Sunday, November 30, 2014

Amygdala damage impairs emotion recognition from music

Study by Nathalie Gosselin, Isabelle Peretz, Erica Johnsen, Ralph Adolphs

The role of the amygdala in recognizing danger is well established for visual stimuli such as the face. Gosselin et al. have decided to study a rare subject (by the name of “S.M.”) who has complete bilateral damage to the amygdala that does not encompass other sectors of the temporal lobe. S.M. is severely impaired in recognizing happiness and other emotions in faces (Adolphes et al., 1994, 1995). Her deficit appears to be limited to facial expressions as she constantly fails to recognize fear, surprise, sadness, and anger. And yet, S.M. has no difficulty in recognizing these same emotions in emotional speech. She also exhibits impaired arousal judgments for scary faces. She finds fearful faces less arousing than normal controls whereas she judges them normally as unpleasant. S.M. has been studied by Adolphs and collaborators for the past decade (as mentioned in this 2007-published paper), as she remains to date, the subject with the most selective and complete amygdala atrophy. With that said, the team aims to assess this time, the specific role of the amygdala in her recognition of fear from music.

Gosselin et al. conducts two experiments in this study. The first one, to rate the intensity of fear, peacefulness, happiness, and sadness from computer-generated instrumental music purposely created to express these emotions. Participants (S.M. along with four “normal” female participants) are to listen to fifty-six musical excerpts containing the previously mentioned emotions. The task is to judge the intensity of each emotion for each musical stimulus on 10-point scales. The participants also rated the arousal and valence of each musical stimulus on another 10-point scale of whether the music sounds pleasant or not, for the duration of 45-minutes. In addition, an error detection task that assesses basic auditory perceptual function was tested, through twenty-four of the fifty-six stimuli used in the emotion task. The twenty-four excerpts are modified to contain a timing error, where participants are to indicate if the pianist (playing in the recording) lost track of what s/he is playing at some point of the piece.

The second experiment aims to determine the musical cues that S.M. (and seven normal female controls this time,) are able to use to recognize emotion in music, through the use of mode and tempo. The stimuli consist of thirty-two musical excerpts; half are chosen to evoke happiness and the other half, sadness. In the “mode condition”, all excerpts are transcribed in the opposite mode (major to minor and vice versa). In the “mode-tempo condition”, the mode and tempi are manipulated respectively.

S.M. performs normally in the task of error detection in basic auditory perceptual function. She has difficulty in recognizing scary and sad music, though is able to recognize happy music. S.M. confounds scary with peaceful music, as the confusion never occurs in normals. She also judges scary music to be less arousing and the peaceful music less relaxing than the other controls. S.M. also confounds sadness with peacefulness and occasionally mistakes sad stimuli as happy. And yet, she judges the happy music more intense than normal controls. S.M. judgment of pleasantness are also influenced by tempo. She judges faster musical tempi as more pleasant whereas normal controls’ judgment are unaffected by tempo. Gosselin et. al. states that S.M.’s impairment is similar to that previously reported in patients with unilateral anteromedial temporal lobe damage.

In the mode and tempo experiment, S.M.’s responses are similar to those of normal subjects. Therefore, she shows normal ability to employ tempo and mode as cues for emotional interpretation.

Music is an appropriate medium to assess emotional processing because music is a powerful emotional trigger and is easy to manipulate for research purposes (Gosselin et al., 2007). Gosselin et al. concludes that the amygdala appears to be necessary for emotional processing of music rather than the perceptual processing itself. They hope to pay special attention to the type of musical cues that convey potential threat to the listener and assess their effects on behavioural or brain responses in future studies.

This is a very fascinating research! Researchers have studied this patient for about two decades (assuming they are continuing their studies with her today, in 2014), and have been monitoring the function of her amygdala (and presumably the other parts of her brain). The number of participants is extremely small, though with much preparation made in order to conduct the study. For example, they hired a “professional composer” (name unmentioned) to compose fifty-six musical excerpts. I do wonder though, how are the mood of the musical excerpts determined? Is mood determined by intervals, dynamics, or texture? Mood is apparently determined by metronome marking (ie. Happy excerpts are written in major mode at the average tempo of 137 metronome marking with the melodic line lying in the medium-high range). But how was the 137 metronome marking determined as “happy”? The paper said that the listening excerpts for each emotion category can be heard on the website at, but the examples don’t exist currently, unfortunately. It would have been interesting to listen to the excerpts.

The first experiment lasted for 45-minutes while the second experiment lasted for 1-hour. That’s a long time for an experiment (especially to listen to fifty-six excerpts). I wonder if the participants eventually ended up tired and/or confused with the amount of music to be heard. Did they take breaks? How was their focus level? Also, the experiments measured pleasant and unpleasant music. I wonder how that is measured.

In the two experiments, S.M. is compared by the other participants’ results. What if the normal participants themselves, had drastically different results themselves? Then how would it be graphed? Lastly, knowing that S.M.’s bilateral damage to the amygdala is extremely rare, I wonder if other subjects with the same damage as her, have been tested. I would be interested to also see their results, and compare it with S.M.’s. This research has proved to me all the more that, the brain is absolutely complex -- even the amygdala on its own. Discovery of this organ of ours, is proven to be all the more important in research to discover the nuances of the role of each part of the brain.


Adolphs, R., D. Tranel, H. Damasio, and A. Damasio. "Impaired Recognition Of Emotion In Facial Expressions Following Bilateral Damage To The Human Amygdala." Nature 374.6507 (1994): 669-72.

Gosselin, Nathalie, Isabelle Peretz, Erica Johnsen, and Ralph Adolphs. "Amygdala Damage Impairs Emotion Recognition From Music." Neuropsychologia 45 (2007): 236-44. 

No comments: