|Remember, information is not knowledge;
knowledge is not wisdom;
wisdom is not truth;
truth is not beauty;
beauty is not love;
love is not music;
music is the best.
– Frank Zappa
The power music has to elicit emotional responses has spanned human history and cultures. Indeed the emotional nature of music may be its most defining feature . It is the most common reason people cite when asked why they listen to music  (Figure 1), and even young children have an innate ability to accurately describe the intended emotional content in musical pieces . However the mechanisms behind the surprisingly strong emotions that music can elicit have been a mystery. A mystery that many philosophers and scientists from Aristotle to Darwin have speculated about . Because music does not have an obviously biological or adaptive role it is unclear what elements would be responsible for the strong and seemingly universal effects it has on the brain, the body and behaviour . Some philosophers have said that it must simply be an "auditory cheescake" - an accident of evolution . However, with recent advances in neural imaging it has been possible to show that music can indeed activate the areas of the brain that are known to be involved in emotions, rewards and core adaptive drives: the limbic, paralimbic and mesolimbic regions . It is therefore starting to become clear how the different elements and structures in music combine to activate these ancient areas in the brain and cause the strong emotional and physiological responses that almost everyone is familiar with.
|Bobby McFerrin demonstrates how well people intrinsically understand the hemitonic pentatonic scale.
Table of Contents
|Figure 1. The reasons most people listen to music is for the emotional experience. |
1. Musical structure and emotion
There are many attributes of musical structure. Some are used in similar ways across cultures like tempo and meter, whereas some, like tonality, differ between cultures quite dramatically . Variations in attributes combine to make each piece of music unique and recognizable. Some of these attributes have been associated with activating specific areas of the brain and thus evoking specific emotional reactions.
1.1 Scale structure
1.1a major and minor keys
Modern western music is generally based on one of three scale structures: major, harmonic minor and melodic minor . These two main modes of tonality, major and minor, are generally immediately recognizable by people who have been exposed to western music from birth . Although trained musicians are more proficient at recognizing these structures, most lay people can discern the overall tone of a musical piece, and usually fairly quickly . As they are so widespread and recognizable, these structures operate as culturally defined emotional cues. The major scale and general consonance has often been associated with happy, cheerful and uplifting emotions while minor and dissonant chords have been associated with sad, lonely, or melancholic emotions . While functional MRI evidence has been somewhat mixed, in general it has been shown that minor and dissonant chords activate the amygdala, retrosplenial cortex, brain stem and cerebellum , whereas major or consonant chords are associated with the bilateral ventral and left dorsal striatum, left anterior cingulate and left parahippocampal gyrus . One of the main reasons for mixed results in some of these studies may have to do with not controlling for differences in active and passive listening. Pallensen et al (2007) found that emotional reactions were more pronounced when the participants were not required to engage in a cognitive task during the listening stage. It has also been suggested that the observed differences in neural processing of major and minor modes is a learned association due to the domination of major modes in modern western music . It may therefore, not be a universal or biological predisposition to a specific abstract structure, but a type of cultural conditioning . For example although western music uses 3 main scales, Hindustani music uses more than 400 scales . People who were raised with Hindustani music may not react to major scale cues with feelings of happiness . There have been mixed results in the literature surrounding this issue due to the fact that some musical attributes do appear to be universal . Despite encultured differences, modal comparisons appear to activate the inferior frontal gyri and the medial thalamus and are hypothesized to be involved in judging tonality .
1.1b scale structure and expectation
Expectation in music has also been connected to scale structure. Most scales are based on unequal sized steps in tones, which add novelty are are more easily recognized and remembered, even by infants . These unequal steps also set up an expectation that when a shorter, and thus unstable, step is taken, that there will be a longer, more stable step in the future - and thus a sense of resolution . This basic and seemingly innate ability humans have for recognizing and remembering unequal tonal patterns is the basis for musical expectation. Indeed, infants are predisposed to be able to process and remember songs that are based on unequal step scales as compared to equal step scales . Although there are musical pieces based on equal step scales such as the chromatic scale, these are not common and do not tend to evoke an emotional reaction, (or much understanding at all), from most people . The occasional uniqueness of the unstable smaller tonal steps may provide a predictive puzzle for the brain and thus engage or activate it much more than patterns of the same sized steps. The expectation and anticipation this creates is one of the fundamental qualities that evokes an emotional response in listeners .
1.2 Acoustic Structure: tempo, complexity, loudness, timbre
Acoustic cues in music tend to be more universal , perhaps because they are processed in similar ways to emotional speech . Indeed, when people are happy or joyful their speech often becomes faster, and with greater ranges of pitch, whereas when they are sad they tend to speak slowly and softly with lower pitches . Listeners tend to perceive these changes in auditory cues very quickly and accurately even when they cannot observe facial expressions or body language . It is not surprising therefore, that musical cues that mimic these qualities are perceived as having the same emotional meaning . However, it is not always simple. These cues are not perceived as single entities but instead are processed and combined in order to evoke a specific emotional reaction in listeners . In addition these cues, and thus the emotional content of a musical piece, can change over time and are not necessarily constant or mutually exclusive . However, in general happy music has the following attributes: fast tempo, major chord structure, wide range of pitches, high amplitude, rhythmically even, and low complexity . Sad music on the other hand tends to be high in complexity, and slower in tempo  and generally stimulates more nervous system arousal .
1.2a Tempo, complexity, loudness, timbre and familiarity
Across cultures people associate fast tempo with happy or angry music and a slow tempo with sad or melancholic music . Highly complex harmony is generally associated with sadness or sorrow while simple harmonic structure is associated with happiness or serenity . Angry music also seems to be associated with high complexity but differs from happier music on other cues (. Soft, quiet music generally evokes sadness while loud music is associated with joy or anger depending on other cues . Instruments with a wide pitch range are often associated with joy, whereas instruments with a softer texture or narrower pitch range are often associated with sadness .
The capacity of people to detect emotion in music is impacted by the level of familiarity both with the specific piece of music and with the cultural tonal system the music is based on . Certainly subjects respond better to culturally native vs non-native music . However, even without being familiar with a piece or its culturally based structures, the universal cues of tempo, loudness, and complexity have an impact on listeners emotional reactions . Peak activation of the mesolimbic structures in the brain tend to happen during subject chosen pieces which are generally not only familiar but favourite pieces of music  . Indeed, a sense of personal meaning and memories with specific pieces dramatically increase the emotional response . Salimpoor et al (2011) also found that activation in the caudate nucleus during anticipation of a musical reward to be higher during familiar pieces.
2. Neuroanatomy of the emotional processing of music
|Figure 2. The diversity of brain regions involved in the processing of musical stimuli|
Music is one of a few stimuli that activates a wide diversity of brain regions including the hippocampus and frontal lobe for memory processing, the cerebellum for pattern recognition, rhythm processing and movement, the prefrontal cortex for expectation and prediction, the visual and motor cortex for playing an instrument or dancing, and the limbic, paralimbic and mesolimbic regions for emotional processing  (figure 2). After music is transduced into electrical impulses in the ear these signals travel to the midbrain and then to the auditory cortex . The auditory cortex analyzes and organizes these signals into a meaningful pattern by interacting with various other regions of the brain . Despite recent advances in understanding the neural underpinnings of musical cognition, the structures that process the emotional experiences are just beginning to be elucidated  (Figure 3). Historically, the focus of affective psychology and neuroscience has been on negative emotions . In addition, any of the research that was done on positive emotions usually employed a direct chemical stimulus . Even the study of music has focused more on development and cognition rather than its emotional effects . However, the study of musical emotion is providing insights into the neural underpinnings of both positive and negative affect that are more naturalistic than direct chemical stimulus .
2.1 Limbic and paralimbic brain regions
The limbic and paralimbic regions of the brain are not always well defined, however, most neuroscientists would agree that the limbic structures include the hippocampus and amygdala . The paralimbic structures include the orbitofrontal cortex and subcallosal cingulate, which are often considered part of the prefrontal lobes as well as the parahippocampal gyrus, which is sometimes classified as being part of the limbic lobe.
The hippocampus is important for learning and memory formation and therefore instrumental in novelty recognition . Thus activation in the hippocampus is often correlated with passive listening to familiar pieces. However, even with unfamiliar music the hippocampus is activated and thus may be important for comparing the novel information to stored information . This may help put music into cultural contextes. The hippocampus is also hypothesized to be important in processing emotion although this is not well understood .
The amygdala is highly correlated with strong emotions, such as fear, and is activated by dissonant music and pieces that are reported to be sad or melancholic . In contrast there have been reports of musically induced pleasure being correlated with the deactivation of the amygdala  . However, it appears to be more complex than that. Different nuclei within the amygdala can be activated by both sad/dissonant music, and by happy /consonant music, and these distinct nuclei innervate different regions in the paralimbic systems of the brain . These musical derived insights into the amygdala may be very important for the ongoing study of mood and anxiety disorders, which are often correlated with over activation or dis-regulation of the amygdala.
Activation in the orbitofrontal (OFC), ventromedial prefrontal cortex (VMPFC), subcallosal cingulate and frontal cortical areas has been correlated with listening to joyful or pleasant music . The OFC and VMPFC areas are important for decision making and are hypothesized to be the site for the cognitive integration of incoming musical reward information from the limbic and mesolimbic regions . The subcallosal cingulate acts as a gateway between the frontal cortex cognitive areas and the limbic emotional regions and has been shown to be important in major depression . When this area is overactive negative emotion tends to over take thought and dis-regulate mood . It is still not understood how music, by activating this area, modulates and improves mood and affect, however it is another area in which the study of musically induced positive emotions could lead to key insights into mood disorders.
|Figure 3. The emotional regions of the brain that are activated by musical stimuli. |
2.2 Mesolimbic brain regions
The mesolimbic pathway is a group of dopamingeric neurons that project from the midbrain into the limbic, paralimbic and cortical areas and have been associated with reward, emotional reactions and core adaptive drives . This area, (which includes the dorsal and ventral striatum), is activated by highly addictive drugs like cocaine, as well as adaptive behaviours such as sex, eating and drinking water when thirsty . Passively listening to music that induces strong feelings of pleasure also activates this area .
2.2a Positive peak emotion
Musically induced feelings of pleasure, along with physiological responses such as chills or shivers as well as corresponding changes in heart rate, breathing, temperature and galvanic skin responses, have been associated with activation in the left ventral striatum, specifically in the nucleus accumbens (NAc) . The NAc receives dopamine input from the ventral tegmental area (VTA) and innervates the amygdala, hippocampus and hypothalamus . Despite the fact that music is intangible the patterns of activation in the NAc during passive listening are similar to the patterns observed during exposure to cocaine and food . In addition peak activation in the NAc and physiological responses both correlate with subjective reports of peak pleasure . Since chills are a byproduct of autonomic nervous system arousal and can be triggered from both pleasant and unpleasant stimuli, it has been shown that the chills are merely a side-effect of the dopamine release in the NAc .
2.2b Negative peak emotion
Although most people appear to associate the feeling of chills with peak positive experiences, it is clear that musically induced sadness, melancholy or nostalgia often result in even greater physiological arousal . Often this result is greater in female listeners than in men listeners .
2.2bi How can sad music make you happy?
Although sad music is classified based on the obvious fact that it elicits sad or melancholic feelings in the listener, it often tends to have a more complicated emotional effect over time. Often, especially upon the culmination of a peak emotional experience, sad music can induce a sense of joy, happiness or serenity . One hypothesis is that sad music triggers the release of prolactin from the pituitary . Prolactin is released during tears of sadness but not tears of joy or irritation and might contribute to the modulation of mood . In addition Levitin (2009) hypothesizes that sad music makes sad people feel more understood and connected to others thus reducing a sense of isolation and encouraging a more positive mood.
|Figure 4. Areas of the striatum that are activated during anticipation and during peak reward. |
Expectation in music is well recognized as being an important structural factor in eliciting emotional response. Leonard Meyer theorized in 1956 that because music is played out over time it allows for a unique experience of pattern recognition and temporal expectancies, which composers can utilize through delay, violation and then fulfillment in order to elicit maximum emotional response in listeners . This unique quality of music may be one of the reasons it can provide peak emotional experiences that are similar in strength to addictive drugs . Interestingly, the NAc, which is highly active during peak pleasure, is not active during the expectation of reward . Instead it is the caudate nucleus in the dorsal striatum that is activated during periods of expectation and anticipation of the peak reward (Figure 4).
Because music is played out over time and provides temporal cues to the listener, it provides a way to examine the temporal aspects of the reward process from anticipation through peak reward. Using music to temporally map the reward process Salimpoor et al (2011) found that activation switched from the dorsal to ventral striatum as music moved from anticipation to peak emotional reward. Anticipation and expectation of a reward that was imminent was found to specifically activate the caudate nucleus in the dorsal striatum during passive listening and not the NAc . This is the first evidence that shows that distinct areas of the reward pathway are involved at different times during the reward process.