So this is my term paper for my “Mind, Machine, and Language” course. It was basically an introductory neuroscience course with an emphasis on linguistics and language acquisition. The topic I chose was “Music Cognition”. I was inspired by the Radiolab episode called “Musical Language“. I wanted to explore the Rite of Spring riot and learn more about Jonah Lehrer‘s explanation, but I didn’t really have time to find his book and read it, so instead just read a ton of journal papers. When I started writing, I was going to try to give a neurological explanation for why I like strange music, and why individual musical opinions can differ so wildly, but after writing all the background information, I had already gone over the number of required words. Also, my grades here don’t matter, so after spending a lot of time on it, I didn’t really care anymore. The result is a somewhat poorly organized, thesis-less paper, but I think it is still quite interesting and has some cool tidbits of information that you probably didn’t know. Also I’d just like to say that I kept coming across awesome papers from my university, McGill, which is always a nice feeling. Turns out McGill does do some great research in almost every field I’ve had to do serious research, especially music perception! Lastly, if you find this stuff interesting, theres a book by a McGill prof named Dan Levitin called “This is Your Brain on Music“. It was bestseller, and I’ve heard its pretty interesting, although probably not quite as deep into the science as some would like. So, I hope someone out there reads this paper, and learns something new. Enjoy!

Introduction

Music is found in all societies and an impressive amount of the cognitive skills required for understanding it develop in young children without their conscious effort (Trehub, 2001). While it is unclear whether we evolved capacities for music as an accidental byproduct of language or for some greater social purpose, it is undeniable that it is deeply engrained in our brains and is a universal part of the human experience.  Even though many conventions of Western tonal music may seem like they must be universal to a Western listener, it is only through repeated exposure to this music, especially during childhood, that these conventions become so fused into our musical minds. While, it is impossible to perceive music from other cultures in the same way that natives of that culture do, some aspects of music do seem to be truly universal and have a neurobiological basis.

Even if certain aspects of music are built into our brains, the way each person experiences a piece of music is very different and as such, individual tastes vary widely, even within a single culture. This can largely be attributed to just some level of subjectivity, but what is truly fascinating, is the extent to which individual tastes in music can differ. Two people can disagree so much that one finds a piece of music deeply moving and pleasurable while the other can find it irritating to the point of refusing to refer to it as music. Why do we have such different opinions of music? The answer must be hidden somewhere in our brains.

Why Study Music?

As listening to music has such widespread effects on the brain, and vast differences in how it is perceived by individuals exist due to differences in musical training and cultural background, it is an extremely difficult topic to study. Additionally, when listening to music that involves sung lyrics, the brain is processing both the music and the language of the lyrics at the same time, complicating matters even further. Nonetheless, music perception is fascinating and has been a topic of great interest for neuroscientists for over a century (Critchley & Henson 1977). The remarkable pattern recognition abilities of the brain are perhaps displayed best by its power to process music. A good example of a complex task that we can perform easily is our ability to recognize songs despite major changes from past examples. Changes to the musical key, instrumentation, amount of noise, or even to the song and melody structure itself are easily filtered out in order to recognize the original. As such, we must have complex mechanisms for storing abstract representations of songs. Additionally, the fact that musical abilities are learned by only a small subset of the general populace through explicit tutoring make it an ideal field in which to explore plasticity. Understanding how we simply perform these complex processes can show us a great deal about our brains and ourselves.

Many studies reduce the inherent complexity associated with music perception by examining the effects of simplified acoustic stimuli such as single tones, chords, phrases (short series of tones as are often used in Western music), or rhythms. While this approach combined with the advent of fMRI and other brain imaging technologies is making some significant progress towards understanding some music related brain phenomena, it misses out on many of the effects that listening to full pieces of music can have on us. As is the case with much of neuroscience, many hypotheses about music processing come from neurological studies of patients with disorders or brain damage. By observing a patient with localized brain damage who has trouble with a specific musical task, researchers can at least point to that part of the brain and confidently say that it plays an important part in the process. Other studies make extensive use of brain imaging techniques but due to the complexity of the response to musical stimulus, often the results are limited to proving the existence of some sort of activity response to a specific event. As such, much of the information we have about the localization of music processing in the brain is very general and the actual processes by which music is processed into a precept largely remain a mystery.

Short overview of acoustics/hearing

Acoustic waves are periodic perturbation of pressure traveling through a medium. These acoustic waves travel from a vibrating source through the medium, while being reflected and modified by the surrounding space. Eventually some of those waves hit our outer ear, which is really an acoustic funnel that collects the sound in front of us and sends it into our ear canal. These variations of air pressure cause vibrations of the tiny bones of our inner ear. These bones act as a gateway for the vibrations to enter the liquid of our inner ear. They are constantly adjusted by tiny muscles that we are unaware of in order to amplify the vibrations to the appropriate levels for our inner ear. It is this automatic adjustment that allows us to hear the detail in very soft noises when it is quiet and at the same time protects our hearing mechanisms from very loud noises. The vibrations travel through the liquid of the inner ear and disturb cilia, or tiny hairs, that are spread in clusters spread along the basilar membrane in the cochlea. This membrane has varying tautness which allows different groups of hairs to react depending on the frequency of the sound waves. When a hair is bent, it opens a gated ion channel which triggers an action potential. The electrical signal of the action potential is sent along the auditory nerve to the brainstem. From the brain stem, the signals are sent around the brain, with a large portion of the processing happening in the auditory cortex. The entire process up until the auditory nerve is understood fairly well, but as soon as the brain involved, our understanding of the process of music perception process quickly fades.  Further, even if hearing mechanisms were completely understood, we still would not know how the sounds we hear are interpreted as music.

What makes it music?

Even from a non-scientific, subjective point of view, the question of what is music is unanswerable. Further, even when listening to a specific piece of audio, sometimes the line between noise and music can be very blurry. This no doubt is due to the large overlap between the cognitive processes involved in language, music, and general sound identification. Obviously all auditory tasks must share most of the same regions of the brain, but it seems that there is at least some separation of specifically music related tasks (Peretz & Zattore, 2005). This question of which sounds are music are pertinent not only for some experimental musical pieces that truly are questionable, but also of a specific individual’s perception of a piece. Therefore, there must be some differences in how different people perceive the sound. Some differences have been shown to be dependent on musical training, cultural exposure, and even somewhat on certain predispositions (Grahn & McAuley, 2009).

It would seem that when hearing any sort of auditory stimulus, part of the auditory cortex is constantly analyzing the sound and looking for patterns. This includes patterns between the relative frequencies of the sounds (pitch and harmony), the timing of the sounds (rhythm), and the overall structure (repetition, and musical conventions).  The brain has been shown to respond to these patterns, at least in the case of pitch (Peretz & Zatorre, 2003), regardless of attention. So it would seem that it is the positive activation of these pattern detecting structures that would lead us to subjectively classify an auditory input as music. Upon recognizing a pattern, whether it be a beat, a melody, or a chord progression, we start to have expectations about what is coming next. When these expectations are fulfilled, our brain continues humming happily along getting ready for what is next, even anticipating upcoming changes (Sridharan et al, 2007). When our expectations are suddenly not fulfilled, there is an increase in brain activity (Regnault et al. 2001). When we can make sense of the unexpected sounds, such as resolving a complex chord into the larger musical context, it generally gives us some pleasure, as any music lover can tell you of when they hear something unexpected but innovative and interesting. However when we cannot make sense of it, we usually perceive it as a mistake and it generally causes some discomfort, as in the case of a wrong note or a beat that is out of time. To understand how our brains can know what to expect, we must take a closer look at how pitch and rhythm are interpreted.

Pitch

Sound is made of waves of pressure. The frequency of these waves is what we perceive as pitch, or how “high” or “low” it sounds. When discussing pitch perception, generally it is relationships between pitches that are important, whether it is between multiple tones played in sequence, as in a musical phrase in Western terms, or in unison, as in a chord. It is these relationships between pitches, or in musical terms their consonance and dissonance, that provide the building blocks of Western tonal music. It is important to note that some aspects of consonance and dissonance are culturally dependant, meaning a certain chord progression may sound qualitatively good to an individual who grew up listening to Western music while the same progression may sound bad to an individual with only a background in Chinese classical music. Regardless, brain responses to pitch relationships can be studied in order to pick out which phenomena are truly universal and biologically based, and which are more related to plasticity and previous exposure.

Consonance describes relationships between notes that a listener would subjectively say sound pleasing while dissonance describes the perception of unpleasing or rough notes. Intuitively, it seems that the musical theory of harmony must be based on physiological properties of our auditory system and neurological properties of the corresponding cognitive systems. When an interval is played, neurons sensitive to frequencies present in the interval respond by firing. For consonant intervals, the timing of the auditory nerve responses contain a representation not only about the heard frequencies, but also of harmonically related pitches (McDermott & Oxenham, 2008). This should make sense to anyone familiar with Western music theory because when playing certain chords, often one has a sensation of hearing tones that are not actually present. It must be this mechanism that allows one to sometimes pick out the root of a chord even when it is missing. By contrast, activity caused by dissonant intervals do not contain representations of related notes and just as the overtones of dissonant notes are close together and cause unpleasant beat patterns, the representations of these tones in the brain are too close together to be resolved. As a consequence, they interfere with one another which can cause fluctuations in the firing of other auditory neurons, causing the perception of unpleasantness. This also seems obvious as a highly dissonant chord can sound very “muddy” and it is very difficult to pick out what notes are present.  This and other related work are finally allowing researchers to show that “the basic pitch relationships governing music may be rooted in low-level sensory processing and that an encoding scheme that favors consonant pitch relationships may be one reason why such intervals are preferred behaviorally (Bidelman & Krishnan, 2009).”

The consonance of a note within the context of a longer series of notes, often referred to as melodic consonance, is a more complex issue, and one that has been studied since the ancient Greeks (Anderson, 2004). Even the theoretical rules of harmony cannot suffice to explain all of melodic consonance. Nonetheless, many cross-cultural studies and studies of infants have shown some biologically based preference for consonant melodies. The neurological foundations for this preference likely are interrelated with the general perception of consonance in combination with some of the memory structures involved in music perception. One compelling attempt to explain the theory behind it uses the overtone series but does not actually go further into doing any brain imaging studies (Anderson, 2004).

More difficult to study still, is the consonance of full chord progressions.  This generally is a culturally-dependent mechanism, but should be possible to study as deviations from harmonic expectancies have been shown to generate robust responses (Regnault et al. 2001). While many of the rules of chord progression consonance are governed by the same music theory rules as those of individual chords and melodies, the brain mechanisms to identify them may appear slightly later in development. One study showed through neuroimaging that by the age of five, the harmonic appropriateness of a chord progression can be judged (Koelsch et al. 2003). Further, it has been shown that listeners learn these principles through only passive exposure to Western music (Bigand 2003).

Despite everything we know about pitch perception, and the fact that we can finally proclaim that certain phenomena like dissonance are real, there is still much to be learned about the entire process. Many pieces of the puzzle have yet to be localized within the brain, and further many questions remain about what aspects are biologically based rather than learned through exposure.

Rhythm

While rhythm, or the temporal structure of music, may at the surface seem simpler than pitch perception, it too involves many areas of the brain and is extremely complex. In dissecting perception of rhythm, it is important to notice the two types of temporal organization present in music. First the segmentation of a sequence into temporal groups based on duration, and second, the underlying regularity of the tempo or beat (Fraisse 1982). While these two tasks are no doubt related in some ways, the currently available research shows evidence of a functional dissociation between these tasks of grouping and identifying regularity (Peretz & Zatorre, 2005).

Throughout all of the research on rhythm, one thing seems clear; a very strong connection between perception and identification of rhythm and parts of the brain usually identified with movement. That is, it seems that when a listener is perceiving the beat, parts of their brain needed to reproduce the beat are being activated. Perhaps this is why rhythm is so primal and often we spontaneously feel the need to move with the music that we are listening to.

Emotional Response

No one can deny the emotional power of music. It seems to be able to bypass the language and abstract thinking parts of the brain and connect directly with our emotional centers. Some aspects of music seem to transcend cultural tendencies and indeed, some emotional cues in musical have been shown to be consistent across cultures (Balkwill et. al. 2004). While the specific pieces of music that cause truly deep emotional responses vary from person to person, many people report having a truly pleasurable response to music and cite this as the reason they listen to music in the first place. Further, a recent study proved this link between music induced pleasure and emotional arousal as measured by certain physiological signals (Salimpoor et. al. 2009). The study showed that music is able to cause certain physiological responses, the peak of which is associated with the “musical chills” response, typically only linked with physical stimuli such as eating food and taking drugs. Music seems unique in this regard as it is able to cause such immense pleasure yet has no tangible physical reward, and no specific functional goal.

Strange & New Music

So why can some people call a piece of music enjoyed by others simply noise? Or more shockingly, how could one of the greatest symphonies of our century have cause a riot at its first performance? This would seem largely to depend on past exposure to music. When listening to music that isn’t understood by the brain, it is constantly working to figure things out and consistently being let down. It is the fulfillment of musical expectations that keep us happy, and when one has no idea what to expect, things can get uncomfortable.  Passive exposure to music as a child would seem to create a sort of perceptual filter, similar to that created by language processing, that is needed in order to fill in the gaps, whether it be in perceiving an implied rhythmic structure, or understanding a complex key change. Without this filter, it is difficult to appreciate the nuance in a piece of music. Contemporary music does sometimes try to push the limits of what music really is, such as the experimental works of John Cage or the musique concrete movement. While much of this work is conceptual, I am curious about the brain activity of listeners of this music. Perhaps those familiar with and truly listening to the music are somehow purposefully redirecting the sounds to the music processing facilities of the brain while a dismissive listener never lets it reach that far. Perhaps they hear it as noise because they don’t give their brain a chance to properly process it.

Bibliography

Anderson, Jared E. (2004) The perception of melodic consonance: an acoustical and neurophysiological explanation based on the overtone series. [Preprint] (Unpublished)

Balkwill, L.-L., Thompson, W.F. and Matsunaga, R. (2004) ‘Recognition of Emotion in Japanese, Western, and Hindustani Music by Japanese Listeners’ , Japanese Psychological Research 46(4): 337–449

Bidelman , Gavin M. and Krishnan, Ananthanarayan. Neural correlates of  consonance, dissonance, and the hierarchy of musical pitch in the  human brainstem. Journal of Neuroscience, 29(42):13165–13171, October 2009.

Critchley M, Henson R, eds. 1977. Music and the Brain. Studies in the Neurology of Music. London: Heinemann

Grahn, Jessica A. and McAuley, J. Devin, Neural bases of individual differences in beat perception. NeuroImage, Volume 47, Issue 4, 1 October 2009, Pages 1894-1903

Langer, Gerald 2005, Neuronal Mechanisms Underlying the Perception of Pitch and Harmony. Annals of the New York Academy of Sciences 1060, 50-52.

McDermott, Josh H. and Oxenham, Andrew J., 2008. Music perception, pitch, and the auditory system. Current Opinions in Neurobiology, 2008 August; 18(4): 452–463.

Peretz I, Zatorre R. 2003. The Cognitive Neuroscience of Music. Oxford: Oxford Univ. Press

Peretz , Isabelle and Zatorre, Robert J, Brain Organization for Music Processing. Annual Review of Psychology 2005, 56, 89-114

Regnault P, Bigand E, Besson M. 2001. Event related brain potentials show top-down and bottom-up modulations of musical expectations. Journal of Cognitive Neuroscence. 13:241–55

Salimpoor VN, Benovoy M, Longo G, Cooperstock JR, Zatorre RJ (2009) The Rewarding Aspects of Music Listening Are Related to Degree of Emotional Arousal. PLoS ONE 4(10): e7487.

Sridharan, D., Levitin, D. J., Chafe, C. H., et al. (2007). Neural dynamics of event segmentation in music: Converging evidence for dissociable ventral and dorsal networks. Neuron, 55, 1–12.

Tramo, M. J., Cariani, P. A., Delgutte, B., & Braida, L. D. (2001). Neurobiological foundations for the theory of harmony in western tonal music. Annals of the New York Academy of Sciences, 930, 92-116.

Trehub, S.E., 2001. Musical predispositions in infancy. Annals of the New York Academy of Sciences 930, 1–16.