Anton Killin, Australian National University
Abstract Stephen Davies defends an analysis of musical expressivity dubbed the ‘contour theory.’ In other work, Davies argues that hominin music could be as much as 500,000 years old. The musical expressivity debate is typically concerned with ‘pure’ (or ‘absolute’) music, taking examples from the Western art music canon as paradigmatic. I register some reservations about applying contour theory to musical expressivity in hominin prehistory, even though we might have reason to think that early musical activities were ‘pure’ insofar as they were wordless (e.g., utilising vocables, not lyrics) and not programmatic.
Over the last few decades Stephen Davies has contributed to a wide-ranging array of topics in the philosophy of the arts. More recently, Davies has been focused on explicating and assessing evolutionary and psychological theories about the arts, and associated empirical research, in addition to advancing his long-standing philosophical research agenda in parallel. The general question I put to Davies in this discussion piece is this: Has your more recent research on the evolution and cognition of the arts impacted your earlier philosophical theories and their application? To narrow the scope somewhat, my focus will be Davies’ contour theory of musical expressivity.
It is a commonplace observation that music induces affective responses in many of those who listen to it. And those who listen to music ascribe properties of an emotional kind to the music they hear, such as sadness. Despite this, music is not—and is not thought to be—sentient. Yet this situation has led to a curious philosophical puzzle. If the music is not itself sad, and is not capable of being sad, how does it express sadness? Philosophers have proposed a number of theories that aim to capture the nature of musical expressivity. This debate is typically concerned only with ‘pure’ (or ‘absolute’) music, taking examples from the Western art music canon as paradigmatic, in order to black-box any expressive effects due to lyrics or programme.
Davies has both motivated and defended the contour theory of musical expressivity. According to this view, musical expressivity is explained by appeal to the manifestation of qualities in instances of music that are characteristic of human emotions, “by virtue of resemblances between their own dynamic structures and behaviours or movements that, in humans, present emotion characteristics”. So an instance of music is expressive of sadness just in case that instance of music is disposed to elicit a response of sadness-ascription, regardless of the occurrence of emotions in the individuals involved, reached via musical resemblances with the manifestations of emotions in human beings. Andrew Kania lists these as “the phenomenology of the experience of the emotion, the emotion’s typical facial expression, the contour of vocal expression typical of a person experiencing the emotion, and the contour of bodily behaviour typical of such a person”. Here Davies has in mind such bodily behaviours as “gait, attitude, air, carriage, posture, and comportment”.
Musical sadness, on this theory, is thus like the sadness of weeping willows: the whispery, downturned structure of a weeping willow resembles that of a downcast, whiskery, hunched-over individual; or the sadness we ascribe to a droopy-looking canine face, given its resemblance to the dynamic characteristics of genuinely sad human faces (perhaps a furrowed brow with deep-set eyes).
Of course, Davies’ contour theory has its detractors. But in the following I take a different tack to that usually deployed against him. I query whether Davies’ account captures the full scope of music that he appeals to in more recent work. Even though Davies circumscribes the kind of music he intends the contour theory to be applicable to (so-called ‘pure’ music), extending the discussion of expressivity to music in an evolutionary context should not be too problematic (in principle), on the plausible assumption that much music in hominin prehistory was ‘pure’ in the sense that it was both wordless (e.g., utilising vocables) and not programmatic.
Davies argues that music may be as much as a half million years old—appearing and incrementally evolving in our Heidelbergensian predecessors. Let me quote Davies at some length:
- heidelbergensis possessed the physiological prerequisites for song: fine tongue and thoracic breath control, descended voice box and appropriate hyoid bone structure, hearing geared to detect and process the pitchbands in which the species vocalized, the neural resources to process and store patterned sound strings, and so on. Moreover, these hominins lived in social groups and depended on coordination, communication, and cooperation, so they had the social capacity to make group music and could benefit from doing so.
Music-making […] is often more about emotional expression and group entrainment and coordination than about abstract or symbolic thought. Individuals with mental deficits can be highly musical. Very young children can participate in group dancing and singing. Music-making is a practical skill that calls for “know-how” but need not require “knowing that,” the capacity to verbally cognize and articulate what is done. What matters, then, is not whether H. heidelbergensis qualified as what we would nowadays call an intellectual but whether she was inclined to vent her feelings in a musical fashion, perhaps interacting with her baby or while cooperating with her fellows. If her group celebrated their successes and mourned their losses, these ancients would have found applications for the musical capacities that they possessed.[… I]t is plausible to think that the earliest music itself was made as much as 500,000 years ago.
It is not my goal here to evaluate Davies’ evolutionary comments. Rather, notice the discussion about music as an expressive outlet. If we are to take seriously the affective nature of music in our evolutionary past, we might take more seriously the expression and arousal theories of musical expressivity. Heidelbergensian crooners and chanters, on this picture, presumably vocalised to one another in order to register and communicate some internal affective state, at least sometimes. In which case, it seems to me that the emotion expressed/aroused via the musical utterance is theirs, regardless of resemblances with other phenomena. Moreover, I think it plausible that capacities for emotional expression/arousal through music would confer adaptive value and be selected for, but perhaps not mere emotional recognition by contour, which might be thought of as a by-product of more general capacities for pattern recognition. The capacity for affective recognition by contour, indeed, might build upon more basic affective expression/arousal capacities as well as pattern recognition. And given that we have come to recognise expressive music in cases where an emotion is actually expressed, all this together does not immediately suggest that a contour-like theory would be correct. I offer the following ethnographic example. This is not intended to be a conclusive test case, yet it should aid our discussion of theories of musical expressivity.
As Inuit communities shifted from small-camp, nomadic-forager ways of life to larger-group, established-settlement societies in the twentieth century, individuals submitted to a sea change in their experience of the social realm. Music took on a clear social management role—extreme cases of hostility among adults was publically ritualised and dealt with through music in the form of song duels:
The song duel […] was resorted to in exceptionally troublesome cases, when the feelings of the antagonists ran too high to allow them to keep silent. In the duel, the two offended parties exchanged scathing songs while an amused audience looked on. […]
In short, the duel embedded conflict in an artistic form, isolated it within a ritualized context, concealed it behind irony and ambiguity of genre, and at the same time publicized it by focusing the attention of the entire community on it.
It seems to me that analysing instances of these song duels for expressivity along the lines of the contour theory would be misguided, and not just for the reason that they involve music with lyrics (i.e., they are not a case of pure music). Rather, song-duel music seems to be straightforwardly expressive of the emotional profile of the duellers. However, the contour theory requires abandoning “the attempt to analyse music’s expressiveness as depending on its connection to occurrent emotions”, and in this context at least, this approach seems to me to be a bug, not a feature, of Davies’ theory.
In my view, beginning at the beginning of music, some kind of hybrid expression-arousal theory of Pleistocene musical expressivity is not only salvageable but very plausible (the main objections to the expression and arousal theories appeal to possibilities that we can, it seems to me, cast aside in speculatively reconstructing Pleistocene music-making). Later developments in music evolution—for example, those that render music’s participatory status as one of trained specialists performing for passive audiences, who hear the performance as, and aesthetically evaluate it as a decoupled art object—utilise our capacities to intentionally express emotion through music and be emotionally stirred by music, and are thus candidates for contour theory. But not all later developments are plausible candidates for contour theory (e.g., Inuit song duels). In brief, then: expression-arousal first and foremost, in hominin musicality; plausibly, contour theory later, in certain contexts. This idea is not, of course, adequately developed or defended here, but I hope I have gone some way towards motivating it, and have given some reason to revisit Davies’ contour theory with a critical eye.
About the Author Anton Killin is a postdoctoral research fellow in the School of Philosophy and the Centre for the Dynamics of Language (CoeDL) at the Australian National University. His current research programme focuses on debates about the evolution of music and language, and especially the overlap between them. His other main research interests include the philosophy of biology, philosophy of music and aesthetics.
Briggs, Jean L. (2000). “Conflict Management in a Modern Inuit Community”. In: Hunters and Gatherers in the Modern World: Conflict, Resistance, and Self-Determination. Ed. P.P. Schweitzer,
- Biesele, and R.K. Hitchcock. New York: Berghahn, pp. 110–124.
Collingwood, R. G. (1938). The Principles of Art. Oxford: Oxford University Press.
Davies, Stephen. (1994). Musical Meaning and Expression. Ithaca: Cornell University Press.
—. (2001). “Philosophical Perspectives on Music’s Expressiveness”. In: Music and Emotion: Theory and Research. Ed. Patrick N. Juslin and John A. Sloboda. Oxford: Oxford University Press, pp. 23–44.
—. (2006). “Artistic Expression and the Hard Case of Pure Music”. In: Contemporary Debates in Aesthetics and the Philosophy of Art. Ed. by Mathew Kieran. Malden: Blackwell, pp. 179–191.
—. (2015). “How Ancient is Art?” Evental Aesthetics 4 (2), pp. 22–45.
Dewey, John. (1934). Art as Experience. New York: Minton, Balch.
Kania, Andrew. (2012). “The Philosophy of Music”. In: Stanford Encyclopedia of Philosophy, Fall 2012 edition. Ed. Edward N. Zalta. URL=
Killin, Anton. (2014). “Musicality in Human Evolution, Archaeology and Ethnography”. Biology & Philosophy 29 (4), pp. 597–609.
Levinson, Jerrold. (2006). “Musical Expressiveness as Hearability-as-Expression”. In: Contemporary Debates in Aesthetics and the Philosophy of Art. Ed. Mathew Kieran. Malden: Blackwell, pp. 192–206.
Matravers, Derek. (1998). Art and Emotion. Oxford: Clarendon.
Watt, Roger J., and Ash, Roisin L. (1998). “A Psychological Investigation of Meaning in Music”. Musicae Scientiae 2 (1), pp. 33–54.
 See e.g. Watt and Ash 1998.
 See e.g. Davies 1994, 2001, 2006.
 Davies 2001, p. 35.
 Kania 2012, §3.1.
 Davies 2006, p. 182.
 For Levinson’s variant, and objections to Davies’ specific account, see Levinson 2006.
 Davies 2015, p. 28.
 Ibid., pp. 28–29.
 Ibid., p. 30.
 See Killin 2014. Expression theorists hold that the emotions expressed in music belong to the composer(s), performer(s), or both; the emotions are behind or expressed through the composition or production of musical sound. Arousal theorists hold that the emotions expressed belong to the listeners of the music; they constitute the affective response evoked in listeners under standard perceptual conditions. Advocates of expression theory include Collingwood 1938 and Dewey 1934. Arousal theorists include Matravers 1998.
 In making this point I am not assuming that that individual’s affective state need be unobvious to other individuals until it is expressed in this way.
 Thanks to Ryan Doran for pressing me on this.
 See Briggs 2000.
 Ibid., p. 111.
 Ibid., p. 112.
 After all, a hypothetical wordless (e.g., vocable) version would serve my purposes, as would analysis of the actual music and musical emotion explicitly decoupled from that of the lyrics. Thanks to an anonymous referee for pressing me on this.
 Davies 2001, p. 34.
 Thanks to Ryan Doran and an anonymous referee for detailed and constructive feedback on previous versions of the manuscript. I have also benefitted from helpful discussions with Johanna Guest.