Essential Limitations in current Neurochemical Studies of Music

by James A. W. Gutierrez, Azusa Pacific University, college of music and art, adjunct prof.

In April, 2013, Mona Lisa Chanda and Daniel Levitin published “The Neurochemistry of Music”, which presents “peer-reviewed scientific evidence” supporting claims that musical influences may correspond directly with neurochemical changes, specifically correlating “musical reward” with dopamine/opioids, stress relief with cortisol, and musical “social bonding” with oxytocin/vasopressin. Ideally, the music-as-medicine pursuit is pure in its intent toward the relief of human suffering, be it behavioral/emotional//physical/social, through a more natural medium than, say, pychosomatic drugs. However, such a strong quantification of music, and generalization of musical elements, invokes the familiar pharmaceutical path where an ambitious medical community responds to a irreducibly complex system of sociobehavioral situations with a grossly oversimplified, pill-sized answer. While there are certainly clinical uses for music, the first mistake a clinician could make, and hence the primary abuse of both music and a patient, would be to attempt to incarnate, confine to physical flesh, the essentially abstract expressive form that is music.

Such extreme reductions in musical semiotics are prevalent throughout current experimentation involving dopamine and opioids. Levitin reports: “Pleasant (consonant) and unpleasant (dissonant) music were contrasted, and the results conformed activation of the ventral striatum during pleasurable music listening.”[2] In tests examining the effect of music on the stress hormone cortisol Levitin reports: “Relaxing music mimics soothing natural sounds such as maternal vocalizations, purring and cooing (soft, low-pitched sounds with a gradual amplitude envelope), which decrease sympathetic arousal.”[3] When observing levels of polypeptides serum oxytocin and vasopressin (currently thought to regulate social behavior) Levitin reports: “a single 30-minute voice lesson was associated with an increase in serum oxytocin levels relative to a pre-lesson baseline in both professional and amateur singers” and “open-heart surgery patients who listened passively to experimenter-selected soothing music for 30 minutes one day after surgery has higher levels of serum oxytocin compared to bed-rest alone.” [4]

The systematic placement of music in such generalized categories as consonant=pleasure/dissonant=stress, “relaxing” music, etc., with the expectation of uniform results only demonstrates the assumption on the part of the experimenter that music, as represented by a particular style/tempo/dynamic range/etc., should behave as a static unit even in the testing of a broad diversity of listeners. Not only does this ignore the music biases of the experimenter, the testing environment all but extinguishes the affective contexts in which real music listening would be experienced. Could not a familiar yet up-tempo progressive rock song be “relaxing”? Perhaps the oxytocin levels post-singing lesson involved factors such as familiarity, personal connection/association, successful performance in front of an intimidating tester, or perhaps it could just maybe have been the lyrics of the song? Could not “dissonant” music be “pleasurable”? It is precisely the paradoxical nature of musical pleasure that makes musical expression unique, and problematizes this whole method of research. As Oscar Wilde observes- “After playing Chopin, I feel as if I had been weeping over sins that I had never committed, and mourning over tragedies that were not my own.”[5] Would this response be observable in his dopamine/opioid levels?

It could be objected that it is merely seeing music in the context of scientific scrutiny that makes a musician uncomfortable, a kind of ‘we don’t belong here’ awkwardness. Could it be that I am simply afraid that music may be demystified if subjected to an empirical testing environment? Absolutely not. Even the previously stated testing is not completely void of value. The last ten years of testing the brain in all subjects surrounding music have yielded a trove of useful information. Laboratory mice have been included in the research: “Two species of ‘singing mice’ which display an unusually complex vocal repertoire exhibit high oxytocin receptor binding within regions related to social memory. Injection of oxytocin increased vocalization levels while oxytocin receptor infant knockout mice engage in fewer vocalization and show marked social deficits and higher stress levels.” [6] These findings at least establish the biological basis for a social component in music, and maybe even supports the notion that music plays an important role in creating social bonds.

Neuroscientists essentially portend to deal with ‘universal’ structures, by nature of their scope. The more they universalize musical elements, the less they are observing actual music, and they run the risk of trivialization all ‘findings’ therein. The ideal of music-based treatments is that they are noninvasive, have minimal or no side-effects, are inexpensive, convenient, and are completely ‘natural’. While the merit of this endeavor cannot be denied, let researchers admit that this reverse-engineering is in its fetal stages of development, where I contend it will remain until a more advanced treatment of musical elements can be introduced into testing. While it is delusional to attempt to incarnate an abstraction, to acknowledge an enigma and conduct research while remaining subject to it can be a step toward real understanding.