Saturday, November 12, 2011

The structural neuroanatomy of music emotion recognition: Evidence from frontotemporal lobar degeneration


Rohani, Omar, Susie M.D. Henley, Jonathan W. Bartlett, Julia C. Hailstone, Elizabeth Gordon, Disa A. Sauter, Chris Frost, Sophie K. Scott, and Jason D. Warren. "The structural neuroanatomy of music emotion recognition: Evidence from frontotemporal lobar degeneration." Neuroimage 2011, June 1; 56(3): 1814-1821.


Despite growing interest in the neurobiology of music, the brain mechanisms that are critical for processing emotion in music remain incompletely understood. Music is universal and highly valued for the powerful emotional responses it engenders: indeed, music activates brain circuitry associated with pleasure and reward and musical emotion judgments and brain responses are consistent amongst members of a musical culture. Certain music can specifically induce an intense arousal response in normal listeners, and this response is mediated by brain structures such as the amygdala and insula that have been implicated in other kinds of salient emotional stimuli. Deficits of musical emotion comprehension have been reported following focal damage of these same structures, located in the medial prefrontal and anterior temporal lobe.

Frontotemporal lobar degeneration (FTLD) is the name for a group of clinically, pathologically and genetically heterogeneous disorders associated with atrophy in the frontal lobe and temporal lobe of the brain. In the over 65 age group, FTLD is probably the fourth most common cause of dementia after Alzheimer’s disease, dementia with Lewy bodies and vascular dementia. Patients with FTLD frequently exhibit derangements of complex social and emotional behaviour. From a clinical perspective, investigation of musical emotion processing and its cerebral associations in FTLD has the potential to improve the understanding of the disease’s phenomenology, and the intrinsic network connectivity in the working brain.

A human brain showing frontotemporal lobar degeneration.

The idea behind this research was to investigate critical neuroanatomical associations of emotion recognition from music using FTLD as a disease model of brain network breakdown. The research included 26 patients with FTLD and 21 healthy control subjects with no history of neurological, or psychiatric illness. Recognition of four emotions (happiness, sadness, anger, and fear) from music, facial expressions and nonverbal vocal sounds was assessed using a procedure in which subjects were required to match each target stimulus with the most appropriate verbal emotion label in a four-alternative-forced-choice model. The music stimuli were short (approx. 11 s) non-vocal (orchestral and chamber) excerpts drawn from the Western classical canon and film scores. MR brain images were acquired in all FTLD patients at the time of behavioural testing, as well as voxel-based morphometry, a neuroimaging analysis technique that allows investigation of focal differences in brain anatomy.

On neuropsychological evaluation, patients with FTLD showed deficient recognition of canonical emotions (happiness, sadness, anger and fear) from music as well as emotional signals conveyed by facial and vocal expressions compared with healthy control subjects. Impaired recognition of emotions from music was specifically associated with grey matter loss in a distributed cerebral network including insula, orbitofrontal cortex, anterior cingulate and medial prefrontal cortex, anterior temporal and more posterior temporal and parietal cortices, amygdala and the subcortical mesolimbic system. This network of the brain is essential for recognition of musical emotion that overlaps with brain regions previously implicated in coding emotional value, behavioural context, conceptual knowledge and theory of mind. The study also found that amygdala damage was associated with impaired emotion recognition only from music, as opposed to emotion recognition of facial and verbal expressions.


The ability that music has to affect and manipulate emotions and the brain is undeniable, and yet largely inexplicable. This research identified regions of the brain associated with music emotion recognition, including insula, orbitofrontal cortex, anterior cingulate and medial prefrontal cortex, anterior temporal and more posterior temporal and parietal cortices, amygdala, and striatum. Identifying the neural mechanisms of musical emotion helps us understand how the brain codes emotional value, and how emotional signals acquire meaning.

Following a similar idea, Petr Janata, associate professor of psychology at UC Davis' Center for Mind and Brain, mapped the brain activity of a group of subjects while they listened to music, and found that the region of the brain where memories of our past are supported and retrieved also serves as a hub that links familiar music, memories and emotion. His research may help to explain why music can elicit strong responses from people with Alzheimer's disease. The hub is located in the medial prefrontal cortex region — right behind the forehead — and one of the last areas of the brain to atrophy over the course of the disease.

In Rohani & al.’s study, subjects with frontotemporal lobar degeneration did not respond well to recognition of emotion in music, unlike Alzheimer’s patients in Janata’s study. This was caused by grey matter loss, including the medial prefrontal cortex region, which is linked to memories and emotion. Does memory affect music emotion recognition, or is it just contained in the same medial prefrontal cortex region as is emotion? How does music succeed in prompting emotions within us? And why are these emotions often so powerful?

Thursday, November 10, 2011

Images of Sonic Objects

Godøy, R. I. (2010, April). Images of sonic objects. Organised Sound, 15(1), 54-62. Cambridge University Press. Retrieved October 10, 2011, from Scholars Portal Journals

Largely based on the theories of Pierre Schaeffer in his Traité des objets musicaux (1966), but also drawing on more recent evidence from the study of musical imagery and support from the theory of embodied cognition, Rolf Inge Godøy, Professor at the Department of Musicology, University of Oslo, argues that the “sonic object” is the most significant timescale of music with regard to human’s ability to form stable memory images of music (sonic images) from continuous sound.

First, Godøy gives some useful background information on musical imagery, which is defined as the “mental capacity for imagining musical sound in the absence of a directly audible sound source”. Placing musical imagery in the broader context of mental imagery, he explains that there is generally a “functional equivalence” between real-world perception and action and imagined perception and action. (For example, recalling the last verse of a song would take longer than the first verse because people usually scan through the song from the beginning.) Furthermore, neuroscientific research shows that mental imagery and real perception and action share much of the same neural substrate. Of particular interest in musical imagery is that auditory and motor imagery seem to be bidirectionally linked. (For example, when professional pianists listen to piano music, the motor areas of the brain are also activated. Vice versa, when the pianists see silent piano performance actions, they also mentally hear the music associated with those actions.) Then, putting musical imagery in the perspective of embodied cognition, which sees perception and cognition as intimately linked with sensations of movement, Godøy argues that body movements are integral to music and that sound-events should be “understood as included in some kind of gesture trajectory”.

All of the above background information helps to prepare the reader for Godøy’s ideas about the nature of sonic objects, which he defines as “holistically perceived fragments of sound, typically with durations in the 0.5 to 5 seconds range”. He justifies this timescale by citing research that shows that listeners can generally recognize salient musical features, such as style, rhythm, texture/timbre, modal/tonal features, and expressivity, within this 0.5 to 5 seconds range. He then points out that theories of memory support the idea of sonic objects as coherent chunks of sound that are perceived and imagined in the present moment (in a series of “now-points”). In this way, an entire piece of music is basically a chain of sonic objects perceived and imagined chunk-by-chunk, moment-by-moment. Godøy describes three types of sonic objects: 1) Impulsive, meaning abrupt attack followed by decay, 2) Sustained, and 3) Iterative, meaning a quick series of fluctuations (e.g. tremolo). Given the integral sound-gesture link in the embodied perspective, he remarks that the three types of sonic objects correlate well with impulsive, sustained, and iterative body gestures. And given the bidirectionality between motor and auditory imagery, Godøy believes the “kinematics and dynamics of sound-related actions can create images of sonic objects”, which carries the implication that action imagery can actually enhance musical imagery and, therefore, can potentially be applied in various contexts, such as musical practice, research, and education.

Though slightly difficult for me to digest, I still found this journal article quite fascinating. Having read a chapter titled “Imagined action, excitation, and resonance” by Godøy (2001) in a book called Musical imagery, which argues that “images of sound-producing actions… can enhance [the] capacity for imagining sonorous qualities” (p. 237), I was curious to find out if Godøy has written anything else on this subject more recently. As it turned out, he indeed has, and I chose this article because it offers more up-to-date information on musical imagery, a topic that I am deeply interested in.

First of all, I was not surprised at all to discover that auditory and motor imagery are linked; I can relate well to the experience of having the urge to move my fingers and “play along” when listening to other pianists performing pieces that I am acquainted with. Being a performer, I have absolutely no doubt that body movements are integral to musical experience. But Godøy’s suggestion that there is an important gestural component to sound would still have seemed a little strange to me had I not taken a course in conducting two years ago, which certainly made me much more aware of how gestures can accurately represent various sound qualities (with a lot of practice, of course).

What impressed me the most about this article was the fact that something as private and seemingly unobservable as imagery could be systematically studied and theorized upon so extensively. I think that Godøy backs up his argument about sonic objects convincingly. What I am primarily interested in, however, is whether action imagery would really prove effective in developing musical imagery in the context of mental practice, as his view implies. Up till now, I have rarely employed the strategy of mental practice myself. But I have always been taught that I must first know what kind of sound I want (in my “inner ear”) before I can experiment with various ways of pressing the keys that would get me closer to realizing that sound. So it seems to me that the music should come first and the action subservient to it. Nevertheless, I suppose that after some physical practice, the sound would become inseparable from the action associated with it, and, at this point, action imagery would be effective in bringing forth musical imagery. So perhaps one needs a certain amount of physical practice on a particular piece before action imagery can be used? Or maybe it would simply be best for one to start developing mental practice skills early on in one's training?

Godøy, R. I. (2001). Imagined action, excitation, and resonance. In R.I. Godøy, & H. Jørgensen (Eds.), Musical imagery (pp. 237-250). Exton, PA: Swets & Zeitlinger Publishers.