CoChat
Sign Up
Cited by 422Action Observation and SynchronizationFace Recognition and PerceptionChild and Animal Learning Development

The neural bases of empathic accuracy

Open access
Zaki, Jamil, Weber, Jochen, Bolger, Niall, Ochsner, Kevin·2009·Proceedings of the National Academy of Sciences
DOI 10.1073/pnas.0902666106PMC PMC2708723
Open this paper in CoChat — free account

Abstract

Theories of empathy suggest that an accurate understanding of another's emotions should depend on affective, motor, and/or higher cognitive brain regions, but until recently no experimental method has been available to directly test these possibilities. Here, we present a functional imaging paradigm that allowed us to address this issue. We found that empathically accurate, as compared with inaccurate, judgments depended on ( i ) structures within the human mirror neuron system thought to be involved in shared sensorimotor representations, and ( ii ) regions implicated in mental state attribution, the superior temporal sulcus and medial prefrontal cortex. These data demostrate that activity in these 2 sets of brain regions tracks with the accuracy of attributions made about another's internal emotional state. Taken together, these results provide both an experimental approach and theoretical insights for studying empathy and its dysfunction.

AI snapshot

Written from the paper’s full text.

TL;DR

This fMRI study adapted an empathic accuracy (EA) paradigm to test whether accurate judgments of others’ emotions depend on shared sensorimotor representations and/or mental state attribution systems. Accuracy was predicted by activity in mirror-neuron-system regions (right IPL, bilateral dorsal PMC) and mental-state-attribution regions (dorsal/rostral MPFC, STS), but not by ACC, AI, SII, or TPJ.

Key findings

  • Perceivers were moderately accurate at inferring target affect, well above chance (mean *r* = 0.46, *t* = 9.72, *P* < 0.001).
  • Whole-brain parametric analysis found that EA was predicted by activity in MSA-related regions—dorsal and rostral MPFC and STS—and in sensorimotor/mirror-neuron-system regions—right IPL and bilateral dorsal PMC.
  • ROI analyses converged: activity in MPFC, IPL, and PMC tracked with EA, whereas activity in ACC, AI, SII, and TPJ did not.
  • Across participants, perceivers who engaged MSA-related regions for accurate inferences also tended to engage SR-related regions (reported in SI Text and Fig. S1).

Methods

  • Two-phase design: 14 targets discussed emotional autobiographical events and later continuously rated their own affect while rewatching; 11 targets’ videos were included after exclusions. In the perceiver phase, 21 perceivers were scanned while watching 6 target videos and continuously rating how positive or negative they believed targets felt; 16 perceivers remained in imaging analyses after exclusions.
  • EA was computed as the time-course correlation between target affect ratings and perceiver ratings of target affect, after z-transforming ratings and averaging into 2-s periods.
  • fMRI data were analyzed with SPM2 using a general linear model; EA scores served as parametric modulators, controlling for number of ratings per minute. Whole-brain maps were thresholded at *P* < 0.005, uncorrected, with 20 contiguous voxels, corresponding to whole-brain false-positive discovery rate < 5%; ROI analyses used peaks from prior SR and MSA studies.

Limitations

  • The paper states that EA is a consensus measure based on the match between two subjective ratings; if targets’ introspective ratings do not match their expressed cues (e.g., in alexithymia), a perceiver could correctly read cues yet appear empathically inaccurate.
  • The paper notes that group-level concurrent activity could obscure individual differences in whether perceivers rely more on SR- or MSA-related systems.
  • The paper states that future work should use both naturalistic and constrained social tasks in the same participants; it also notes that the absence of accuracy-related AI, ACC, SII, and TPJ activity may reflect the specific complex emotional states studied rather than a general absence.
  • Evident from the text: the imaging sample was small (16 perceivers, 11 targets), each perceiver rated only 6 videos, and the design is correlational, so it cannot establish causal roles for the identified regions.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1073/pnas.0902666106

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 25 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § Results

    Activity in ACC, AI, SII, and TPJ did not track with EA.

    “Data extracted from these regions of interest (ROIs) were consistent with the results of the whole-brain analysis: Activity in MPFC, IPL, and PMC tracked with EA, but activity in the ACC, AI, SII, and TPJ did not (Table S3).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.
  • Findingown data
    § Discussion

    Perceivers who engaged MSA-related regions to make accurate inferences also tended to engage SR-related regions.

    “To examine this possibility, we correlated mean activity in each type of system across participants and found that perceivers who engaged MSA-related regions to make accurate inferences also tended to engage SR-related regions (described in SI Text and Fig. S1).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.
  • Findingown data
    § Results

    EA was predicted by activity in sensorimotor regions in the mirror neuron system thought to support SRs: right inferior parietal lobule (IPL) and bilateral dorsal premotor cortex (PMC).

    “and activity in sensorimotor regions in the mirror neuron system thought to support SRs [right inferior parietal lobule (IPL) and bilateral dorsal premotor cortex (PMC) (5, 6)].”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.
  • Findingown data
    § Discussion

    Activity in the TPJ was unrelated to accuracy when using both kinds of analyses in the current study.

    “It was similarly notable that activity in the TPJ, which has been implicated in MSA, also was unrelated to accuracy when using both kinds of analyses in the current study.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.
  • Findingown data
    § Results

    EA was predicted by activity in 3 regions implicated in MSA: dorsal and rostral subregions of medial prefrontal cortex (MPFC) and the superior temporal sulcus (STS).

    “Results revealed that EA was predicted by activity in 3 regions implicated in MSA [dorsal and rostral subregions of medial prefrontal cortex (MPFC) and the superior temporal sulcus (STS; see Fig. 2 and Table S1; for inaccuracy-related activations, see SI text and Table S2) ]”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.
  • May combine two claims. 10
  • Classified as probably not a substantial claim. 12

  • May combine two claims.

    Findingnone given
    § Discussion

    The positive and negative affective states discussed by social targets in the current study were complex and naturalistic and, by and large, did not include direct or implied displays of pain or disgust.

    “The positive and negative affective states discussed by social targets in the current study were complex and naturalistic and, by and large, did not include direct or implied displays of pain or disgust.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    Perceivers were moderately accurate at inferring target affect, and their accuracy was well above chance (mean r = 0.46, t = 9.72, P &lt; 0.001).

    “Overall, perceivers were moderately accurate at inferring target affect, and their accuracy was well above chance (mean r = 0.46, t = 9.72, P &lt; 0.001), replicating previous behavioral results using this task (25, 26).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § The neural bases of empathic accuracy

    The design allowed for a direct examination of the neural sources of accurate interpersonal understanding.

    “This design allowed for a direct examination of the neural sources of accurate interpersonal understanding, which we explored by using 2 analytic techniques.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Discussion

    Accurate, as opposed to inaccurate, judgments of target affect relied on regions in premotor and inferior parietal cortex spatially similar to those found in studies of motor imitation, and medial prefrontal structures similar to those found in previous studies of MSA.

    “We found that accurate, as opposed to inaccurate, judgments of target affect relied on regions in premotor and inferior parietal cortex spatially similar to those found in studies of motor imitation, and medial prefrontal structures similar to those found in previous studies of MSA.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning
    § Discussion

    Naturalistic tasks like the one used here have the potential to show how multiple networks work simultaneously in supporting accurate social inferences.

    “Whereas simplified social cognitive paradigms may isolate individual networks, and therefore shed light on only 1 piece of a larger processing puzzle, naturalistic tasks like the one used here have the potential to show how multiple networks work simultaneously in supporting accurate social inferences.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.

  • May combine two claims.

    Interpretationown data
    § Results

    Brain regions involved in monitoring actions and motor intentions, as well as regions involved in MSA, support accurate judgments about social targets' affective states.

    “These findings provide converging support for the conclusion that brain regions involved in monitoring actions and motor intentions, as well as regions involved in MSA, support accurate judgments about social targets' affective states.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.

  • Classified as probably not a substantial claim.

    Premisereasoning
    § The neural bases of empathic accuracy

    Whether SRs, MSAs, or other neural systems underlie accurate understanding of target states remains unknown because extant methods cannot address this question.

    “However, whether these or other neural systems underlie accurate understanding of target states remains unknown, because extant methods are unable to address this question.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation
    § The neural bases of empathic accuracy

    Both shared representations (SRs) and mental state attributions (MSAs) are involved in processing information about targets' internal states.

    “These 2 lines of research converge to suggest that both SRs and MSAs are involved in processing information about targets' internal states (14, 19, 20).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [14] source

      Not checked in the free run.

    • cites [19] Decety J, Jackson PL (2004) The functional architecture of human empathy. Behav Cognit Neurosci Rev 3:71-100. source

      Not checked in the free run.

    • cites [20] Olsson A, Ochsner KN (2008) The role of social cognition in emotion. Trends Cognit Sci 12:65-71. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § Discussion

    The TPJ is engaged when assessing false beliefs, detecting deception, and more generally when shifting attention away from invalid cues.

    “The TPJ is engaged when assessing false beliefs (14), detecting deception (32), and more generally when shifting attention away from invalid cues (34).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [14] source

      Not checked in the free run.

    • cites [32] Saarela MV, et al. (2007) The compassionate brain: Humans detect intensity of pain from another's face. Cereb Cortex 17:230-237. source

      Not checked in the free run.

    • cites [34] Mitchell JP (2008) Activity in right temporo-parietal junction is not selective for theory of mind. Cereb Cortex 18:262-271. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data
    § Methods

    Data from 3 target participants were excluded (2 refused video use, 1 showed insufficient rating variability), leaving 11 targets in the sample.

    “Twotarget participants refused to have their videos used in the subsequent EA paradigm, and a third showed insufficient variability in the ratings to allow meaningful ratings of EA; data from these 3 participants were excluded, leaving 11 targets in our sample (6 female, mean age = 25.4).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § Methods

    A regressor of no interest corresponding to the amount of affect ratings perceivers made per minute during each video was included to control for the possibility that increased number of ratings, rather than accuracy per se, drove brain activity during accurate blocks.

    “We also included a regressor of no interest corresponding to the amount of affect ratings perceivers had made per minute during each video; this allowed us to control for the possibility that an increased number of ratings, and not accuracy per se , would be driving brain activity during accurate blocks.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Targets' continuous ratings of their affect while discussing events accurately reflect the affect they experienced while discussing those events.
    • Rests on an assumption: Perceivers' continuous ratings accurately reflect their beliefs about the target's affect at each moment.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationnone given
    § Discussion

    EA and subsequent memory paradigms share a broad conceptual similarity: both provide meaningful performance measures that can help constrain theory.

    “That said, EA and subsequent memory paradigms share a broad conceptual similarity: both provide meaningful performance measures that can help constrain theory.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § Discussion

    Interventions teaching people with ASD to recognize emotions from simplified stimuli often improve performance on simplified tasks but not in real-world social interactions.

    “Further, performance on simplified social cognitive tasks does not always track with the severity of social deficits in ASD (37, 38), and interventions teaching people with ASD to recognize emotions from simplified stimuli often improve performance on simplified tasks but not in real-world social interactions (39, 40).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [39] Golan O, Baron-Cohen S (2006) Systemizing empathy: Teaching adults with Asperger syndromeorhigh-functioningautismtorecognizecomplexemotionsusinginteractive multimedia. Dev Psychopathol 18:591-617. source

      Not checked in the free run.

    • cites [40] Hadwin J, Baron-Cohen S, Howlin P, Hill K (1997) Does teaching theory of mind have an effect on the ability to develop conversation in children with autism? J Autism Dev Disord 27:519-537. source

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductcitation
    § Methods

    Activation maps were thresholded at P < 0.005, uncorrected, with an extent threshold of 20 contiguous voxels, corresponding to a false-positive discovery rate of < 5% across the whole brain as estimated by Monte Carlo simulation implemented using AlphaSim in AFNI.

    “Resulting activation maps were thresholded at P &lt; 0.005, uncorrected, with an extent threshold of 20 contiguous voxels, corresponding to a false-positive discovery rate of &lt; 5% across the whole brain as estimated by Monte Carlo simulation implemented using AlphaSim in AFNI (48).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [48] Cox RW (1996) AFNI: Software for analysis and visualization of functional magnetic resonance neuroimages. Comput Biomed Res 29:162-173. source

      Not checked in the free run.

    • Rests on an assumption: Targets' continuous ratings of their affect while discussing events accurately reflect the affect they experienced while discussing those events.
    • Rests on an assumption: Perceivers' continuous ratings accurately reflect their beliefs about the target's affect at each moment.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning
    § Discussion

    A perceiver could correctly assess the cues a target expresses but appear empathically inaccurate because the perceiver's ratings of the target's experience will not match the target's own ratings.

    “A perceiver could then correctly assess the cues the target expresses but appear to be empathically inaccurate because their ratings of the target's experience will not match the target's own ratings.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Targets' continuous ratings of their affect while discussing events accurately reflect the affect they experienced while discussing those events.
    • Rests on an assumption: Perceivers' continuous ratings accurately reflect their beliefs about the target's affect at each moment.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § Methods

    Data from 5 perceivers were removed (2 for image artifacts, 3 for large motion artifacts), leaving 16 subjects (9 female) in the imaging analysis.

    “Data from 2 subjects were removed because of image artifacts induced by intensity spikes and signal dropout. Additionally, because the blocks in our study were unusually long, our data were especially sensitive to motion artifacts, and data from 3 subjects were removed because of large motion artifacts, leaving a total of 16 subjects (9 female) in our analysis.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § Discussion

    Regions AI, ACC, and SII have been implicated in the empathic sharing of affective and sensory states, with research focusing on sharing of pain, disgust, touch, and posed emotional expressions.

    “The research supporting the role of these regions, however, has focused on the sharing of pain, disgust, touch, and posed emotional expressions (2, 3, 9, 11), all of which may depend on representations of somatovisceral states coded in these regions (31).”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [2] source

      Not checked in the free run.

    • cites [3] source

      Not checked in the free run.

    • cites [9] source

      Not checked in the free run.

    • cites [11] source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § Discussion

    Recent theories of social cognition suggest that cortical areas related to MSA and mirror neuron system structures related to SRs should both be engaged when perceivers make inferences about target states.

    “Recent theories of social cognition recognize this likelihood and suggest that cortical areas related to MSA and mirror neuron system structures related to SRs should both be engaged when perceivers make inferences about target states (20, 28, 29).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [20] Olsson A, Ochsner KN (2008) The role of social cognition in emotion. Trends Cognit Sci 12:65-71. source

      Not checked in the free run.

    • cites [28] UddinLQ,IacoboniM,LangeC,KeenanJP(2007)Theselfandsocialcognition:Therole of cortical midline structures and mirror neurons. Trends Cognit Sci 11:153-157. source

      Not checked in the free run.

    • cites [29] Keysers C, Gazzola V (2006) Towarda unifying neural theory of social cognition. Prog Brain Res 156:379-401. source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § Methods

    Target and perceiver rating data were z-transformed across the entire session to correct for interindividual variation in use of the rating scale.

    “Target and perceiver rating data were z -transformed across the entire session to correct for interindividual variation in use of the rating scale.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Targets' continuous ratings of their affect while discussing events accurately reflect the affect they experienced while discussing those events.
    • Rests on an assumption: Perceivers' continuous ratings accurately reflect their beliefs about the target's affect at each moment.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Discussion

    In some cases, such as for alexithymic targets, experience and expression of emotion could fail to match because these targets have limited ability to introspectively rate their emotions.

    “In some cases, such as for alexithymic targets(42), experience and expression of emotion could fail to match because these targets have limited ability to introspectively rate their emotions.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [42] Sifneos PE (1996) Alexithymia: Past and present. Am J Psychiatry 153(Suppl):137-142. source

      Not checked in the free run.

20 of 20

  • The paper relies on this without support.

    Target self-reports of their moment-to-moment affect accurately reflect what they actually felt.

    “the match between a perceiver's judgment of what a target was feeling and the report that target provided of what he/she was actually feeling.”

  • The paper relies on this without support.

    Targets' continuous ratings of their affect while discussing events accurately reflect the affect they experienced while discussing those events.

    “Again, it was emphasized that targets should concentrate on the affect they had felt while discussing events and not during the events themselves.”
    “Targets' affect ratings were then correlated with perceivers' affect ratings of targets; resulting coefficients are referred to as EA for that perceiver/clip combination (see Fig. 1 B ).”

  • The paper relies on this without support.

    Perceivers' continuous ratings accurately reflect their beliefs about the target's affect at each moment.

    “Perceivers were instructed that if the cue ''OTHER'' appeared before a video, they would use 2 buttons to continuously rate how positive or negative they believed the target felt at each moment, using the same 9-point scale targets themselves had used (anchored at very negative on the left and very positive on the right).”
3 of 3

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The two networks associated with empathy and Theory of Mind are jointly activated for complex evaluations of someone else's feelings such as empathic accuracy.

    “The two networks associated with empathy and ToM are also jointly activated for complex evaluations of someone else's feelings such as empathic accuracy, the ability to infer what someone else is feeling [38].”

    From On the interaction of social affect and cognition: empathy, compassion and theory of mind (2018) · cites this paper as [38]

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. S Baron-Cohen Mindblindness (MIT Press, Cambridge, MA, 1994).

Cited by

Sorted by citations

Related

Authors
Zaki, J., Weber, J., Bolger, N., & Ochsner, K.
Affiliations
University of Oregon (US)
ORCID
Venue
Proceedings of the National Academy of Sciences
Publisher
National Academy of Sciences
Published
2009-07-07
Volume
106
Issue
27
Pages
11382-11387
ISSN
  • 0027-8424 (ISSN-L)
  • 1091-6490
DOI
10.1073/pnas.0902666106
PMC
PMC2708723
OpenAlex
W2169547194
Semantic Scholar
14568bc5ab6d255efbe379abd2f0efa7fb1a3ed1
Cited by
422
References
49
Citations by year
2026: 19 · 2025: 19 · 2024: 24 · 2023: 28 · 2022: 23 · 2021: 33 · 2020: 28 · 2019: 33 · 2018: 22 · 2017: 17 · 2016: 23 · 2015: 22 · 2014: 20 · 2013: 26 · 2012: 35
FWCI
36.42
Type
journal article
Language
en
Open access
Locations
  • Proceedings of the National Academy of Sciences (publishedVersion)
  • PubMed Central (publishedVersion)
  • PubMed (publishedVersion)
Funders
  • Autism Speaks
  • National Institute on Drug Abuse (R01DA022541)
  • Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01HD069178)
Topics
  • Social Sciences › Psychology › Social Psychology › Action Observation and Synchronization
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Face Recognition and Perception
  • Social Sciences › Psychology › Developmental and Educational Psychology › Child and Animal Learning Development
Keywords
empathic accuracy, empathy, mirror neuron system, superior temporal sulcus, medial prefrontal cortex, mental state attribution, emotional understanding, neural basis of empathy
MeSH terms
Adult, Behavior, Empathy, Female, Humans, Male, Nervous System Physiological Phenomena, Task Performance and Analysis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-10.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-10.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review