The paradox of sad music has been noticed and commented upon many times in the past. Nearly 400 years ago, Robert Burton observed:

"Many men are melancholy by hearing music, but it is a pleasing melancholy that it causeth; and therefore to such as are discontent, in woe, fear, sorrow, or dejected, it is a most present remedy." [1]

Common experience shows that Burton was right. We often turn to sad music for comfort and for pleasure. The survey presented by Liila Taruffi and Stefan Koelsch is a most thorough and convincing exploration of the phenomenon. Their survey examines different aspects of the problem: What are the different emotions evoked? What are the differences between listeners? Between cultures? etc. Almost certainly there are several reasons we enjoy sad music. The authors wisely describe the problem as "multifaceted" and "multidimensional." They find that "listening to sad music can lead to beneficial emotional effects such as regulation of negative emotion and mood as well as consolation." The authors' description of these beneficial effects is excellent.

The question I have is WHY this should be so. Why is it that sadness -- sadness in particular, and not anger or fear -- can be so soothing, so comforting? Quite a number of proposals have been made, including Davies, Levinson, Schubert, and others cited in the present study. I would like to suggest one more, based on the premise (not universally accepted, of course) that human emotions are to some degree the product of natural selection, as argued by Darwin, Tooby & Cosmides, Nesse, and others.

Emotions perform functions of benefit to the animal. Fear, for instance, prompts an animal to flee in certain circumstance -- anger to confront in other circumstances. These responses tend to enhance the animal's ability to survive. What is the function of sadness? The principal physiological characteristic of sadness is slowness of body and mind. Emotions “are specific, efficient responses that are tailored to problems of physical and social survival.” [2]. We must therefore ask: What problems of physical and social survival did sadness address? What is the adaptive value of slowness?

Many have conjectured that the original function of sadness – millions of years ago – was to withdraw energy from a task or goal that was no longer worthwhile [3]. Our ancestors would often have found themselves engaged in an enterprise that was failing – that was becoming too costly in energy compared to the potential gain, or too dangerous, or simply impossible. An unsuccessful hunt. An unsuccessful courtship. That was the problem – a recurring and serious problem. The “specific, efficient response” was the emotion of sadness:

"The experience of sadness signals to the individual that a loss, that she can do little about, has or is about to take place. Thus, it motivates the individual to conserve energy and resources through the reduction of effort and withdrawal or by triggering the action tendency to withdraw into oneself." [4]

Once an animal withdraws from a losing situation, the loss of energy serves a further function: it gives the animal time to rest, gather its strength and lick its wounds – literally lick its wounds – an activity that is both healing and comforting. To some extent at least, retreat is rewarded by pleasure. It is a relief – and therefore a pleasure – to escape from a bad situation, to relish one’s safety. It is a pleasure to rest after exertion. And as the cognitive abilities of our ancestors grew, they became able to evaluate the situation. Suffering a defeat or a disappointment, a person has something to think about and the opportunity to think – to ruminate – while wounds are licked. Rest and rumination help an animal survive.

When we disengage from a difficult situation, we are out of harm’s way and resting. That is a desirable state. We listen to sad music in order to reach that state. Sadness in real life has real causes – but not in music. We can savor the retreat and the rest without having suffered through a real cause. We may be in a situation – perhaps a situation we do not even consciously recognize – that is unpleasant or dangerous. We may therefore be feeling anxious or angry or lonely. We may be feeling frustration, shame, jealousy, exhaustion, desperation. Sad music is a way of dealing with these bad feelings, at least for a while. Sadness is the evolved response to loss, disappointment and failure. It is the sounding of Retreat. By evoking sadness, we buy the entire emotional package: admission of failure, retreat, relief, rest and rumination. That is the pleasure. A sad song will evoke that pleasure. Even a passing note or phrase in a larger work might touch some distant chord within the listener – a momentary resonance of sadness – perhaps literally an harmonic resonance in the nerves.

We would like to disengage -- not only from particular disappointments, such as the failure to win a desired mate -- but also from the many futilities of modern life. We would like to retreat and rest. Through sad music we invoke, we tap into, the biological process of sadness to do so. Although music is a recent development in the history of our species, perhaps no older than 60 or 70,000 years, “the emotional qualities of music are encoded in ancient subcortical regions of the brain that are foundational for our. . . nature” – so says the brilliant Jaak Panksepp. [5]. We turn to sad music as an imaginative disengagement, a respite and a shelter – not just from this or that trouble – but from the troublesome business of living in the modern world.

These conjectures are by no means inconsistent with the empirical results of Taruffi and Koelsch. Retreat is entirely consistent with nostalgia, empathy, and peace. (Grief, as opposed to sadness, is another matter.) Emotions are, at least in part, biological phenomena. "Nothing in biology makes sense except in the light of evolution." [6]


[1] Burton, Robert (1621). The Anatomy of Melancholy (Jackson, H., ed.). New York, NY: New York Review Books (2001). Part 2, p. 118.

[2] Keltner, D. and Gross, J. (1999). Functional accounts of emotions. Cognition and Emotion, 13(5), 467-480.

[3] Klinger, E. (1975). Consequence of commitment to and disengagement from incentives. Psychological Review, 82(1), 1-25, and many others.

[4] Hackenbracht, J. and Tamir, M. (2010). Preferences for sadness when eliciting help: Instrumental motives in sadness regulation. Motivation and Emotion, 34(3), 306-315.

[5] Panksepp, J. (2009). The emotional antecedents to the evolution of music and language. Musicae Scientiae, 13/2 supp., 229 – 259. p. 242.

[6] Dobzhansky, T. (1973). Nothing in biology makes sense except in the light of evolution. The American Biology Teacher, 35(3), 125-129.