Differential neural decoding of alarm and avoidance from vocal alarm calls in humans

• Differential neural decoding of alarm and avoidance information from vocal alarm calls has been demonstrated in humans.¹
• Human neural decoding of vocal meaning (alarm/avoidance) supports speech and communication decoding research.¹
• Nature-published; human neuroscience with indirect BCI relevance.^{1 1}

Weekly enrichment (2026-07-20)

• Published in Communications Biology (2025), the study used fMRI to map how the human brain decodes “alarm” versus “avoid-approach” information from scream calls, testing the idea that human screams vary in alarm level rather than always triggering full threat responses.^{2 3}
• Brain activity was recorded from 35 participants (21 female, 14 male; mean age 27.2 years, SD 6.22) after 3 of 38 initial volunteers were excluded for excessive movement or too many missing responses.^{3 4}
• Participants completed two binary forced-choice tasks on each scream: an ALARM task (alarm vs. non-alarm) and an approach-avoidance (AA) task (avoid vs. approach), with two blocks per task.^{3 4}
• Stimuli were 84 validated 800 ms scream calls covering six scream types (pleasure, joy, sadness, pain, fear, anger) plus a neutral vowel /a/ category, produced by 3 male and 3 female speakers and selected from a larger 420-scream set.^{3 4}
• Alarm and avoid-approach decisions showed minimal neural overlap: alarm decisions recruited sensory-affective cascades for action preparation, while avoid-approach judgments engaged spatio-affective systems for social decision-making.^{2 3}
• The authors derived an “alarm quotient” and an “avoid quotient” linking the likelihood of an urgent response to slowed selection of the safe option, quantifying risk arbitration in choice behavior.^{2 3}
• The avoid quotient positively predicted activity across a broad cortico-limbic (auditory-frontal) network, whereas the alarm quotient predicted neural suppression with increasing alarm levels, notably reduced amygdala activity.^{2 3}
• This points to the amygdala being involved in choice arbitration rather than in evaluating the threat signaled by the scream itself, refining models of limbic contributions to vocal threat processing.^{2 3}

Footnotes

1. https://news.google.com/rss/articles/CBMiX0FVX3lxTFBJTXRXbFV4LXJ3cG9vbGF6TzRwVU02UVVNME9kRTZHTTlwX1o3UmpuXzFudXRTbllEZU5mRDV6SW44T3RNODlJQ1pqWHpCWnowQ0hBNzNObWFCOE1KME9Z?oc=5 ↩ ↩² ↩³ ↩⁴

2. https://www.nature.com/articles/s42003-025-08248-9 ↩ ↩² ↩³ ↩⁴ ↩⁵

3. https://pmc.ncbi.nlm.nih.gov/articles/PMC12117006/ ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

4. https://doi.org/10.1038/s42003-025-08248-9 ↩ ↩² ↩³