Themes and highlights



Salzman Lab

Columbia University Medical Center

Department of Neuroscience

Department of Psychiatry

Columbia Neuroscience

Mahoney-Keck Center for Brain and Behavior Research

Neural Mechanisms Underlying
Emotional Learning and Behavior

Emotions often occur in response to sensory stimuli that have been associated with either rewards or punishments. The Salzman lab studies the neural mechanisms that underlie our ability to learn such associations, and our ability to regulate our emotions based on those associations. We conduct neurophysiological experiments in behaving animals, trying to understand the relationship between neural activity and processes such as appetitive and aversive reinforcement learning, the regulation and control of neural representations during and after learning, and the influence of emotions on different cognitive processes.
Initially, our work has investigated the physiological responses of amygdala neurons during reinforcement learning. The amygdala is a limbic brain structure likely to be critical to the process of associating sensory stimuli with reinforcement. Thus far, we have found evidence that amygdala neural responses to visual stimuli reflect the value -- or valence -- of reinforcement associated with such visual stimuli. Some neurons encode positive value, i.e. they respond more strongly to visual stimuli associated with rewards than with punishments. Other neurons do the opposite, encoding negative value. Furthermore, we have discovered that the amygdala processes surprising reinforcement in a manner consistent with its role in different aspects of emotion. Surprising rewards or punishments are known to trigger a variety of cognitive and emotional responses. Different populations of amygdala neurons respond preferentially to either or both unexpected rewards and punishments. These neuronal populations reflect the role of the amygdala in valence-nonspecific processes, such as the enhancement of attention, arousal, and memory formation, as well as valence-specific emotional processes, such as fear or reward-seeking behavior.

In current experiments, we now often monitor neural activity in multiple brain areas simultaneously to understand the complex neural circuitry that underlies emotional learning and behavior. In particular, we have begun to study the orbitofrontal cortex (OFC) and the amygdala together, since OFC is intimately connected with the amygdala. In the future, we are interested in trying to delineate neural circuits more precisely by using techniques that could reveal the specific synaptic mechanisms and connections responsible for reinforcement learning and emotional regulation.


Click on any person's name to see where they are in the Neurotree

Daniel Salzman
Assistant Professor

cds2005 [at] columbia [dot] edu
212.543.6931 x400

mailing address:
C. Daniel Salzman, M.D., Ph.D.
Depts. of Neuroscience and Psychiatry
Columbia University
1051 Riverside Drive, Unit 87
NYSPI Kolb Research Annex, Rm 561
New York, NY 10032

Karen Marmon
Research Technician

km2432 [at] columbia [dot] edu

Alex Saez
Postdoctoral Fellow

as3171 [at] columbia [dot] edu

Chris Peck
Graduate Student

cp2413 [at] columbia [dot] edu

Ellen Peck
Graduate Student

eld2123 [at] columbia [dot] edu

Jerome Munuera
Postdoctoral Fellow

jm3570 [at] columbia [dot] edu

Jalal Baruni
Graduate Student

jkb2128 [at] columbia [dot] edu

Latoya Palmer
Mahoney Center Administrator

lp2104 [at] columbia [dot] edu
212.543.6931 x100

Holly Cline
Lab Administrator

hc2471 [at] columbia [dot] edu
212.543.6931 x102


Marina Belova
Graduate Student, 2002-2008

Life Sciences Specialist (Associate Consultant)
L.E.K. Consulting, Boston, MA

belovamarina [at] hotmail [dot] com

Joe Paton
Graduate Student, 2002-2008

Group Leader
Champalimaud Neuroscience Program
Instituto Gulbenkian de Ciencia
P-2781-901 Oeiras, Portugal

jpaton [at] igc [dot] gulbenkian [dot] pt

Nicholas Macfarlane
Research Technician, 2008-2009

Sara Morrison
Graduate Student, 2004-2010

Research Associate
Nicola lab, Dept. of Psychiatry
Albert Einstein College of Medicine, Bronx, NY

sara [dot] morrison [at] einstein [dot] yu [dot] edu

Crista Barberini
Postdoctoral Research Scientist

crista [at] stanfordalumni [dot] org

Brian Lau
Postdoctoral Fellow

brianlau [dot] phd [at] gmail [dot] com

Rebecca Schoer Saez
Graduate Student

ras2159 [at] columbia [dot] edu

Anna Ipata
Postdoctoral Fellow

ai2019 [at] columbia [dot] edu

February 2008

    Morrison SE, Saez A, Lau B, Salzman CD. (2011). Different time courses for learning-related changes in amygdala and orbitofrontal cortex. Neuron, 71, 1127-1140. [PDF]

    Morrison SE, Salzman CD. (2011). Representations of appetitive and aversive information in the primate orbitofrontal cortex. Ann N Y Acad Sci, 1239, 59-10. [PDF]

    Morrison SE, Salzman CD. (2010). Re-valuing the amygdala. Curr Opin Neurobiol, 20, 221-230. [PDF]

    Salzman CD, Fusi S. (2010). Emotion, cognition, and mental state representation in amygdala and prefrontal cortex. Annu Rev Neurosci, 33, 173-202. [PDF]

    Rigotti M, Ben Dayan Rubin D, Morrison SE, Salzman CD, Fusi S. (2010). Attractor concretion as a mechanism for the formation of context representations. Neuroimage, 52, 833-847. [PDF]

    Lau B, Salzman CD. (2009). The rhythms of learning. Nat Neurosci, 12, 675-676. [PDF]

    Morrison SE, Salzman CD. (2009). The convergence of information about rewarding and aversive stimuli in single neurons. J Neurosci, 29, 11471-11483. [PDF]

    Lau B, Salzman CD. (2008). Noncholinergic neurons in the basal forebrain: Often neglecte but motivationally salient. Neuron, 59, 6-8. [PDF]

    Belova MA, Paton JJ, Salzman CD. (2008). Moment-to-moment tracking of state value in the amygdala. J Neurosci, 28, 10023-10030. [PDF]

    Belova MA, Paton JJ, Morrison SE, Salzman CD. (2007). Expectation modulates neural responses to pleasant and aversive stimuli in primate amygdala. Neuron, 55, 970-984. [PDF -- includes supplementary material]

    Salzman CD, Paton JJ, Belova MA, Morrison SE. (2007). Flexible neural representations of value in the primate brain. Ann N Y Acad Sci, 1121, 336-354. [PDF]

    Paton JJ, Belova MA, Morrison, SE, Salzman CD. (2006). The primate amygdala represents the positive and negative value of visual stimuli during learning. Nature, 439, 865-870. [PDF -- includes supplementary material]

    Salzman CD, Belova MA, Paton JJ. (2005). Beetles, boxes and brain cells: Neural mechanisms underlying valuation and learning. Curr Opin Neurobiol, 15, 721-729. [PDF]

    Previous work on visual motion processing and perceptual decision-making:

    Salzman CD, Newsome WT. (1994). Neural mechanisms for forming a perceptual decision. Science, 264, 231-237. [PDF]

    Murasugi CM, Salzman CD, Newsome WT. (1993). Microstimulation in visual area MT: Effects of varying pulse amplitude and frequency. J Neurosci, 13, 1719-1729. [PDF]

    Newsome WT, Salzman CD. (1993). The neuronal basis of motion perception. Ciba Found Symp, 174, 217-230.

    Salzman CD, Murasugi CM, Britten KH, Newsome WT. (1992). Microstimulation in visual area MT: Effects on direction discrimination performance. J Neurosci, 12, 2331-2355. [PDF]

    Salzman CD, Britten KH, Newsome WT. (1990). Cortical microstimulation influences perceptual judgments of motion direction. Nature, 346, 174-177. [PDF]

Themes and highlights

The amygdala

The amygdala is part of the limbic system, a network of interconnected brain structures traditionally associated with the processing of emotion. Our understanding of the composition and function of the limbic system has evolved over time, and it is now clear that limbic areas are involved not only in emotion, but also in behavior and memory (LeDoux 2000). Headway into the particular functions of the amygdala was initially made in studies of fear conditioning, one of the simpler examples of associating a stimulus with emotional value. Plasticity in the amygdala is required for such conditioning (Davis 1992). The scope of amygdalar function, however, has yet to be determined. We focus on the amygdala because it appears to be well-situated to play a central role in emotional learning and behavior, given its extensive sensory input, and connections with other brain areas known to be involved in processing value (Cardinal et al 2002; Sah et al 2003).

Cardinal RN, Parkinson JA, Hall J, Everitt BJ. (2002). Emotion and motivation: The role of the amygdala, ventral striatum and prefrontal cortex. Neurosci Biobehav Rev, 26, 321-352.

Davis M. (1992). The role of the amygdala in conditioned fear. In Emotion: Theory, Research and Experience (R Plutchik and H Kellerman, Eds.). Orlando, FL: Academic Press, 255-306.

LeDoux JE. (2000). Emotion circuits in the brain. Ann Rev Neurosci, 23, 155-184.

Sah P, Faber ES, Lopez de Armentia M, Power J. (2003). The amygdaloid complex: Anatomy and physiology. Physiol Rev, 83, 803-834.
Experimental design

Fractal images and contingency reversal are frequently central features of the tasks we use. For visual stimuli, we access an essentially limitless bank of computer-generated fractal images, allowing us to use a different set of images in each experiment which are distinct and recognizable, have similar image statistics, and which won't have idiosyncratic associations for the subject (as images of real-world objects might). In our paradigms, we employ reversal, switching the contingency between the stimulus (the CS) and the outcome (reward or airpuff, the US), so as to differentiate responses related to value from responses related to visual features and image identity.

Paton et al, 2006, Figure 1A (see Publications).

RESULT:   Value signals in the amygdala

Single amydala neurons reverse their responses to visual stimuli when the value of these stimuli are reversed through conditioning. Amygdala neural activity thus reflects image value. Some amygdala neurons respond to positive value, while others respond to negative value. The figure shows a neuron that responds to positive value.

Paton et al, 2006, Figure 3C-F (see Publications).

Cover of Neuron

Montage created by artist Ellen K. Levy. She depicts the amygdala metaphorically in relation to Dante's Spheres of Heaven and Hell, which here symbolize rewards and punishments in a surprising context. Neurons extend from an MRI of the amygdala towards these Spheres, thereby depicting how the amygdala integrates information about surprising reinforcement of both valences.

Cover of Neuron, Volume 55, in association with Belova et al 2007 (see Publications).
RESULT:   Modulation of reinforcement responses in the amygdala by expectation

Amygdala neurons respond to conditioned stimuli, both during stimulus presentation and during delay periods between stimuli and reinforcers. These same neurons also respond to reinforcers. Moreover, reinforcement responses are modulated by expectation, as shown in the figure. Responses to expected reinforcement were measured during a conditioning task, while responses to unexpected reinforcement were measured during random uncued delivery of reinforcers. The nature of the responses to visual stimuli and reinforcers suggests that amygdala neurons may be integrating value-related information during conditioning.

Belova et al, 2007, Figure 2 (see Publications).

Neuroscience seminars at Columbia

Neurobiology and Behavior Seminar
(Department of Neuroscience seminar)
noon Thursday All seminars listed on the Columbia Neuroscience Seminars page

Neurotheory seminars also listed on the Center for Theoretical Neuroscience calendar
Neurotheory Seminar 11:30am Friday
Special presentations:
  • Kavil Lecture on Neuronal Plasticity
  • Colleen Giblin Memorial Lecture
  • Thesis presentations



available through Columbia University Libraries
(off campus access prompts for Columbia login)

Last updated 07/30/12.