As you scroll through this blog, you are probably also listening to music on Spotify, YouTube, Apple Music, iTunes, a CD, a vinyl, a cassette tape, your Walkman, what have you. The near-universal fascination of humans with music is a tale as old as time. Interestingly, in contrast to other pleasure-inducing concepts such as food and sex, music does not have an evolutionary value that justifies its prevalence in our lives. In Darwin’s words, “As neither the enjoyment nor the capacity of producing musical notes are faculties of the least use to man in reference to his daily habits of life, they must be ranked  the most mysterious with which he is endowed”. If music is not essential to our survival, why does it hold such a significant place in our society? The answer lies in the rewarding effects of music – but what exactly are they?

“Reward System” Who?

To ensure the survival of our species and the fulfillment of basic needs, our brains developed a “reward system” which involves dopaminergic neurons that fire in response to a learned reward or a reward-predicting cue, producing a “dopamine high”. This neurotransmitter comes in two forms: tonic, the general level of dopamine available, and phasic, released in bursts during reward processing. Dopamine release occurs following the temporal difference learning model, which states that prediction errors, i.e. the difference between the outcome of a situation and our expectation, guide reward-associated learning. For instance, a positive prediction error (PPE)  when the outcome of a situation is better than expected, accompanied by phasic dopamine bursts, while a negative prediction error (NPE) happens when the outcome is worse than expected, and results in a decrease in dopamine.  

A Bit of Music Theory

So how does music act upon the reward system? To answer this question, we need to look at the musical features that make this art form rewarding. Musical pieces follow a culture-specific structure that allows listeners to create expectations and predictions. For instance, they have well-defined and easily predictable beginnings, middles and ends. Furthermore, Western music follows specific rules: the direction of the next pitch change can be predicted from the current pitch change size; the melody is often arch-shaped, with a rise in the middle and a return at the end; the distribution of tones follows certain statistical rules, e.g., the 5ths occur more often than the 6ths and the 7ths; and chord progressions follow set sequences. While these statistical rules might not be explicitly learned, music listeners subconsciously internalize those rules after years of exposure to Western music and use them to create expectations during subsequent music-listening experiences.

Why Do I Keep Hitting “Repeat”?

Now that we learned about the basic Western music structure, we can look at how the resulting expectations work in conjunction with the reward system to give us that “high” we feel from listening to our favourite pieces. Music theorist Leonard Meyer suggests that reward from music is not related to its emotion-inducing property but is instead the result of the way we process sequential events. Indeed, at every moment throughout a musical piece, we create predictions of what the next note, chord, or musical event will be, which will be either met, resulting in a PPE, or not. These expectancies can be either based on the veridical knowledge of the piece, if the listener is familiar with it, or on the implicit knowledge of the above-mentioned rules of the musical system, if the listener is exposed to a novel musical piece.

What happens with dopamine during prediction confirmation? When we first listen to a piece, the largest dopamine burst occurs from the closure of a phrase or of a section, associated with the completion of a predicted sequence and resolution of tension. However, once we become more familiar with the piece, the most significant dopamine release will occur at the beginning of such a section, acting as a cue for the reward – the completion of the section – that will arrive shortly.

But wait…isn’t the release of dopamine associated with an outcome that is BETTER than predicted?

Indeed, the prediction confirmation hypothesis, which states that dopamine is released following an ACCURATELY-predicted sequence of music, stands opposite to the previously discussed positive prediction error hypothesis, according to which, dopamine is released when the outcome is BETTER than predicted.

So which one is it?!

There is significant evidence for the prediction confirmation hypothesis. For instance, prior exposure to a piece increases familiarity, making it more predictable, and is associated with increased liking of it. Furthermore, pieces that follow musical conventions, i.e. easier to predict, are more appreciated than unconventional pieces. Lastly, musical fluency, which depends on individual levels of music-listening skills, helps predict musical sequences more easily, and is also associated with greater enjoyment of music.

However, there are limitations to the prediction confirmation hypothesis. For instance, over-familiarity with a piece – listening too many times in a brief period – will lead to a reduction in its liking. Additionally, very simple, conventional and predictable pieces are less appreciated by musically-fluent individuals than more complex pieces. So, the rewarding aspect of a musical experience is based on integrating accurate predictions and positive prediction errors, and the contributions of each theory depend on individual prior exposure to music.

What a popular song sounds like when our predictions are not met: https://www.youtube.com/watch?v=lXMskKTw3Bc

“Sound of Music”

Neuroimaging studies show that music processing depends on both cortical and subcortical structures. The subcortical structures, e.g. the striatum (including the NAcc), the amygdala and the hippocampus, are part of the reward system. The cortical structures, e.g. the prefrontal cortex (PFC) and the auditory cortex, including the superior temporal gyrus (STG), are associated with value-guided decision-making and perceptual processing of music. The auditory cortex is crucial to the listening experience, because parts of the temporal lobe store memories of previously heard music that serve as auditory templates. When these templates connect with the reward regions, listening to music triggers dopamine release and turns into an emotional experience. Of course, everyone has a unique music history – which makes everyone’s templates, predictions and rewarding experiences different. Certain core features that differ from one style to another, such as the timbre of instruments and chord changes, are stored in these templates, which create different expectations for music listened to subsequently. Since music exposure depends on demographic factors, culture and generation, individuals have unique templates, resulting in unique ways of processing and appreciating music.

“Sound of Silence”

The strength of the connectivity between these cortical and subcortical structures is important for music processing: it affects our appreciation of a musical piece. Salimpoor and colleagues found that functional connectivity between the NAcc and the STG predicted how much money people were willing to pay for a certain musical piece. Zatorre and colleagues recently demonstrated that stimulating the dorsolateral PFC, which modulates the release of dopamine in the dorsal striatum, resulted in participants increased payment and ratings for musical excerpts they already liked, while inhibiting it resulted in decreased liking and not willing to pay as much for those same excerpts.  

Is it possible that some people cannot appreciate and get any reward from music whatsoever? Mas-Herrero and colleagues tested whether “music anhedonics”, people who according to a self-report questionnaire have average general reward sensitivity but don’t get reward out of musical stimuli, have a different psychophysiological response to music from individuals who find music rewarding (“hedonics”). The authors found that as opposed to the latter, music anhedonics did not experience an increase in skin conductance or heart rate as a function of an increase in liking music stimuli. They did, however, experience the same psychophysiological changes to monetary reward as hedonics, demonstrating that their reward system is not damaged. This anhedonia is also not due to a lack of music exposure or inability to verbally associate emotions with musical pieces, as participants were also tested on these capacities. One potential explanation for this phenomenon is the lack of interaction between the cortical and subcortical structures, with both systems intact separately, but this hypothesis can only be supported with further neuroimaging and lesion results.  Thus, music anhedonia demonstrates once more how complex music processing is, compared to simpler stimuli and basic needs.

Enjoy your dopamine rush!