nonreflexive behaviors. Because motivation plays an important part in instrumental conditioning, and the types of behaviors involved are much more variable, the neural circuits that underlie instrumental conditioning are much more complex than those that subserve simple forms of Pavlovian conditioning. Consequently, the neural substrates of instrumental conditioning have not been as clearly delineated as those involved in Pavlovian conditioning. However, in recent years there has been considerable research on the neural substrates of appetitive instrumental conditioning and instrumental avoidance conditioning. In the case of appetitive instrumental conditioning, this increase in research interest has been based on important theoretical advancements in understanding the psychology of instrumental learning. In the case of instrumental avoidance learning, this increase in research interest has been based on the recognition that studies of Pavlovian fear conditioning fail to adequately capture several features of the multifaceted fear response.

Appetitive Instrumental Conditioning

Animals readily learn to emit actions that procure a rewarding outcome. In the laboratory setting, rats are typically trained to lever-press for a food reward. Seminal behavioral work by Dickinson, Balleine, and colleagues, reviewed in Balleine and Dickinson 1998, has identified and described the multiple learning processes that control behavior in this setting as a function of (1) the animal’s motivational state (i.e., instrumental incentive learning), (2) the presence of other stimuli (a Pavlovian CS) predictive of the reward (i.e., Pavlovian instrumental transfer), and (3) the amount and kind of lever-press training (i.e., goal-directed versus habit learning). This behavioral research, in turn, has been used to guide investigations into the neural mechanisms of instrumental appetitive conditioning. This literature, including data from primates and rodents, is reviewed in Balleine and O’Doherty 2010. Among the key findings from this literature are those reported in (1) Yin, et al. 2008, that whereas ventral parts of the striatum are important for Pavlovian appetitive learning, more dorsal parts of the striatum are important for instrumental appetitive learning; (2) Corbit, et al. 2001 and Corbit and Balleine 2005, that the amygdala as well as ventral striatum are important for Pavlovian–instrumental transfer effects; and (3) Balleine and Doherty 2010 and Killcross and Coutureau 2003, that the transition from goal-directed to habit-based responding is associated with recruitment of distinct striatal and prefrontal cortical circuits. The involvement of the amygdala in Pavlovian appetitive learning is especially interesting given the role of amygdala in fear learning. Balleine and Killcross 2006 provides an important analysis of these two roles and proposes a novel model of amygdala function in associative learning.

Instrumental Avoidance Conditioning

A stimulus that signals imminent threat or danger elicits various conditioned fear responses, as discussed in the section on fear conditioning, but it also results in the animal preparing to avoid or escape the danger. These instrumental responses and the neural substrates that underlie them have been examined using instrumental avoidance procedures. There are various different procedures that have been used to study instrumental avoidance learning, such as passive avoidance, escape from fear, unsignaled (Sidman) avoidance, and active avoidance. Generally it is thought that a conditioned fear response is necessary to motivate the avoidance behavior, and consequently there has been considerable research on the role of the amygdala in instrumental avoidance learning. The most important initial study was Killcross, et al. 1997, which reports a double dissociation between the role of different amygdala nuclei in Pavlovian fear conditioning and instrumental avoidance learning. More recently, Choi, et al. 2010 extended this work and showed that the central nucleus of the amygdala may in fact constrain instrumental avoidance learning by the generation of a competing Pavlovian response, such as freezing. Cain and LeDoux 2008 provides a review of the evidence for the involvement of the amygdala in escape from fear and signaled active avoidance learning and concludes that the lateral nucleus and the basal nucleus but not the central nucleus of the amygdala is important for instrumental avoidance learning. Michael Gabriel and colleagues have made significant contributions to understanding the brain mechanisms involved in signaled active avoidance learning in the rabbit. This work, which is reviewed in Gabriel and Talk 2001, emphasizes the involvement of multiple different brain regions, including the amygdala, thalamus, cingulate cortex, and auditory cortex in the coordination of the learning and performance involved in instrumental avoidance conditioning. Although we still have only a rudimentary picture of how all of these different brain regions are integrated in instrumental avoidance conditioning, Gabriel, et al. 2003 presents a hypothesis about how the amygdala might initiate learning-related plasticity in other parts of the brain. Part of the interest in the mechanisms of instrumental avoidance learning is because avoidance is thought to play an important role in maintaining pathological fears in human anxiety disorders. Recent evidence from human imaging studies of instrumental avoidance during fear conditioning in Delgado, et al. 2009 showed that amygdala-striatal interactions underlie the acquisition of the avoidance response, which is consistent with the animal data.

  • Cain, Christopher K., and Joseph E. LeDoux. 2008. Brain mechanisms of Pavlovian and instrumental aversive conditioning. In Handbook of anxiety and fear. Edited by Robert J. Blanchard, D. Caroline Blanchard, Guy Greibel, and David Nutt, 103–124. Amsterdam: Academic Press.

    Save Citation »Export Citation »E-mail Citation »

    Reviews the literature on the brain areas and circuitry thought to underlie Pavlovian and instrumental aversive conditioning.

    Find this resource:

  • Choi, June-Seek, Christopher K. Cain, and Joseph E. LeDoux. 2010. The role of amygdala nuclei in the expression of auditory signaled two-way active avoidance in rats. Learning & Memory 17:139–147.

    DOI: 10.1101/lm.1676610Save Citation »Export Citation »E-mail Citation »

    Showed that discrete or combined lesions of the lateral and basal amygdala performed after acquisition of the avoidance response impaired subsequent avoidance responses, whereas a lesion of the central nucleus of the amygdala had little effect. However, lesions of the central amygdala were able to rescue avoidance responding in animals that had failed to acquire the avoidance response after three days of training.

    Find this resource:

  • Delgado, Mauricio R., Rita L. Jou, Joseph E. LeDoux, and Elizabeth A. Phelps. 2009. Avoiding negative outcomes: Tracking the mechanisms of avoidance learning in humans during fear conditioning. Frontiers in Behavioral Neuroscience 3:33.

    Save Citation »Export Citation »E-mail Citation »

    In a fear conditioning paradigm in humans, participants learned that one stimulus predicted the occurrence of a shock, a second stimulus predicted the absence of the shock, and a third stimulus predicted the occurrence of the shock that could be avoided if they chose the correct action. Successful learning of the avoidance response was associated with brain activity, as measured by fMRI, in the striatum.

    Find this resource:

  • Gabriel, M., L. Burhans, and A. Kashef. 2003. Consideration of a unified model of amygdala associative functions. Annals of the New York Academy of Sciences 985:206–217.

    DOI: 10.1111/j.1749-6632.2003.tb07083.xSave Citation »Export Citation »E-mail Citation »

    Attempts to integrate the models of the role of the amygdala in Pavlovian fear and appetitive conditioning with its role in instrumental conditioning, with particular reference to the influence of the various nuclei of the amygdala and other brain regions involved in instrumental avoidance conditioning.

    Find this resource:

  • Gabriel, Michael, and Andrew Talk. 2001. A tale of two paradigms: Lessons learned from parallel studies of discriminative instrumental learning and classical eyeblink conditioning. In Model Systems and the neurobiology of associative learning: A festschrift in honor of Richard F. Thompson. Edited by Joseph E. Steinmetz, Mark A. Gluck, Paul R. Solomon, and Richard Thompson, 149–185. Mahwah, NJ: Lawrence Erlbaum.

    Save Citation »Export Citation »E-mail Citation »

    Reviews the evidence with respect to the contributions of different brain regions in an active avoidance paradigm in the rabbit. The authors show how results from multisite, single-unit recording, pharmacology, and lesion studies have implicated the amygdala, thalamus, cingulate cortex, and auditory cortex in the acquisition and retention of instrumental avoidance learning.

    Find this resource:

  • Killcross, Simon, Trevor W. Robbins, and Barry J. Everitt. 1997. Different types of fear-conditioned behaviour mediated by separate nuclei within amygdala. Nature 388:377–380.

    DOI: 10.1038/41097Save Citation »Export Citation »E-mail Citation »

    Showed that lesions of the central nucleus of the amygdala impaired conditioned fear responses in rats but did not impair their ability to avoid further presentations of the CS, whereas lesions of the basolateral amygdala impaired avoidance of the CS but left conditioned fear responses intact. This represents a double dissociation between the role of the different amygdala nuclei in Pavlovian fear conditioning and instrumental avoidance learning. Available online for purchase or by subscription.

    Find this resource:

Associative Learning and the Hippocampus

The importance of the medial temporal lobe and the hippocampus in particular for memory was firmly established as the result of the temporal lobe resection in patients such as H. M. and others like him, neurological case evidence that has been systematically documented in Milner 1972. Subsequently, animal models of human memory capabilities have been developed, primarily in the monkey and the rat, to experimentally examine the anatomical substrates of various aspects of human memory. Cohen and Eichenbaum 1994 provides a review of this work on the role of the hippocampus in memory formation in humans and animals. Both the human and the animal work suggest that the hippocampus, as well as other medial temporal lobe structures, is important for the fast formation of associations between stimuli in different sensory modalities. This evidence is reviewed in the section on associative memory in humans and animals.

  • Cohen, Neal J., and Howard Eichenbaum. 1994. Memory, amnesia, and the hippocampal system. Cambridge, MA: MIT Press.

    Save Citation »Export Citation »E-mail Citation »

    This text focuses exclusively on the role of the hippocampal system in memory in humans and other animals. The authors review cellular, neuroanatomical, behavioral, and cognitive findings relating to the role of the hippocampus in memory.

    Find this resource:

  • Milner, B. 1972. Disorders of learning and memory after temporal lobe lesions in man. Clinical Neurosurgery 19:421–446.

    Save Citation »Export Citation »E-mail Citation »

    Provides a summary of the types of deficits of learning and memory that are seen in humans as a consequence of damage to the temporal lobe.

    Find this resource:

Associative Memory in Humans and Animals

Associative memory is the ability to learn and remember the association between two items, such as the name and appearance of a person someone just met or where an individual left his car keys. There is a growing body of evidence to suggest that the hippocampus, with the surrounding entorhinal, perirhinal, and parahippocampal cortices, are important for encoding the associations among stimuli. For an excellent review of this evidence in both humans and animals, see Squire et al. 2004. The study of the neuroscience of associative memory is aided by the possibility of assessing similar questions in both humans and nonhuman animals. To examine the neural mechanisms of associative memory in animals, a number of tasks have been developed that require the formation of new associations to perform appropriately. In rats, the most extensively used tasks have been the paired odor task developed by Michael Bunsey and Howard Eichenbaum (see Bunsey and Eichenbaum 1993) and the flavor-location task developed by M. Day (see Day, et al. 2003). Langston, et al. 2010 reviews the evidence from rats on the role of the various hippocampal subregions in these forms of associative memory. In monkeys, the acquisition and expression of associative memory have been examined predominately using a scene-location task. Suzuki 2007 provides a review of work examining changes in neural activity in monkeys during the acquisition and expression of scene-location associations. There is some controversy about whether the hippocampus is specialized for learning the association between unrelated items, rather than a more general ability to remember items. Giovanello, et al. 2003 provides human evidence to suggest that the hippocampus is specialized for learning associations, but Stark and Squire 2003 presents alternative evidence to suggest the hippocampus is equally important for single-item memory as it is for conjunctions. In a similar vein, there is controversy about the role of the hippocampus in familiarity and recollection of items. Familiarity involves knowing that the item has been seen before but does not involve knowledge about the context in which the item was seen, whereas recollection involves knowing that the item has been seen before and in what particular context it was seen. Eichenbaum, et al. 2007 provides a comprehensive review of the neuropsychological, neuroimaging, and neurophysiological studies of humans, monkeys, and rats with respect to the role of the subregions of the medial temporal lobe in familiarity and recollection. The authors conclude that the hippocampus is critical for recollection but not for familiarity.

  • Bunsey, Michael, and Howard Eichenbaum. 1993. Critical role of the parahippocampal region for paired-associate learning in rats. Behavioral Neuroscience 107.5: 740–747.

    DOI: 10.1037/0735-7044.107.5.740Save Citation »Export Citation »E-mail Citation »

    Developed a paired-associate task in which rats were exposed to pairs of odors that were rewarded and to recombinations of these pairs of odors that were not rewarded. The authors showed that normal animals were able to learn this task but that rats with parahippocampal damage failed to learn. Available online for purchase or by subscription.

    Find this resource:

  • Day, M., R. Langston, and R. G. M. Morris. 2003. Glutamate-receptor-mediated encoding and retrieval of paired-associate learning. Nature 424:205–209.

    DOI: 10.1038/nature01769Save Citation »Export Citation »E-mail Citation »

    Developed a paired-associate task in which the flavor of a food was paired with a particular spatial location. The authors showed that encoding and initial retrieval of the flavor-location paired associate depends on the hippocampus.

    Find this resource:

  • Eichenbaum, H., A. P. Yonelinas, and C. Ranganath. 2007. The medial temporal lobe and recognition memory. Annual Review of Neuroscience 30:123–152.

    DOI: 10.1146/annurev.neuro.30.051606.094328Save Citation »Export Citation »E-mail Citation »

    Reviews the literature on the role of the medial temporal lobe in recollection and familiarity based recognition. The authors conclude that the hippocampus is critical for recollection, whereas the perirhinal cortex is necessary for familiarity-based recognition. Available online for purchase or by subscription.

    Find this resource:

  • Giovanello, Kelly Sullivan, M. Verfaellie, and Margaret M. Keane. 2003. Disproportionate deficit in associative recognition relative to item recognition in global amnesia. Cognitive, Affective, & Behavioral Neuroscience 3.3: 186–194.

    DOI: 10.3758/CABN.3.3.186Save Citation »Export Citation »E-mail Citation »

    Found that patients with relatively selective hippocampal damage performed within normal limits on an item-recognition test but were significantly impaired in word-word recognition in a word-word learning task.

    Find this resource:

  • Langston, Rosamund F., Cassie H. Stevenson, Claire L. Wilson, Ian Saunders, and Emma R. Wood. 2010. The role of hippocampal subregions in memory for stimulus associations. Behavioural Brain Research 215.2: 275–291.

    DOI: 10.1016/j.bbr.2010.07.006Save Citation »Export Citation »E-mail Citation »

    Reviews the literature on the roles of the various hippocampal subregions in associative recognition tasks in rodents. Available online for purchase or by subscription.

    Find this resource:

  • Squire, Larry R., Craig E. L. Stark, and Robert E. Clark. 2004. The medial temporal lobe. Annual Review of Neuroscience 27:279–306.

    DOI: 10.1146/annurev.neuro.27.070203.144130Save Citation »Export Citation »E-mail Citation »

    Reviews the findings from humans, rats, and monkeys on the function of the hippocampus and the perirhinal, entorhinal, and parahippocampal cortices. The authors discuss some of the controversies about the specificity of function of the hippocampus and the adjacent medial temporal lobe structures, such as associative versus nonassociative memory, episodic versus semantic memory, and recollection versus familiarity. Available online for purchase or by subscription.

    Find this resource:

  • Stark, Craig E., and Larry R. Squire. 2003. Hippocampal damage equally impairs memory for single items and memory for conjunctions. Hippocampus 13.3: 281–292.

    DOI: 10.1002/hipo.10085Save Citation »Export Citation »E-mail Citation »

    Found that patients with hippocampal damage were equally impaired at distinguishing pairs of words that they had previously seen from novel pairings of familiar words, pairs of words with one old and one new, and entirely novel pairs where both words were new. Thus hippocampal damage did not seem to selectively impair the associative component of memory. Available online for purchase or by subscription.

    Find this resource:

  • Suzuki, Wendy A. 2007. Making new memories: The role of the hippocampus in new associative learning. Annals of the New York Academy of Sciences 1097:1–11.

    DOI: 10.1196/annals.1379.007Save Citation »Export Citation »E-mail Citation »

    This article reviews the behavioral and neurophysiological studies of new association formation in the macaque monkey with a particular focus on the role of the hippocampus in associative memory formation. Available online for purchase or by subscription.

    Find this resource:

LAST MODIFIED: 11/29/2011

DOI: 10.1093/OBO/9780199828340-0080

back to top

Article

Up

Down