Navigation – Plan du site

Response to comments on “Postures of listening“

Victor A. Stoichita et Bernd Brabec de Mori


The authors of the initial proposal on “Postures of listening“ reply to the comments received from Jérôme Dokic, Robert S. Hatten, Tim Ingold, Michel Kreutzer and Elizabeth Tolbert.

Haut de page

Texte intégral

1We very much appreciate the comments and critiques that these five distinguished scholars from very different backgrounds have formulated for our debate, and we are most thankful that they share our interest in shedding more light on the rather obscure topic of what humans actually hear when they listen to the world. We are also grateful to Emmanuel de Vienne for his suggestion to organize the debate, and for his careful editing of the comments.

2There is little doubt that any human being endowed with functional ears can listen. Is that one big competence applied to various objects (voices, birds, storms, drums . . .), or are there several ways to listen? We started with the latter premise. Our inquiry led us to identify three modes of listening that appear to be available to all human beings with functional hearing, independently of their cultural background.

Are there several ways of listening?

3Scholars who are used to thinking in terms of cultural diversity are often led to emphasize that there are “more” ways of doing things than people in “the West” would think. In this debate, however, Ingold as an anthropologist is inclined to think that “all listening is enchanted” (albeit not in our terms). Tolbert, on the other hand, as an ethnomusicologist, advises that all listening should be understood as referential: the objects of our three listening postures “are not sound objects per se, but are referents of sonic signs”.

4With a specific focus on music, Hatten also finds that three listening modes are too many. He thinks it dubious that “any listener actually listens according to the authors’ strict structural posture” and also deems that enchantment “might be better addressed by an extended theory of virtual agency that encompasses both indexical and structural listening”. Dokic, too, finds it objectionable to distinguish between indexical and structural postures. He considers “clearly false” our argument that abstract structures are not proper auditory objects. He would rather hold that “phonemes are both abstract and audible, and that the auditory Gestalt that survives transposition is a bona fide object of auditory perception”. But he flips the argument and also considers “necessarily false” that enchanted beings could have what has just been granted to phonemes: proper auditory existence. Dokic then advises redefining enchantment as the superimposition of a “non-perceptual ontology onto the proper objects of auditory experience”.

5From the vantage point of ethology and neuroscience, Kreutzer easily grants that animal brains can analyse auditory sensations in several contrasted ways. Structural listening counts amongst them, and enchantment, possibly, too. Kreutzer asks, however, why we think that these processes should result in “alternatives” of listening. Could they not be combined, instead, in achieving humour comprehension, for instance? (A similar question is asked by Hatten with respect to the perception of “nuanced gestures” in music.) On the other hand, Kreutzer also believes that to encompass the attribution of states to animate and inanimate beings within one category (indexical listening) is to frame it too widely.

6This array of opposing ideas comforts our impression that, however futile it might seem, a coherent description of human ways of listening is still very much to be sought. Perhaps the biggest difference between our proposal and the positions held by our critics lies in how we deal with listeners’ choices.

Can we choose how we listen?

7Ingold, Tolbert, and Dokic express views on what listening is and what its (true) objects are. This stands opposed to our claim that distinct listening modes are available to any human being, and that each of us is able to choose amongst them. If listening is fundamentally one thing, our other accounts of it either are “historically specific” (Ingold on structural listening), or “rely heavily on Western aesthetic concepts” (Tolbert on enchanted listening), or attribute truth to narratives which are in fact “necessarily false” (Dokic on the same mode). Hatten is mainly concerned with how people listen to “music”. He notes that people have some choice in this respect when it comes to their modes of engagement with “virtual agency”. It does not seem, however, that they have the choice to disregard the virtual agent and not perceive it at all (to perceive “music” as something not “musical”). For Kreutzer, the problem of choice seems altogether irrelevant since he disputes that modes of listening are alternatives. Brains process auditory input in several parallel ways. Birds, for instance, can listen to identify conspecifics and to gain hedonistic pleasures at the same time. Why would they need to choose? Why would we?

8Let us suppose that we all agree on one thing: listening is an action. It is not just parsing sensory information according to some predefined rules. Such parsing exists, subconsciously, but when we refer to listening, we refer to something that happens in our consciousness. We actively focus our attention on specific auditory streams, we analyse them according to our specific needs or interests. Action means choice. Bergson once argued that from an evolutionary point of view, the main reason to develop a brain, rather than a mere spinal cord, was that a brain enables its possessor to choose: “The brain is like a switch having the faculty of putting the current [received from the senses] in communication with one or other of the motor mechanisms of the spine, chosen at will. . . . The brain is the organ of choice.” (Bergson 1911: 30).

9Since Bergson wrote this, neuroscience and cognitive psychology have considerably advanced our understanding of “brains”. It remains nevertheless true that a theory which claims to address how brains engage in the act of listening should probably leave some freedom of choice to the listener. Freedom does not mean that choices are entirely unpredictable, but simply that the listener can voluntarily deviate from what has been predicted. There is a “default” mode (indexical listening), and others that are triggered in specific circumstances. Quite often, there is just one way to listen “appropriately” in any given circumstance. What we say is that this is merely a culturally defined preference. It amounts to habits, not to rules. The listener remains free to “commute” intentionally from the culturally appropriate mode to another one.

10Dokic states that “[i]n many cases, we cannot help but hear sounds as belonging to given phonemes”. He is right indeed. But why only “in many cases”? Why not “always”? Possibly because common experience tells us that we can have a conversation with an individual while other people talk nearby (a condition nicknamed the “cocktail party effect” in psychoacoustics). Even when there is only one person speaking, structural listening is not automatically granted (teachers, for instance, know well that students in the classroom don’t necessarily parse their words for meaning). We agree with Dokic that under certain circumstances, which require specification, a person’s ability to choose how he or she listens – or not – to something can be diminished. We simply see no reason to place the starting point in constraint rather than in freedom. We should probably repeat here that our model is not about how sounds are processed in the brain (they can be processed in many ways at once), but about attention and awareness to sounds (which imply selective focusing).

Who can choose?

11Ingold criticizes our proposal on the grounds that “both the indexical and the structural modes seem to be artefacts of analytical moves that the authors make themselves”. Indeed, we are talking about analytical moves. But the real question is not whether we make them, but whether any other human being can make them, too.

12Similarly, if, in Tolbert’s words, our account of sonic beings “owes a debt to Western philosophical concepts such as Langer’s ‘forms of feeling’ . . . or Hanslick’s ‘sonically moving forms’”, then what exactly is at stake? Tolbert alludes to “problems of transcendence, sonic immediacy, dichotomy between form and content, and universalist aesthetics”. But firstly, if one looks at other musical traditions, it might prove surprisingly difficult to find those which do not insist on transcendence, do not attribute immediate effects to sounds, and do not take their aesthetics to have some kind of universal meaning. Secondly, we are well aware that generalizing from theories of Western music has been one of the pitfalls of early comparative musicology. This is why we refuse, for instance, to tweak the Western concept of “music” so as to make it seem universal (as in Small’s “musicking”, which Tolbert advises us to adopt). But once we set aside these considerations – the concept of “music” included – we may start to look at Hanslick’s and Langer’s proposals with milder ethnographic eyes. Instead of counter-examples for ethnomusicology students, they could also stand as pretty good indigenous theories. They show indeed what it means, for a certain class of Euro-American people, to listen in enchantment. The question then is really whether things like “sonically moving forms” (Hanslick), or “virtual” agents displaying “motivated actions” in “virtual” environments (Hatten) can appear to any human being. This is not asking whether the things themselves exist universally or transcendentally. It is asking whether all human beings have the ability to make such things stand out in their auditory experience. Is this ability a result of specific cultural backgrounds, or does it exist universally, as a shared competence of our species?

Who chooses to listen like that?

13Ingold opposes our depiction of structural listening on the grounds that “without writing . . ., the very idea of language, as a structured system of differences given independently and in advance of its realisation in speech, would be inconceivable”. The argument here is again misplaced, since our discussion has never been about the “very idea of language”. The question is whether people can recognize patterns in sound and link them to stable meanings, despite the fact that every voice sounds different. If they can do it, then they demonstrate structural listening. Incidentally, structural listening can also be applied to non-linguistic activities. Many “musical” traditions do not use writing at all, and still have explicit theories with rich vocabularies referring to abstract sound structures.

14Further on, Ingold seems also to question whether indexical listening is really practised in all societies. He asks: “[W]ho said the bird exists as an object, separate from its sound?” The quick answer would be: probably everyone. Just ask around whether a bird and its sound are different or not: the answer will probably be the same anywhere, even around Mount Bosavi where the Kaluli live. Ingold, however, thinks that in their world “every bird is its song . . . an active and affective presence in which any barrier between mind and world is effectively dissolved”. Feld’s ethnography plainly supports the idea that the Kaluli consider their birds as “active and affective presences”. As Jubi (one of Feld’s informants) summarizes, to him they are “voices in the forest”, not just “birds” (Feld 1990: 44–46). It does not follow, however, that the Kaluli can get by with no distinction between what the birds are and what the birds do. If it were true, as Ingold writes, that “every bird is its song” or that “the bird is its singing; it is not an object that sings”, then the Kaluli would be unable to distinguish between objects (birds) and events (their sounding). Their world would be just a phenomenal mess condemning them (and us) to a “surrender of understanding” (Descola 2016 in another debate with Ingold). But things stand otherwise. Feld’s ethnography shows that the Kaluli distinguish indeed not only between the “voices in the forest” and the beings from which these voices emanate, but also between various kinds of birds, some of which do not “sing” at all (they “make noise”, “whistle”, “weep”, “only sound”, etc.). Kaluli grammatics confirm it if needed: “bird” – ɔ̃bẽ – is a noun; “to sound” – ganalema – is a verb.

15That sounds occur is often overlooked in anthropology and ethnomusicology. This could be because they do not resemble what we usually hold to be an event. Scruton calls them “pure events” because they do not seem to happen “to” anything. In ordinary events, “physical objects undergo change: a car crash is something that happens to a car and to the people in it. But a sound is not a change in another thing, even if it is caused by such a change. Nor does anything participate in the sound in the way that the car participates in the crash” (Scruton 2009: 50). Sounds are, then, ontologically unstable. Enchantment is just one step away.

Can we truly hear inaudible things?

16Can people really choose to hear things that – by their own accounts – are inaudible? Or is it rather, as Dokic proposes, that they “hear something as something else”? This is a serious question. Narratives all over the world show that people frequently have “strong experiences with audition” (e.g. Gabrielsson 2011). A wide range of auditory beings is reported in these descriptions, from “mild” sensations of light, colour, or taste, to full-fledged sonic agents acting of their own will. But of course, anywhere in the world, people can be wrong. They can also use metaphors, humour, poetry, and many other figures of speech which should not be taken to mean literally.

17Before going further, let us take an example. In his Dancing Prophets: Musical Experience in Tumbuka Healing, Steven Friedson (1996) investigates how the Tumbuka in Malawi relate to various kinds of spirits through dancing and drumming. On one occasion, he ends up in front of the playing drums during a vimbuza session. He describes the feeling of what it means to be “danced by the drums”: “[W]hen I got my hips to synchronize with the basic six-pulse pattern of the drums, I felt as if the drums were moving my hips for me. Although the dancing was physically exhausting, it nevertheless seemed effortless while I was doing it” (Friedson 1996: 15). There is little doubt that Friedson did not intend to write in poetical, humorous, or otherwise metaphorical terms. He simply tried to convey a lived experience in an ethnology book.

18Observe that to convey this experience, he did not need to mention audition as such. We have to reconstruct it from encyclopaedic knowledge: drums, in fact, make sounds; they lack hands resting directly upon the dancer’s hips. That they have movement is far more difficult to elude. Being “danced by the drums” requires that the dancer truly feels a movement inside the drumming. This is the core signature of the narrated experience. Without that, the phenomenon is lost.

19Dokic is right in pointing out that what we term enchanted listening is the experience of “something as something else”. We also agree with him that this “typically requires more than purely auditory or more generally perceptual abilities”, and that the additional ability might be termed “imagination”. Dokic mentions how Wittgenstein touched upon this topic in his Philosophical Investigations, though without reaching a definite conclusion. Our proposal owes more to Sartre’s subsequent inquiry into the phenomenology of imagination (Sartre 1940).

20The relation of “perception” to “imagination” depends on the precise meaning given to these terms. One could hold, with Sartre, that seeing an elephant in a cloud or hearing a movement in a series of drum beats is a species of imagination, not of perception. The imagined object is plainly there, but it is simultaneously, and very consciously, unreal. It bears the mark of the counterintuitive, the magical, the supernatural, the simulacrum. Or else, in a distributed account of cognition, one could say that the imagined object arises from the superimposition of external and internal structures. An example on how people “see” constellations is given by Hutchins (1995: 68): The space in which the superimposition occurs can gain autonomy and function as a cosmological frame of reference, like the Oceanian sidereal compass, which literally encompasses the world. Both these ways of framing enchantment could help in understanding how it arises and what its ontological implications are. A deeper inquiry into the modes of existence of imaginary objects exceeds, however, the limits imposed on our current discussion. We can only agree with Dokic that an account which would attribute reality to sonic beings would be false. But denying their existence, too, would only take us from the frying pan into the fire, as it would contradict a wide range of empirical observations (no more “being danced by the drums”, only an illusion, or a figure of speech).

21Between perceiving and imagining, there is indeed some “conceptual tension”, as Dokic rightly comments. Actually, the tension is not just conceptual: all over the world, objects of imagination are treated as “special”, “magical”, “aesthetic” (really another kind of magic), “spiritual”, “demonic”, and so on. Maybe “virtual” is a better word for our own societies, as Hatten’s elaborate theory of musical agency suggests. We simply maintain that it takes a distinct auditory ontology to make such objects exist. They are unreal, stand “by themselves” in audition, and become “possessed” by things which are normally inaudible (see, e.g., Brabec de Mori 2015).

22Tolbert’s remark about ritual wailing is right in this respect: to many people at a funeral, “the lament is specifically not an instance of crying”. This is precisely why we mentioned it as an example of a faked index. We disagree, however, with Tolbert’s conclusion that this makes lamenting “an iconic performance of crying”. People at funerals are not supposed to engage in iconic performances. They do not describe proper laments as imitations of crying. To the Armenian Yezidi, for instance, a good lament is “true” (rast) or “sincere” (bî namus), even though it is notoriously difficult to understand the lament as a referential statement, and even when performed by professional mourners (Amy de la Bretèque 2013: 142; 2016: 50). Good laments bring the cries here and now. Rather than imitations, they are augmented instances of crying. Sadness and sorrow possess (cf. Sartre 1940) the lamenting voice if they are to move the audience to tears, just as movement possesses the drums’ sounds if these are to move the dancer’s hips.

Do we need to choose?

23Some of our commentators question whether choice is necessary at all. Why insist on “alternative” listening modes while some examples seem to demonstrate that they can be combined? Kreutzer argues that “when a message is humorous, the tone of voice often contradicts the verbal content. And yet these two aspects – semantic content, on the one hand, and “musical” intonation, on the other – are not analysed by the same nervous structures. Thus, in order to understand the message, we should not exclude one of these two modes of listening, but combine them instead.” Similarly, Hatten highlights the “importance of intonational and other non-verbal aspects of speech communication that can account for so much of a listener’s inferences”.

24Brains do many things at once and there is no reason to think that they treat auditory sensations in just one way at a time. But our inquiry starts with attention. It is quite clear that we can focus our attention on specific auditory streams. Other animals do it too, as Kreutzer’s commentary confirms. What does it mean to focus attention on an auditory stream? While one stream is focused, the others receive less attention, or none at all. Evidence for this can be found in a variety of experiments, especially those related to “change deafness” (we cited some examples in our FAQ). It has been repeatedly shown that when we focus our auditory attention, we tend to “miss” otherwise significant changes in what we send to the background. This fact alone shows that listening implies mutually exclusive choices. Background streams are still parsed, but their parsing is less efficient, and sometimes the result does not reach our consciousness at all.

25The case of humour could be different, however. To say things with a given tone of voice can convey distance towards the utterance, and distance towards the utterance is the general hallmark of irony (Sperber & Wilson 1981: 310; see also Anolli et al. 2002; Kumon-Nakamura et al. 1995). Is intonation, then, a good way to distinguish ironic utterances from non-ironic ones? One problem here is that irony (and humour in general) can be conveyed just as well in written form. Sound is not mandatory. Moreover, attempts to identify a specifically “ironic” tone of voice have led Bryant and Fox Tree (2005: 272) to conclude that “what seems like an ironic tone of voice is likely an emergent product of interpretations informed by multiple sources of information, many not acoustic”. In other words, we might be tempted to attribute more to voicing than it demonstrably holds. Voicing provides a “setting” for the ironic act. But to that effect, background processing could be enough.

26On a different tack, Hatten also argues that listeners need not necessarily choose between the modes which we deem alternative. When they listen to something as music, he writes, “Western listeners interpret nuanced gestures all the time, perhaps more often than they recognize ‘abstract’ musical structures”. All over the world, people are “capable of blending structure with gesture (and, by extension, expression) in a single listening mode”. But is this really a single listening mode? By the definition which we gave to this term, it should then have a definite kind of object, something with a specific kind of affordance.

27Hatten proposes that “basic inferences for music’s energies lead progressively to an actant that acts ‘in’ the music”. This is indeed a kind of being that affords quite specific relations. In Hatten’s terms, a “leap” occurs, after which the listener’s experiential world is different. Instead of enchanted, Hatten would prefer that listening be considered indexical throughout, albeit “virtualized” after the leap. To us, if a “virtual” realm brings forth a distinct set of objects, with distinct affordances, it manifests a distinct posture of listening.

28Nuances are good examples of that. In musicology, nuances are defined as categories of loudness, or variations thereof. A musicologist can say that the theme is first played pianissimo and then crescendo. This is often written on the score, and it can be perfectly heard, too. Now what if, instead of a nuanced gesture (pianissimo), the listener experiences quietness, as the property of an environment? Then physical gesture has been transformed into something else: a medium which is presently still, and which will soon be animated by the crescendo. It is probably not false to say that the listener is “capable of blending structure with gesture”. But if the resulting object is of a different kind than both structures and gestures (e.g. to the listener it is an “environment”), then we would rather address this as a specific posture of listening with a specific ontology.

Can other animals choose, too?

29Dokic, Hatten, and presumably Kreuzer agree to a certain extent that indexical listening, or the way of gathering ecologically relevant information with our ears, is the “default mode” or “our basic evolutionary response to any sound” (Hatten). Indeed, all animals with functional ears seem to be able to listen indexically – somehow connecting what they hear to their surroundings. In zoology, as in zoomusicology, however, it is unclear whether non-human animals are able to do what humans do when understanding speech, or appreciating musical variation.

30Kreutzer signals that birds engage in structural listening even though they have no language. Birds are sensitive to syntagmatic relations and recognize paradigmatic classes: “A bird can implement a number of different strategies for recognizing songs. If the phonology of the notes is altered too much, it will rely on their syntax; if the syntax is modified, it will rely on the phonology.” It is easy to see why the ability to recognize categories of sounds and auditory patterns proved advantageous in evolution, even for those animals that did not develop “language” in the human sense.

31What about enchanted listening then? Why would it be available to all human beings? And is it only available to them? Kreutzer’s comments shed an interesting light on this debate, and we are most thankful to him for bringing to our attention the existence of “hedonistic listening” amongst birds (Kreutzer 2014; Kreutzer & Aebischer 2015). That birds can experience pleasure in audition seems well established by now, and can be recognized by dopaminergic activation in their brain (Kreutzer 2014). “Hedonistic listening” could indeed be a path to what we call enchantment.

32On the other hand, auditory pleasure might be sufficiently well explained by structural listening alone. Kreutzer and Aebischer (2015) note that cognitive categories tend to have prototypes: members that are “better exemplars” of the category than others. Then, “if some males are better than others to be chosen as sexual partners, they can be considered to be prototypes. Moreover, if emotions and pleasure are experienced at the sight of these prototypes, then we are at the heart of the category ‘beautiful partners’. The ‘sense for beauty’ could work” (Kreutzer & Aebischer 2015: 275). The hypothesis is convincing, but if hedonistic listening works like this, then its “sense of beauty” is quite different from enchantment. Besides pleasure and beauty, many human traditions of auditory enchantment seek to procure feelings of sadness, tragedy, despair – all states of mind that in theory are quite antagonistic to the recompense circuitry of dopaminergic pleasure. Sonic beings need not be “beautiful", and can procure pleasure and discomfort just as readily. Hedonistic listening might have been an antecessor of enchantment, but the latter, as we witness it now amongst humans, is a realm where many kinds of relations and appreciations are possible. There would of course be more to say about the evolutionary relevance of enchantment. In an article elaborated subsequently to the present proposal, Brabec de Mori (2017) tackles the evolutionary implications mentioned by Tolbert (“why would such a thing [enchanted listening] have evolved in the first place?”). It would, however, go beyond the scope of the present debate to discuss it here in full.

33With that, we would like to heartily thank all the commentators for the work and thoughts they have invested in this debate. We are very confident that future inquiry, discussion, and especially interdisciplinary interaction and dialogue will lead us to a sharper picture of how we construct our auditory realms. We also thank the editors of Terrain for allowing space for this debate and hope for an ongoing engagement with the world and with listening in particular as a subject of scholarly thought.

Haut de page


Paroles mélodisées. Récits épiques et lamentations chez les Yézidis d’Arménie, Paris, Classiques Garnier.

–, 2016.
“Self-Sacrifice, Womanhood, and Melodized Speech. Three Case Studies from the Caucasus and Anatolia”, Asian Music no. 47/1, pp. 29–63.

“From ‘Blame by Praise’ to ‘Praise by Blame’. Analysis of Vocal Patterns in Ironic Communication”, International Journal of Psychology no. 37, pp. 266–276.

“Life and Consciousness”, The Hibbert Journal no. 10, pp. 24–44.

“Sonic Substances and Silent Sounds. An Auditory Anthropology of Ritual Songs”, Tipití. Journal of the Society of the Anthropology of Lowland South America no. 13/2, pp. 25–43.

–, 2017.
“Musical Spirits and Powerful Voices. On the Origins of Song”, Yearbook for Traditional Music no. 49, pp. 114–128.

“Is There an Ironic Tone of Voice?”, Language and Speech no. 48/3, pp. 257–277.

“Biolatry. A Surrender of Understanding (Response to Ingold’s ‘A Naturalist Abroad in the Museum of Ontology’)”, Anthropological Forum no. 26/3, pp. 321–328.

Sound and Sentiment. Birds, Weeping, Poetics, and Song in Kaluli Expression, Philadelphia, University of Pennsylvania Press [2nd ed.].

Dancing Prophets. Musical Experience in Tumbuka Healing, Chicago, University of Chicago Press.

Strong Experiences with Music, Oxford, Oxford University Press.

Cognition in the Wild, Cambridge, MA/London, MIT Press.

“Un demi siècle de chants d’oiseaux”, in Martine Clouzot & Corinne Beck (eds.), Les oiseaux chanteurs. Sciences, pratiques sociales et représentations dans les sociétés et le temps long, Dijon, Éditions universitaires de Dijon, pp. 25–45.

“The Riddle of Attractiveness. Looking for an ‘Aesthetic Sense’ within the Hedonic Mind of the Beholders”, in Thierry Hoquet (ed.), Current Perspectives on Sexual Selection. What’s Left After Darwin?, Dordrecht, Springer Netherlands, pp. 263–287.

“How About Another Piece of Pie. The Allusional Pretense Theory of Discourse Irony”, Journal of Experimental Psychology. General no. 124/1, pp. 3–21.

L’imaginaire. Psychologie phénoménologique de l’imagination, Paris, Gallimard.

“Sounds as Secondary Objects and Pure Events”, in Matthew Nudds & Casey O’Callaghan (eds.), Sounds and Perception. New Philosophical Essays, Oxford, Oxford University Press, pp. 50–68.

“Irony and the Use–Mention Distinction”, in Peter Cole (ed.), Radical Pragmatics, New York, Academic Press, pp. 295–318.

Haut de page

Pour citer cet article

Référence électronique

Victor A. Stoichita et Bernd Brabec de Mori, « Response to comments on “Postures of listening“ », Terrain [En ligne], Lectures et débats, mis en ligne le 29 novembre 2018, consulté le 16 décembre 2018. URL :

Haut de page


Victor A. Stoichita

Articles du même auteur

Bernd Brabec de Mori

Articles du même auteur

Haut de page
  • Logo CNRS
  • Logo Université Paris Lumière
  • Logo Ministère de la Culture
  • Logo Université Paris Ouest
  • Logo MAE - Maison Archéologie Ethnologie
  • Logo FMSH-Diffusion
  • OpenEdition Journals