Navigation – Plan du site

The "Sound of Power": Investigating Polyphone Actions and the Perception of Polyphony

Beatrice Kobow


Cet article aborde la polyphonie sous l’angle de la théorie de l’action et propose une comparaison entre deux domaines – la théorie de la musique et la théorie de l’action. La notion d’« action polyphone » se définit en tant que métaphore du phénomène de l’intentionnalité collective dans lequel elle décrit l’attitude des agents ainsi que leur action. Un modèle de « constructivisme pluriel » est suggéré comme conservant les prémisses de l’individualisme méthodique pour comprendre l’importance d’autrui dans la formation du soi. Il est aussi question ici de la normativité relative à la capacité à entendre la musique en général, et la polyphonie en particulier. La compréhension des structures polyphoniques reflète la capacité humaine à « faire ensemble » dans des cas proto-normatifs, mais également dans des cas déontologiques.

Haut de page

Texte intégral

1A musical piece is called polyphone if created by multiple voices each pursuing a melodic line, sounding in accord. Through interplay and coordination of these different voices the intended audible outcome is composed. If performing a polyphone piece can itself be a group action (involving more than one player), I want to see how far this notion of polyphony carries as a metaphor for group agency, in general. The concept of polyphony, I argue, can make a philosophical account of human agency, especially group agency, more comprehensive. I briefly sketch what could be regained for the theory of music, especially the investigation of the origins of musical polyphony, from an application of the term to action theory: the ability to hear polyphone sounds and “make meaning” of them points to the general ability of the human observer to focus her attention from the perception of a single voice to a perception of multiple voices, or from single agent-causation to group-actions. This I-mode/we-mode perception is implemented in the biological background. It is trained and developed normatively through the cultural background. Perception of polyphony thus already indicates our particular ability to impose norms collectively.

2In a theory of action, the intentional stance marks the difference between actions performed by different agents independently and different actions “adding up” to an overall group-action. Understanding collective intentionality in acting means understanding the right kind of causation due to the right kind of intention. I show how actions can be described as polyphone either by describing the intentional stance of the agents – the action is we-intended, or by characterizing the action as one that has to be performed by two or more agents. Taking into account these two possible descriptions, I stipulate two things about agency:

3– that understanding and self-imposing normativity on actions is based in an understanding of collective intentionality in general and

4– that linguistic agency (making meaning) is constituted by such an understanding, as well.

  • 1 I wish to only propose an idea about polyphone music as deeply ingrained in the make-up of human be (...)

5A discussion of polyphony clarifies how different forms of proto-normativity are constituted, such as commitment to procedures (finishing a course of action) and an expectation of brute fact normalcy, and how they relate to the general ability to hear music. If music is developed from a perception of the natural world, I suggest for an understanding of the origins of polyphony in music to look towards the “natural world” of human interaction as potentially polyphone in structure.1

Describing Group-Agency as Polyphone

6Actions can be understood as different from behavior in that the agent is free to perform the action or refrain from acting. Agency describes the intentional stance of the agent, not any physical movement. If there is a force binding the agent, we will not characterize her (albeit: causal) involvement in the event as agential. For example, if I open the window at gunpoint, or you raise my arm in an auction and I make a bid, we won’t say that I voluntarily opened the window or bought that painting. I won’t count as an agent in the full sense. A single agent’s participation in group efforts has to be agential in the same sense of being free, otherwise we won’t characterize her involvement in the group-action as full agency. Yet, what kind of intentions does the agent need to hold to perform in a group? How can she share intentions, or as prerequisite, join attention with others? Furthermore, looking at the (mostly unpleasant) side-effects of our ability to perform together, at accidental patterns and outcomes such as man-made draught or systemic discrimination, at intended actions with consequences for which no single agent seems to be responsible such as corporations polluting rivers, or at systems constituting group actions that may not be based on any sharing of intentions such as market capitalism, we see that a better understanding of agency in groups is necessary for a clarification of these cases and a fair attribution of responsibility.

  • 2 This suggestion is close to Tuomela’s distinction between I-mode, pro-group I-mode and pro-group we (...)
  • 3 I largely follow Searle and use his terminology in this description, though, later on, I wish to su (...)

7I suggest describing group-agency as polyphone, homophone or monophone.2 Introducing this terminology has the benefit and limitation of giving us access to the phenomenon first through a metaphor. Polyphone melodies consist of several “equal” voices constituting the piece. Relating the metaphor, if the group-members hold we-intentions that motivate their further intentions and bodily movements, this might constitute a polyphone action. For example, my playing my part of the Bach-Sonata in our string quartet could be described as polyphone action, because I hold the suitable we-intentions; my action is motivated by my intention that “we-play-the-Sonata”.3 An action would be monophone if performed alone and involving only I-intentions directed at my reaching my goal. For example, if I am alone in my room practicing my part of the Sonata. An action would be called homophone if some participants are only helping an agent reach her goal, where, again, only I-intentions are involved, this time some of them directed towards someone else reaching her goal. For example, if I practice my part and you help me by turning the pages of my sheet music, we both hold I-intentions, but yours are directed to assist my goal. That is, in cases of monophone or homophone actions there are no proper we-intentions motivating the action, the participants are not in “pro-group we-mode” (Tuomela). The polyphony described is on the level of the intentions motivating the action, not on the level of the concrete result of the action. Whether such we-intentions are present isn’t to be decided for all members of the group performing, but for each of them individually. Equally, it has to be decided for each instance of a potentially polyphone action (on token level). It could be that in some cases the agents hold we-intentions, but in some other cases the agents don’t. The same outcome, the same observable event in the world, might count as polyphone, monophone or homophone action depending on the intentions held by the agents. Consider as example three different versions of reaching the same outcome: If we play a Bach piece together, or if I play all voices of the polyphone piece myself one after the other for a recording, or if per chance you perform and I perform alone at home and the passer-by on the street hears Bach and stops to listen, each time the audible event will be polyphone on the level of music, yet, polyphone on the level of action only in the first case, monophone in the others.

8Alternatively, we could also think of polyphone actions as those that are only possible for several agents performing together, such as playing a symphony or a game of soccer. Then, polyphony describes the action itself, not the intentional states of the agents. Introducing polyphony as a term helps illuminate how the two descriptions are linked for the individual agent. Knowing that and how some actions are only possible with others is a prerequisite for we-intending and pro-group we mode. Actions possible only as group-efforts aren’t limited to games or actions with constitutive-rule-structures, nor actions constituting deontic rights-duty-obligations. Yet, in actions in games, following constitutive rules, or enforcing deontic structures it is most evident that the other participants aren’t simply used as means to an ultimately individualistic end. Polyphony then describes a) that actions are we-intended, b) that they are only possible as we-intended actions, and c) that the agents are aware of this fact.

9There are two main differences between the perception of polyphony in music and the analysis of actions as polyphone: first, an action originating as a group effort does not have to consist of several actions performed at the same time. We can make the plan to compose a symphony together where I will write the first movement and you will write the second and so on. We will take turns, this might take months at a time and no action is performed simultaneously with any other since they depend for their performance on the completion of the previous action. The we-intending of the single group members, not the simultaneity of several actions defines the specific difference of group-actions. A discussion of the concept of polyphony in music might reflect this point, though polyphony is a quality attributed musical pieces which are typically heard at one time.

10Secondly, any and all sounds surrounding a listener, let us say: the human ear, can be heard, that is, understood, as creating a polyphone multi-melodic musical piece together. Like all functions imposed by an observer, the discreteness of the object – in this case: the multi-layered melodic texture of sounds making up one overarching “melody” - is imposed by her as well. So, in a way, the perceived melody is “up to the listener” whereas whether an action is or isn’t a we-intended action is not up to the observer, but up to the agents performing.

11The metaphor of polyphony reaches its limits where polyphony as a quality of music can be part of a single composer’s strategy, but polyphony in action always characterizes we-intentional states and/or actions (necessarily) performed by more than one person.

  • 4 For reasons of simplicity, but also reasons of strictness of the concept of agency employed which i (...)

12To summarize: The concept of polyphony for action theory might help us understand how the intentions of the agents and the action itself can be characterized as polyphone. This, in turn, suggests that the agents themselves must exhibit an awareness of a certain form of normativity that is implied by polyphony in the sense used above to successfully perform polyphone actions. I will now use this stipulation to propose a way of accounting for agency that links up to this idea of polyphone actions, yet retains the premise that all actions, single or group actions, bottom out in individuals’ mental states, that there are nothing but individual mental states motivating actions (a premise sometimes referred to as “methodological individualism”4). Based in an understanding of polyphony as a normative awareness of agents, I suggest a normative account of group agency.

13There are three different ways of accounting for agency while maintaining the premise of methodological individualism. Among the methodological individualists, we find descriptive accounts of single agency as basic (such as Meggle’s), descriptive accounts of group agency as primitive (such as Searle’s) and normative accounts for single agents (such as Korsgaard’s). The descriptive account of single agency as basic denies the existence of we-intentions altogether and claims that all group-actions can be reconstructed as consisting of I-intentions together with mutual beliefs. In opposition, the descriptive account of group agency as primitive claims that we-intentions are not reducible to I-intentions with mutual beliefs while, conforming to the premise of methodological individualism, it states that we-intentions are in the heads of the individual agents performing together. Here, I will not discuss the merits and difficulties of these two accounts. I label both of them descriptive accounts because they stand in opposition to the existing normative accounts. They maintain that for the agent agency is like any other biologically rooted phenomenon, that agency is for the agent something like vision or digestion. Just as the agent sees due to corresponding visual states, the agent acts if she has the right corresponding intentional states. Normative accounts, in contrast, believe that someone counts as an agent not simply by means of the correct intentional motivational state, but that the agent has to self-constitute, that is, impose normative restrictions on her own behavior. The descriptive accounts assume that someone is an agent in acting, whereas normative accounts think of someone as becoming an agent in setting standards for her actions.

14Korsgaard shows a way from a Kantian, normative self-constitution to the world of others in a public sphere, but she assumes that the individual does not need the other in order to self-constitute in the first place. Existentialist thought originates from just this dilemma: that the self should be free to constitute itself, but that, at the same time, it is bound to the world of others for its very constitution (that is: not free). Sartre’s reflexive self, for example, needs the gaze of the other to grasp itself as existing and in order to put the two modes of the self (reflecting and existing) together, again, we need to, at least, imagine the other as present and as consciousness, as world-shaping agent. Inspired by these analyses and convinced of the relevance and truth of their observation, but aware of their status of remaining stuck between the rock of individual freedom and the hard place of our need for others for this freedom, I want to propose an account of group agency as normative. This entails keeping methodological individualism as a premise and understanding agency as constituted by the agent (much as Korsgaard’s agent) through self-imposed normative standards, but introducing the dependency of this constitution upon the realization of the importance of others for the understanding of normativity. If to perform polyphone actions the agent has to have a basic (possibly tacit) sense of group-originating normativity, a normative account of group agency is in order. Yet, doesn’t the agent just perform with others without ever thinking or reflecting about her need of these others for her self-constitution? Putting herself together as an agent is something that the agent does “naturally”, yet, upon closer inspection, we find that what appears natural is in fact part of a culturally learned and trained ability, one that was biologically given, but shaped through (cultural, linguistic) interaction with others. The concept of polyphony as the ability to perceive single voice and multi-voice melodies, single or multi-agent action reflects our self-constitution through others.

The Common Ground: Normativity and Linguistic Agency

15What kind of normativity or normative standards does the individual need to grasp in order to become an agent and how are these normative standards linked to collectivity? Any agent needs to be able to disambiguate contexts in order to act; this involves understanding contexts and also understanding some contexts as “normal”. Animals and humans alike are able to perceive a certain “brute fact normalcy”, and will be ill at ease in a situation that varies wildly from their expectation of normalcy. This expectation is quite independent of collective constitution. A solitary cat will feel it as well as a dog in a pack, as well as myself. Secondly, there is a proto-normativity that comes from a commitment to a course of action. It is implied not so much by the logic of going through with what I have begun (as Searle has suggested) where it would be too costly to abandon a course of action. Rather, it has to do with the logic of complying with the “right way of doing” what I have chosen to do. If I want to play piano, I typically use my fingers, not a bow, for example (though it is conceivable that I would do so). This procedural proto-normativity is independent of collective constitution, as well. There are forms of full agency that can be described independently of any form of collectively constituted normativity. In the case of a rupture in the expectation of brute fact normalcy, any of us (the cat, the dog, myself) might just run in a panic. Single agency is independent of collectively imposed norms where just procedural considerations are concerned. Yet, there is a third form of normativity that is dependent on collective constitution and awareness thereof. I will call this form “linguistic” normativity and the agency that it affords “linguistic” agency. It comes with understanding that some signs as naturally meaningful and that some signs as non-naturally meaningful, a distinction introduced into the literature by Grice. It is based in an understanding of what it means to communicate, to impose meaning on arbitrary signs to get someone to understand because they understand that this is what I am trying to do. I think we can tell the story of the perception of this difference from the point of view of the child who learns to see what her mother shows her (whereas her dog will never learn that). Equally, she learns what it means to go through with an action, but also learns what it means to act following rules because the action in question wouldn’t be possible without the rules. She and her dog will go on walks together, but they will never play chess together. She will also come to know that some actions require the help of others, some are only possible if performed together with others. Self-imposed normativity in acting is a result of understanding these different possible forms of acting and linguistic agency. It is dependant on understanding communicative structures and collectively imposed normativity.

Origins of Polyphony

16This essay is titled somewhat extraordinarily “the Sound of Power”. I will now give an explanation of this title. I have asked myself if music is meaningful for me in terms of a (non-linguistic) experience, or rather than meaningful for us, that is, conventionally meaningful; is it meaningful for me, as “music of the spheres”, or is it culturally coded and conditioned? Listeners testify to both. They appreciate music on a deeply personal level that goes beyond mediation to others. It is a solipsistic or private joy. But there is also evidence that what kind of music one likes depends on one’s culture, and furthermore, lovers of music appreciate experiencing music with others, as part of a tradition, and they judge it according to the norms of the canon that the composer belongs to.

17If I am able to perceive different sounds reaching my ear in progression as melodic, as somehow ordered “meaningfully” and as having a discrete form that distinguishes them from other sounds surrounding me, what kind of normativity does this entail? Is it the same normativity as the one that I understand in understanding non-natural signs as communicative devices of a speaker? Maybe the meaning of music is just “meaning for me”; maybe experiencing music is not only non-literal, but also not to be paraphrased in terms of common, conventional meanings, maybe it always touches on the sublime and relates us to the grandeur of the world (like a beautiful sunset, or an awesome thunderstorm) more than having a communicative aspect to it. The question of its communicative component applies not only to music, but to all works of art. It is the question of the conventional and author-intentional aspect of meaning in artworks. Walton, for example, claims that works of art don’t have to be intended by anyone to mean anything, but that the perceiver is the maker of the work of art in question. Walton doesn’t address the ability to hear music, but it could be a key example for him, since there are on the surface no transmittable “meanings” to sounds much like there isn’t meaning in this sense to numbers. It seems that not only the logical relationships between numbers, but also the harmonic relationships between musical notes are not simply communicative in nature. Music is not simply another symbolic code. According to Walton, seeing clouds as castles is just an exercise of our ability to pretend as-if. I stipulate that pretending as-if is itself an ability that results at least from our nature, if not from our self-knowledge as gregarious, communicative and deonticity creating animals. The as-if of children’s games is a mimesis of the structure of social facts. Music then would be an application of our ability to hear sounds as something else, for example, or as parts of a melody, and in both cases we would be imposing conditions of satisfaction upon already existing conditions of satisfaction (Searle). Music reflects the linguistic sphere in that it reflects our ability to impose meaning on sequences of sounds. But, what does this mean outside of the context of linguistic or conventional meaning?
Polyphone music mirrors our world, the world constituted with others through our doings together. It gives us – literally – a sense of power, because hearing single-voice melodies and polyphone melodies reflects our different abilities to act alone or with others. Other forms of acting together, like dancing together, or playing music together, can be seen as proto-forms of “linguistic” agency that involves the attribution of rights and obligations to the members of the performing group. The perception of polyphony, one could say, is an expression and the presupposition of our ability to act in concert.

Haut de page


BURGE, Tylor, Origins of Objectivity, Oxford, Oxford University Press, 2010.

GRICE, H.Paul, Studies in the Way of Words, Cambridge-Massachusetts, London, Harvard University Press, 1989.

KORSGAARD, Christine, Self-Constitution, Oxford, Oxford University Press, 2009.

MEGGLE, Georg, “Mutual Knowledge and Belief”, in MEGGLE, Georg (ed.), Social Facts & Collective Intentionality, Frankfurt am Main, Dr. Hänsel – Hohenhausen AG, 2002, p. 205-223.

MEGGLE, Georg (ed.), Social Facts & Collective Intentionality, Frankfurt am Main, Dr. Hänsel – Hohenhausen AG., 2002.

SARTRE, Jean-Paul, Being and Nothingness, translated and with an introduction by Hazel E. Barnes, New York, Washington Square Press, 1992.

SEARLE, John, Construction of Social Reality, New York, Free Press, 1995.

SEARLE, John, Rationality in Action, Cambridge-Massachusetts, MIT Press, 2001.

SEARLE, John, Making the Social World, Oxford, Oxford University Press, 2009.

TOMASELLO, Michael, The Cultural Origins of Human Cognition, Cambridge-Massachusetts, Harvard University Press, 1999.

TOMASELLO, Michael, Origins of Human Communication, Cambridge-Massachusetts, MIT Press, 2008.

TUOMELA, Raimo, The Importance of Us: A Philosophical Study of Basic Social Notions, Stanford, Stanford University Press, 1995.

WALTON, Kendal, Mimesis as Make-Believe – On the Foundation of the Representational Arts, Cambridge-Massachussetts, London, Harvard University Press, 1990.

Haut de page


1 I wish to only propose an idea about polyphone music as deeply ingrained in the make-up of human beings as gregarious and communicative animals, nothing about a historical type of polyphony in music or its cultural origins.

2 This suggestion is close to Tuomela’s distinction between I-mode, pro-group I-mode and pro-group we-mode. I think it might be worth discussing the term “polyphone” because it allows us the dual description of the agents’ involvement and the event described.

3 I largely follow Searle and use his terminology in this description, though, later on, I wish to suggest an account of group agency that is different from his, in that it accounts for a normative component that I believe is necessary for a comprehensive description of agency.

4 For reasons of simplicity, but also reasons of strictness of the concept of agency employed which is tied to a strict concept of what it means to act freely I will continue to support MI as premise. I hope to show that one can still have a comprehensive account of the world of the agent as socially constructed, that is, take into account the importance of others for the agent.

Haut de page

Pour citer cet article

Référence électronique

Beatrice Kobow, « The "Sound of Power": Investigating Polyphone Actions and the Perception of Polyphony », Transposition [En ligne], 1 | 2011, mis en ligne le 01 février 2011, consulté le 26 avril 2017. URL : ; DOI : 10.4000/transposition.137

Haut de page


Beatrice Kobow

Beatrice Kobow received her PhD in Philosophy from the University of Leipzig. She has been a research assistant to Prof. Georg Meggle in Leipzig and Prof. John Searle in Berkeley and lecturer at both universities. She was a fellow at the “Kulturwissenschaftliches Kolleg” at the University of Konstanz. Currently, she works as researcher at the CRAL/EHESS in Paris with a grant from the DAAD & MSH. In 2011, she will commence a Dilthey Fellowship of the Volkswagen Foundation at the University of Leipzig with her project “The Leap into Language - The Constitution of the Social Self between pre-intentional Background and self-reflexive Cogito”.

Haut de page

Droits d’auteur

© association Transposition. Musique et Sciences Sociales

Haut de page
  • Logo Centre de recherche sur les arts et le langage - CRAL
  • Logo Ecole des hautes études en sciences sociales - EHESS
  • Logo Philharmonie de Paris
  • Les cahiers de