Abstract

The amygdala is an almond-shaped cluster of nuclei in the medial temporal lobe, one in each hemisphere, and a hub of the limbic system. It attaches emotional and motivational significance to sensory events and is best known for detecting threat and acquiring conditioned fear. Beyond fear, it modulates the consolidation of emotionally arousing memories, sharpens perception of salient stimuli, and represents stimulus value along a positive-to-negative dimension. Human lesion studies, single-neuron recordings, and optogenetic dissection converge on a structure that is heterogeneous rather than unitary: distinct nuclei and cell populations route information toward opposite behavioral outcomes. This article surveys its anatomy, its fear circuitry, its memory functions, the human lesion evidence, and its valence-coding circuits, with three interactive demonstrations.

Keywords: amygdala, fear conditioning, emotional memory, limbic system, valence

The amygdala (from the Greek for almond) is a subcortical grey-matter structure buried in the anterior medial temporal lobe, comprising roughly a dozen anatomically and functionally distinct nuclei (Šimić et al., 2021). It is a central node of the limbic system, positioned to receive highly processed sensory information and to influence the hypothalamic and brainstem systems that generate autonomic, endocrine, and behavioral responses. Its emotional importance was first revealed when bilateral temporal-lobe removal in monkeys produced tameness, loss of fear, and indiscriminate approach to objects, a pattern later named the Klüver–Bucy syndrome (Klüver & Bucy, 1997). A half-century of animal and human research has since established the amygdala as essential for learning which stimuli predict danger or reward and for tagging significant experiences for durable storage (LeDoux, 2000; Phelps & LeDoux, 2005).

Key Takeaways
  • The amygdala assigns emotional and motivational value to sensory events, not fear alone.
  • It is heterogeneous: the basolateral complex is the input side, the central nucleus the output side.
  • Fear conditioning depends on synaptic plasticity in the amygdala; a fast subcortical route can trigger responses before awareness.
  • The amygdala modulates memory consolidation elsewhere in the brain rather than storing declarative memories itself.
  • Human lesion cases such as patient S.M. show impaired fear recognition and blunted threat responses.

What the Amygdala Is

Each amygdala sits anterior to the hippocampus and the inferior horn of the lateral ventricle, deep within the temporal lobe. It is not a single nucleus but a collection of them, conventionally grouped into a basolateral complex, a centromedial division, and a cortical (corticomedial) group, each with distinct connections and cell types (Šimić et al., 2021). The basolateral complex is the principal sensory gateway, receiving input from thalamic and cortical areas; the central nucleus is the principal output station, projecting to hypothalamic and brainstem effectors that produce freezing, autonomic arousal, and stress-hormone release (Davis, 1992). This input–output architecture is why the amygdala can convert a learned signal of danger into a coordinated bodily response.

Figure 1

Input and Output Organization of the Amygdala

Amygdala input and output schematic Sensory thalamus and sensory cortex send arrows into the basolateral complex, which connects to the central nucleus, which sends arrows out to the hypothalamus and brainstem producing autonomic and behavioral responses. Sensory thalamus Sensory cortex Amygdala Basolateral Central Hypothalamus endocrine Brainstem autonomic
Note. The basolateral complex receives sensory input; the central nucleus drives autonomic and endocrine output. Original schematic.

Types of Amygdala

In the Medical Subject Headings (MeSH) classification the amygdala is filed under the broader structures Limbic System and Basal Ganglia, and it is subdivided into three direct subtypes corresponding to its major nuclear groups. These subtypes are anatomical partitions rather than mutually exclusive functional modules: a single behavior such as conditioned fear recruits more than one of them in sequence. MeSH is an indexing vocabulary used to organize the biomedical literature, so the list below reflects how work on the amygdala is catalogued, not a theoretical claim that these are the only meaningful divisions of the structure.

SubtypeIn brief
Basolateral Nuclear ComplexThe input side, receiving sensory information from thalamus and cortex; the principal site of associative plasticity in fear learning.
Central Amygdaloid NucleusThe output side, projecting to hypothalamus and brainstem to drive autonomic, endocrine, and behavioral fear responses.
Corticomedial Nuclear ComplexReceives olfactory and pheromonal input and is linked to social, defensive, and reproductive behaviors.

Fear and Threat Processing

The amygdala's best-characterized role is in Pavlovian fear conditioning, a form of classical conditioning in which a neutral conditioned stimulus comes to predict an aversive unconditioned stimulus and thereby elicits defensive responses (Maren, 2001). Convergent sensory pathways reach the basolateral complex, where coincident activity strengthens synapses so that the conditioned stimulus alone can later activate the circuit (LeDoux, 2000). Lesion and stimulation work in animals showed that the central nucleus is required for the full expression of learned fear, including the fear-potentiated startle reflex whose amplitude rises in the presence of a danger signal (Davis, 1992).

A recurring theme is speed. Sensory information can reach the amygdala by a fast, coarse subcortical route through the thalamus as well as by a slower cortical route that carries finer detail, so a threat can begin to mobilize the body before it is consciously identified (LeDoux, 2000). Human neuroimaging supports rapid, partly automatic engagement: the amygdala responds more strongly to fearful than to neutral or happy faces (Morris et al., 1996), and it does so even when the emotional face is rendered invisible by backward masking, without the viewer reporting having seen it (Whalen et al., 1998).

Explore

Two Routes to the Amygdala

A threatening image can reach the amygdala two ways: a fast, coarse route through the thalamus that arrives before the image is identified, and a slower cortical route that carries the detail needed for conscious recognition. Change how the stimulus is shown. Under masking the cortical route never yields a seen face, yet the subcortical route still activates the amygdala — the basis of unaware threat responses.

050100150200time since stimulus onset (ms, schematic)low roadthalamo-amygdalahigh roadthalamo-cortico-amygdala
subcortical route (fast, coarse)cortical route (slow, detailed)
With a clearly visible face, the subcortical route reaches the amygdala at about 40 ms, ahead of the cortical route at about 120 ms. The cortical route also delivers a conscious identification, so recognition and the amygdala response proceed together.
An illustrative comparison of the fast subcortical route and the slower cortical route by which a visual threat reaches the amygdala. Latencies are schematic fixed values chosen to show the ordering, not measured timings.

Emotional Memory and Consolidation

Emotionally arousing events are remembered better than neutral ones, and the amygdala is the reason. Rather than storing declarative memories itself, it modulates the consolidation of memories held in other systems, notably the hippocampus and neocortex, through stress hormones and noradrenergic signalling released during arousal (McGaugh, 2004). Blocking this modulation weakens the memory advantage for emotional material, while enhancing it strengthens later recall, which is why the mechanism is described as a consolidation gate rather than a storehouse (Cahill & McGaugh, 1998). This influence over memory consolidation links the amygdala to the durability of long-term memory for significant experiences.

Conditioned fear is itself a learned association whose strength grows with repeated pairing and can later be reduced, though not erased, by extinction. The demonstration below models acquisition with a simple error-correction rule, the delta rule of the Rescorla-Wagner model, in which the associative strength of a cue is updated on each trial by a fraction of the discrepancy between the outcome and the strength already accrued; this is the standard formal account of associative learning throughout conditioning research, and the amygdala's synaptic plasticity is widely read as its neural implementation (Rescorla & Wagner, 1972). Extinction learning that suppresses the conditioned response depends on the prefrontal cortex acting on the amygdala (Quirk & Mueller, 2008).

Model It

Acquisition and Extinction of Conditioned Fear

A conditioned stimulus that predicts an aversive outcome gains associative strength in ever-smaller steps, because each trial closes a fixed fraction of the remaining gap to the ceiling. At the default learning rate of 0.25 the first five trials trace 0.25, 0.44, 0.58, 0.68, 0.76 — the worked example above. Add extinction trials, where the outcome is withheld, and the same rule carries strength back down.

Learning rate α0.25
Acquisition trials (CS paired with aversive US)5
Extinction trials (CS presented alone)6
0.000.250.500.751.00strength Vacquisitionextinction
At a learning rate of 0.25, the 5 acquisition trials give V = 0.25, 0.44, 0.58, 0.68, 0.76, reaching 0.76 of a maximum of 1. The increments shrink as the prediction error falls, tracing the negatively accelerated fear-acquisition curve. Over 6 extinction trials strength falls to 0.14. The model returns to zero, but the animal does not: extinction is new prefrontal-dependent learning, so the fear can recover.
An exact run of the delta rule V ← V + α(λ − V), computed locally in the browser and not stored. The rise is acquisition toward λ = 1; the fall is extinction toward 0. The curve is the model, not measured data.

The Human Amygdala: Lesion Evidence

Rare patients with selective bilateral amygdala damage have been decisive for human research. Patient S.M., whose amygdalae were destroyed by Urbach–Withe disease, is impaired at recognising fear in facial expressions while recognition of other emotions is relatively spared (Adolphs et al., 1994). She also fails to experience fear in situations that reliably frighten others, from handling snakes to visiting reputedly haunted houses, yet inhaling carbon dioxide can still induce panic in her, showing that the amygdala is necessary for externally cued fear but not for every route to the feeling (Feinstein et al., 2011).

The human amygdala also shapes perception and choice. Lesions impair the normal enhancement of perception for emotionally salient stimuli, so that events which would ordinarily capture attention lose their privileged access (Anderson & Phelps, 2001). In a double dissociation with the hippocampus, amygdala damage abolishes the conditioned autonomic response to a stimulus while leaving explicit knowledge of the contingency intact, whereas hippocampal damage produces the reverse, a separation that informs how emotional value guides decision making (Bechara et al., 1995). Its responsiveness to emotional facial expressions also connects it to face perception and to the reading of social signals.

Circuits and Valence

Modern circuit tools have replaced the picture of a single fear centre with one of intermingled, genetically and anatomically distinct populations that route information toward opposite outcomes (Janak & Tye, 2015). Optogenetic control of a basolateral-to-central projection can bidirectionally and reversibly increase or decrease anxiety, demonstrating that specific pathways, not the structure as a whole, carry specific functions (Tye et al., 2011). Within the basolateral complex, separate neurons projecting to different downstream targets preferentially signal positive or negative value, so appetitive and aversive information leave the amygdala by divergent routes (Beyeler et al., 2016). This valence coding is why the amygdala matters for reinforcement of both reward and punishment, and its integration of internal bodily state with external cues connects it to interoception.

Explore

Divergent Routing of Positive and Negative Value

Within the basolateral amygdala, distinct neurons projecting to different targets preferentially signal appetitive or aversive value, so positive and negative information leave by divergent routes rather than from one undifferentiated centre. Slide the experience from aversive to appetitive and watch which population dominates and which downstream target it drives.

Experienceneutral
basolateral amygdalaaversive targetavoidance / defenseappetitive targetapproach / reward
positive-valence populationnegative-valence population
The positive-valence population is driven to 0.50 and the negative-valence population to 0.50. The two channels are balanced, so appetitive and aversive routing are equally engaged.
An illustrative model of valence routing: two basolateral projection populations carry opposite value to different targets, with activity set exactly by the slider. The widths encode drive; they are a schematic of the routing principle, not recorded firing rates.
DomainAmygdala contributionKey evidence
Fear conditioningAcquires and expresses learned defensive responsesMaren (2001); LeDoux (2000)
Emotional memoryModulates consolidation of arousing eventsMcGaugh (2004)
PerceptionEnhances processing of salient stimuliAnderson & Phelps (2001)
Valence and anxietyRoutes positive and negative value by distinct circuitsTye et al. (2011); Beyeler et al. (2016)

Worked Example

Consider how the strength of a conditioned fear response grows across training trials. Let the associative strength of a conditioned stimulus be V, updated after each trial by the delta rule VV + α(λ − V), where λ = 1.0 is the maximum association supported by the aversive outcome and α = 0.25 is the learning rate. Starting from V = 0, the prediction error (λ − V) is largest on the first pairing and shrinks as learning approaches its ceiling.

Trial 1: error = 1.00, so V rises to 0.25. Trial 2: error = 0.75, V = 0.44. Trial 3: error = 0.56, V = 0.58. Trial 4: error = 0.42, V = 0.68. Trial 5: error = 0.32, V = 0.76. The conditioned response therefore climbs steeply at first and then decelerates, tracing the negatively accelerated acquisition curve seen in fear conditioning. If the unconditioned stimulus is then withheld, λ becomes 0 and the same rule drives V back down: extinction follows a mirror-image curve, and because it is new prefrontal-dependent learning rather than unlearning, the original association can recover (Quirk & Mueller, 2008). The second demonstration lets these parameters be varied directly.

Discussion

The amygdala illustrates a general principle of affective neuroscience: emotion is implemented not by a single structure computing a single feeling but by dedicated circuits that link appraisal to action. Its lesions do not abolish emotion wholesale; they selectively disrupt the learning of danger signals, the emotional enhancement of memory and perception, and the recognition of threat in others, while sparing much else (Phelps & LeDoux, 2005). The structure's clinical relevance is broad, because dysregulated amygdala function is implicated in anxiety disorders, post-traumatic stress, and mood disorders, where the same threat-learning and valence-coding mechanisms that are adaptive in ordinary life become maladaptive. Understanding the amygdala as a set of separable pathways, rather than a monolithic alarm, is what makes targeted intervention conceivable.

Current Directions

Contemporary work reframes the amygdala from a fear detector into a general engine for assigning behavioral significance. Single-neuron recordings during naturalistic behavior show that amygdala ensembles encode sustained internal states, such as defensiveness or exploration, rather than only transient stimulus reactions, integrating information across time to bias ongoing behavior (Gründemann et al., 2019). Circuit studies reveal competing inhibitory microcircuits within the central amygdala that arbitrate between active and passive defensive responses, so that the same threat can produce flight or freezing depending on which population prevails (Fadok et al., 2017). Reviews now emphasise the amygdala's multidimensional processing, in which many stimulus features and behavioral variables are represented in overlapping populations (Gothard, 2020). In parallel, a two-system framework distinguishes the nonconscious defensive circuits centred on the amygdala from the cortical networks that generate the conscious feeling of fear, with direct implications for how anxiety is treated (LeDoux & Pine, 2016).

Common Misconceptions

The amygdala is the brain's fear center.
The amygdala processes emotional and motivational significance broadly, including appetitive and rewarding stimuli, not fear alone; distinct populations signal positive as well as negative value (Janak & Tye, 2015). The fear label persists because early lesion and conditioning studies emphasised defensive behavior, which was the easiest response to measure (LeDoux, 2000).
The amygdala is where the feeling of fear is generated.
The amygdala controls defensive responses that can be triggered without awareness, but the conscious feeling of fear is assembled by cortical networks; the two can be dissociated (LeDoux & Pine, 2016). Patient S.M., who lacks a functioning amygdala, still experiences panic when inhaling carbon dioxide, showing that the feeling has routes the amygdala does not gate (Feinstein et al., 2011).
The amygdala is a single structure.
It is a heterogeneous collection of roughly a dozen nuclei with different connections, neurochemistry, and functions, conventionally grouped into basolateral, central, and cortical divisions (Šimić et al., 2021).

Glossary

Amygdala.
An almond-shaped group of nuclei in the medial temporal lobe that assigns emotional and motivational significance to stimuli.
Basolateral complex.
The input region of the amygdala, receiving sensory afferents and forming the associations that underlie fear learning.
Central nucleus.
The main output region of the amygdala, projecting to hypothalamic and brainstem systems that produce emotional responses.
Conditioned response.
A learned reaction elicited by a stimulus that has been paired with a biologically significant event.
Corticomedial complex.
The amygdala division receiving olfactory and pheromonal input, associated with social and reproductive behavior.
Emotional memory.
The enhanced retention of arousing experiences, which the amygdala promotes by modulating consolidation elsewhere.
Extinction.
The reduction of a conditioned response when the conditioned stimulus is repeatedly presented without the outcome; new learning, not erasure.
Fear conditioning.
Pavlovian learning in which a neutral stimulus comes to predict an aversive event and elicits defensive responses.
Fear-potentiated startle.
The increase in the startle reflex when an organism is in the presence of a learned danger signal.
Limbic system.
A set of interconnected subcortical and cortical structures involved in emotion, motivation, and memory, of which the amygdala is a part.
Low road.
The fast, coarse subcortical pathway carrying sensory information from thalamus to amygdala, capable of triggering responses before conscious identification.
Nucleus.
A cluster of neuron cell bodies in the central nervous system that functions as a processing unit.
Rescorla-Wagner model.
The standard formal model of associative learning, in which a cue's associative strength updates by a fraction of the prediction error; the delta rule used to describe fear-conditioning acquisition in the amygdala.
Salience.
The property of a stimulus that makes it stand out and command processing resources; the amygdala boosts the salience of emotional events.
Unconditioned stimulus.
A stimulus that elicits a response without prior learning, such as a shock producing fear.
Valence.
The positive or negative value of a stimulus, encoded in the amygdala by partly separate neuronal populations.

Key Researchers

Ralph Adolphs. Bren Professor of Psychology, Neuroscience, and Biology at the California Institute of Technology; he showed that bilateral amygdala damage impairs the recognition of fear from facial expressions. Faculty Page - ORCID

Michael Davis. Professor Emeritus of Psychiatry and Behavioral Sciences at Emory University; he mapped the amygdala circuitry underlying the fear-potentiated startle reflex. Faculty Page

Joseph E. LeDoux (b. 1949). Professor of Neural Science at New York University; he traced the subcortical fear-conditioning pathway from thalamus to amygdala and proposed a two-system framework separating defensive responses from conscious feelings. Faculty Page - Wikipedia

James L. McGaugh (b. 1931). Distinguished Professor Emeritus at the University of California, Irvine; he demonstrated that the amygdala modulates the consolidation of emotionally arousing memories. Faculty Page - Wikipedia

Elizabeth A. Phelps. Pershing Square Professor of Human Neuroscience at Harvard University; she extended amygdala research to humans, showing its role in emotional perception and memory. Faculty Page - ORCID

Kay M. Tye. Professor at the Salk Institute for Biological Studies; she used optogenetics to dissect amygdala circuits controlling anxiety and to trace divergent routes for positive and negative valence. Faculty Page - ORCID

Frequently Asked Questions

What is the amygdala?
The amygdala is an almond-shaped group of nuclei in the medial temporal lobe that assigns emotional and motivational significance to stimuli and is central to fear and emotional memory (Šimić et al., 2021).

Is the amygdala only involved in fear?
No; it processes stimulus value broadly, with distinct neuronal populations signalling positive and rewarding outcomes as well as aversive ones (Janak & Tye, 2015).

What happens if the amygdala is damaged?
Bilateral damage impairs recognition of fear in faces and blunts learned fear responses, as documented in patient S.M. (Feinstein et al., 2011).

Does the amygdala store memories?
It does not store declarative memories itself; it modulates the consolidation of emotionally arousing memories held in the hippocampus and cortex (McGaugh, 2004).

What is the low road to the amygdala?
It is a fast, coarse subcortical pathway from the thalamus that can activate the amygdala before a stimulus is consciously identified (LeDoux, 2000).

Can the amygdala respond to things people do not consciously see?
Yes; the amygdala responds to fearful faces even when they are masked so that viewers do not report seeing them (Whalen et al., 1998).

How is a conditioned fear reduced?
Through extinction, in which the conditioned stimulus is presented without the aversive outcome; this depends on prefrontal cortex acting on the amygdala and does not erase the original memory (Quirk & Mueller, 2008).

Does the amygdala have different parts?
Yes; it comprises multiple nuclei, with the basolateral complex serving as the sensory input and the central nucleus as the response output (Gothard, 2020).

References

Adolphs, R., Tranel, D., Damasio, H., & Damasio, A. (1994). Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala. Nature, 372(6507), 669-672. https://doi.org/10.1038/372669a0

Anderson, A. K., & Phelps, E. A. (2001). Lesions of the human amygdala impair enhanced perception of emotionally salient events. Nature, 411(6835), 305-309. https://doi.org/10.1038/35077083

Bechara, A., Tranel, D., Damasio, H., Adolphs, R., Rockland, C., & Damasio, A. R. (1995). Double dissociation of conditioning and declarative knowledge relative to the amygdala and hippocampus in humans. Science, 269(5227), 1115-1118. https://doi.org/10.1126/science.7652558

Beyeler, A., Namburi, P., Glober, G. F., Simonnet, C., Calhoon, G. G., Conyers, G. F., Luck, R., Wildes, C. P., & Tye, K. M. (2016). Divergent routing of positive and negative information from the amygdala during memory retrieval. Neuron, 90(2), 348-361. https://doi.org/10.1016/j.neuron.2016.03.004

Cahill, L., & McGaugh, J. L. (1998). Mechanisms of emotional arousal and lasting declarative memory. Trends in Neurosciences, 21(7), 294-299. https://doi.org/10.1016/S0166-2236(97)01214-9

Davis, M. (1992). The role of the amygdala in fear and anxiety. Annual Review of Neuroscience, 15, 353-375. https://doi.org/10.1146/annurev.ne.15.030192.002033

Fadok, J. P., Krabbe, S., Markovic, M., Courtin, J., Xu, C., Massi, L., Botta, P., Bylund, K., Müller, C., Kovacevic, A., Tovote, P., & Lüthi, A. (2017). A competitive inhibitory circuit for selection of active and passive fear responses. Nature, 542(7639), 96-100. https://doi.org/10.1038/nature21047

Feinstein, J. S., Adolphs, R., Damasio, A., & Tranel, D. (2011). The human amygdala and the induction and experience of fear. Current Biology, 21(1), 34-38. https://doi.org/10.1016/j.cub.2010.11.042

Gothard, K. M. (2020). Multidimensional processing in the amygdala. Nature Reviews Neuroscience, 21(10), 565-575. https://doi.org/10.1038/s41583-020-0350-y

Gründemann, J., Bitterman, Y., Lu, T., Krabbe, S., Grewe, B. F., Schnitzer, M. J., & Lüthi, A. (2019). Amygdala ensembles encode behavioral states. Science, 364(6437), eaav8736. https://doi.org/10.1126/science.aav8736

Janak, P. H., & Tye, K. M. (2015). From circuits to behaviour in the amygdala. Nature, 517(7534), 284-292. https://doi.org/10.1038/nature14188

Klüver, H., & Bucy, P. C. (1997). Preliminary analysis of functions of the temporal lobes in monkeys. Journal of Neuropsychiatry and Clinical Neurosciences, 9(4), 606-620. (Original work published 1939) https://doi.org/10.1176/jnp.9.4.606

LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155-184. https://doi.org/10.1146/annurev.neuro.23.1.155

LeDoux, J. E., & Pine, D. S. (2016). Using neuroscience to help understand fear and anxiety: A two-system framework. American Journal of Psychiatry, 173(11), 1083-1093. https://doi.org/10.1176/appi.ajp.2016.16030353

Maren, S. (2001). Neurobiology of Pavlovian fear conditioning. Annual Review of Neuroscience, 24, 897-931. https://doi.org/10.1146/annurev.neuro.24.1.897

McGaugh, J. L. (2004). The amygdala modulates the consolidation of memories of emotionally arousing experiences. Annual Review of Neuroscience, 27, 1-28. https://doi.org/10.1146/annurev.neuro.27.070203.144157

Morris, J. S., Frith, C. D., Perrett, D. I., Rowland, D., Young, A. W., Calder, A. J., & Dolan, R. J. (1996). A differential neural response in the human amygdala to fearful and happy facial expressions. Nature, 383(6603), 812-815. https://doi.org/10.1038/383812a0

Phelps, E. A., & LeDoux, J. E. (2005). Contributions of the amygdala to emotion processing: From animal models to human behavior. Neuron, 48(2), 175-187. https://doi.org/10.1016/j.neuron.2005.09.025

Quirk, G. J., & Mueller, D. (2008). Neural mechanisms of extinction learning and retrieval. Neuropsychopharmacology, 33(1), 56-72. https://doi.org/10.1038/sj.npp.1301555

Rescorla, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In A. H. Black & W. F. Prokasy (Eds.), Classical conditioning II: Current research and theory (pp. 64-99). Appleton-Century-Crofts.

Šimić, G., Tkalčić, M., Vukić, V., Mulc, D., Španić, E., Šagud, M., Olucha-Bordonau, F. E., Vukšić, M., & Hof, P. R. (2021). Understanding emotions: Origins and roles of the amygdala. Biomolecules, 11(6), 823. https://doi.org/10.3390/biom11060823

Tye, K. M., Prakash, R., Kim, S.-Y., Fenno, L. E., Grosenick, L., Zarabi, H., Thompson, K. R., Gradinaru, V., Ramakrishnan, C., & Deisseroth, K. (2011). Amygdala circuitry mediating reversible and bidirectional control of anxiety. Nature, 471(7338), 358-362. https://doi.org/10.1038/nature09820

Whalen, P. J., Rauch, S. L., Etcoff, N. L., McInerney, S. C., Lee, M. B., & Jenike, M. A. (1998). Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge. Journal of Neuroscience, 18(1), 411-418. https://doi.org/10.1523/JNEUROSCI.18-01-00411.1998