• 1  Tom Johnson, The Voice of new music (New York City 1972-1982, A collection of articles originally (...)

Perhaps the most fascinating thing about this kind of one-note music is not the little changes which the composer causes in the sound, but the additional ones which somehow seem to happen in one’s ear. Sometimes I found it hard [sic] to focus on some aspect of the sound which I knew, simply by the logic of what was going on, was acoustically present. Sometimes it was hard to tell if one thing was getting softer or if another thing was getting louder. Sometimes one of the overtones would change its color and it would sound like everything else was changing too. And often I could tell that something was changing, but would not be able to hear exactly what it was. Of course, it could be that as my ears become more accustomed to hearing music like this they will stop playing these weird tricks on me. But I have the feeling that illusions of this nature will always be created when pure electronic tones are sustained in a context of this nature. We know quite a bit about optical illusions, but I have never heard the term “aural illusions.” For some reason no one seems very interested in these phenomena, but they certainly exist, at least in electronic pieces like this one. Maybe someday there will be a musical equivalent of Josef Albers, who will work out a theory to explain how they happen. In the meantime, we can just listen and be pleasantly mystified.
Tom Johnson, 17 August 1972, about Rhys Chatham’s Narcissus Descending.1

• 2  François-Xavier Féron, Des illusions auditives aux singularités du son et de la perception : l’imp (...)

2We will first deal with the polysemic nature of the term “illusion,” before we present the bases of our classification method. To do so, we will build on investigations carried out as part of our doctoral thesis, which aimed to gather a great variety of these unusual phenomena, so as to describe their aural properties, explain them scientifically and illustrate them musically.2

• 3  George A. Miller and Joseph C. R. Licklider, “The intelligibility of interrupted speech,” Journal (...)
• 4  Richard M. Warren and Roslyn P. Warren, “Auditory Illusions and Confusions,” Scientific American, (...)
• 5  To learn more on Warren’s researches and listen to sound samples, please refer to: Richard M. Warr (...)
• 6  Jean-Claude Risset, “Composer le son : expériences avec l’ordinateur, 1964-1989,” Contrechamps, no (...)
• 7  Diana Deutsch, “An auditory illusion,” Nature, no. 251, 1974, p. 307-309.
• 8  Diana Deutsch, Musical Illusions and Paradoxes [CD], Philomel Records, 001, 1995. Diana Deutsch, P (...)
• 9  To learn more on Albert Bregman’s researches and listen to sound samples, see: Albert S. Bregman, (...)
• 10  David L. Wessel and Jean-Claude Risset, “Les illusions auditives,” in Claude Malric (ed.), Musique (...)

4As compared to optical illusions, which have been studied for centuries, auditory illusions have emerged more recently. Their elaboration can originally be connected to the development of sound reproduction techniques, beginning with Thomas Edison’s invention of the phonograph in 1877. They later thrived upon Max Mathews’ 1957 development of computer sound synthesis. It is only in the 20th century, and especially in its second half, with the emergence of magnetic tape recording, that auditory illusions began to be scientifically studied by several psychologists. In the early 1950s, based on a speech that had been sectioned a dozen times per second, George Miller and Joseph Licklider realized that the speech is once again intelligible and seems continuous when loud noises are added into the gaps.3 Authors do not describe this phenomenon in terms of illusion, but as the “PicketFence” effect, which will later be given new labels: barrier, tunnel, continuity effect... In the 1960s, Richard Warren highlights the verbal transformation phenomena: when syllables or words are quickly repeated in a loop, listeners perceive semantic changes, effects that are related to the variability of perception under conditions of invariable sensory stimulation.4 Throughout his entire career, Warren focus remained on phenomena of an illusory nature, especially the continuity and completion effects.5 The creation of computer-assisted sound synthesis marks a turning point in the history of auditory illusions. The analytical works carried out in the Bell laboratories, at the instigation of Max Mathews, have revealed that certain remarkable properties of audition could be used in order to produce extremely confusing sound effects. As explained by Jean-Claude Risset, synthesis made it possible to “facilitate the emergence of illusions by building sound structures specially designed to play on certain perceptive mechanisms.”6 In this way, genuinely incredible sounds are created throughout the 1960s, such as Roger Shepard’s endless scale and Risset’s infinite or paradoxical glissandi, which we will further examine in the third section of this article. In 1974, Diana Deutsch discovers a truly surprising spatial illusion by alternating two sounds separated by an octave at a fast tempo: when the low sound appears on the left, the high sound simultaneously appears on the right and vice versa. The vast majority of right-handed listeners then locate all the high sounds on the right and all the low ones on the left.7 Like Warren, Deutsch will never cease to study these curious sound effects, which she released to the general public via two CDs.8 It is also in the 1970s that Albert Bregman begins his research on the analysis of auditory scenes and especially on the creation of auditory streams, a phenomenon that is at the root of numerous trompe-l’oreille effects.9 This brief overview of the researches on auditory illusions and related phenomena does not aim at comprehensiveness, but rather highlights the increasing interest they spark in the scientific community. It is in this context that the French National Audiovisual Institute (INA), Radio France and the group of French-speaking acousticians came together to discuss this topic in 1977. Two years later, auditory illusions are referenced in the Encyclopedia Universalis for the first time, in an article written by David Wessel and Jean-Claude Risset.10

• 12  Steve McAdams, “L’utopie d’une psychoacoustique musicale,” Les Cahiers de l’IRCAM – Recherche et M (...)
• 13  David L. Wessel and Jean-Claude Risset, “Les illusions auditives,” in Claude Malric (ed.), Musique (...)
• 14  Jean-Claude Risset, “Illusions sonores et musicales,” Composer le son. Repères d’une exploration d (...)
• 15  Jean-Claude Risset, “Illusions sonores et musicales,” Composer le son. Repères d’une exploration d (...)

5Therefore, illusion reflects discord, insofar as it does not relate to any objective reality. But what is this “objective” reality, exactly? Shall we refer only to the physical properties of the acoustic signal or also take the properties of auditory perception into account? Everything would be an illusion, if we should not take the psychoacoustic reality into account, i.e. “the set of strict and highly formalized experimental paradigms which aim to establish a quantitative link between the parameters of the acoustic world and those of the listening experience.”12 Besides, as Wessel and Risset explained, one must not reduce illusions to misinterpretations; they should instead be considered with Purkinje’s famous phrase in mind as, “a deception of the senses, truths of perception.”13 In addition to the idea of a certain discordance with reality, of disjunction between what should “normally” be perceived and what is “really” perceived, illusion also refers to an “appearance void of reality.”14 For instance, the sound reproduction techniques allow to create the illusion of a presence, by having us believe that a certain source is physically present, while the sound signal has “no visual compensation or vibratory causality.”15 Therefore, defining illusion is not simple, though the notion of deception is a constant. Let us briefly get back to how auditory illusions used to be listed, in order to appreciate their polysemic nature.

• 17  John Pierce, The Science of musical sound, San Francisco, W. H. Freeman and Company, 1983.
• 18  Jean-Claude Risset, “Illusions sonores et musicales,” Composer le son. Repères d’une exploration d (...)
• 19  Jacques Ninio, La Science des illusions, Paris, Odile Jacob, 1998.

13Wessel and Risset list and describe a certain number of illusory effects and classify them, either according to the procedures of perception, or according to the sound or musical parameters in play. Some of these effects reveal natural mechanisms of auditory perception and are perfectly well-rooted in our listening habits; while others are artificially constructed and present themselves as impossible sound “objects.” The term “illusion” encompasses here a wide array of phenomena, most of which are described in John Pierce’s The Science of Musical Sound17 (backed by numerous sound examples), as well as in another article by Risset.18 In his book La Science des illusions,19 Biologist Jacques Ninio lays the foundations for a classification method, drawing inspiration from Richard Gregory’s researches, who is a leading expert in optical illusions. Ninio proposes to classify visual and auditory illusions in accordance to the great procedures of perception (perceptive thresholds, contrasts, segregations/fusions, completion, adaptation, consistencies, points of reference/localizations, arbitration) rather than in compliance with the parameters that are prone to misinterpretations. The typology we have developed draws on these investigations, while striving to encompass every sound phenomenon that could be considered, in one way or another, as unusual.

• 20  Jacques Ninio, La Science des illusions, Paris, Odile Jacob, 1998, p. 13.

14How can we determine whether a phenomenon is unusual or not? What would be the definition of an unusual phenomenon? By definition, what is unusual arouses astonishment and surprise through its out of the ordinary nature, which goes against custom and rules. However, each one has their own knowledge and points of reference, so that what is unusual to one may not be to the other. Besides, what used to be unusual in the past may not be today. As Ninio notes: “We used to believe, on the basis of the information collected by our sensory organs, that things were that way, while culture and reason have made us aware that they are different.”20 In this way, the phenomena regarded as illusory by Lucretius (echo, reflection in a mirror...) are now included in the legitimate representations of reality. The typology that we aim to develop shall remain open, so that every phenomenon that seems or would seem somewhat unusual might be included. We are fully aware of the utterly subjective nature of such a premise, but it allows us to lay the foundations for a generic classification method that would include a great variety of known phenomena, but also ones that are yet unknown. The changes that occur in the ear, to which Johnson refers when dealing with Chatham’s music, are not necessarily illusions. Nonetheless they reflect remarkable sound specificities (such as differential sounds, beat phenomena, resonance phenomena) intentionally emphasized by the compositor who wishes to surprise the listeners, and offer them a singular experience. These phenomena cannot be mistaken with illusions, since they do not reveal misinterpretations. However, we must find a way to include them in our typology. In order to better understand the development of this typology, let us take a concrete example: overtone singing, a truly unusual sound phenomenon which may be a source of illusions to the novice’s ear.

• 21  CD edited by Hugo Zemp, Les Voix du monde. Une anthologie des expressions vocales, Le chant du mon (...)

fig. 1

Agrandir Original (jpeg, 1,2M)

Overtone singing xöömij (Mongolia, Ulan Bator), spectrogram (0-4.500 Hz) produced with the Audiosculpt software (Hanning window, 6.000 samples).21

• 22  Jens Blauert, Spatial Hearing. The Psychophysics of Human Sound Localization [Räumliches Hören, St (...)

17Drawing directly on German engineer Jens Blauert’s reflections,22 who makes the distinction between sound events pertaining to the field of physics, and auditory events relating to the perception of sound, we have chosen to dissociate two types of sound phenomena:

• 24  Lionel S. Penrose and Roger Penrose, “Impossible object, a special type of visual illusion,” Briti (...)

20By nature, illusion tends to deceive the listener in any way and often implies a true misinterpretation, or in other words a discordance between what should be and what is perceived. Computer-assisted sound synthesis offers full control over sound and allows for the creation of truly remarkable illusions of infinity, comparable to the famous endless staircase designed in 1958 by Lionel Penrose and his son Roger Penrose24 (fig. 3).

• 25  François-Xavier Féron, “L’art du trompe-l’oreille rythmique,” Intermédialités – Histoire et théori (...)

fig. 3

Agrandir Original (jpeg, 324k)

Revisited Penrose staircase.

© François-Xavier Féron and Karen Brunel-Lafargue, 2010.25

• 26  Roger N. Shepard, “Circularity in judgements of relative pitch,” Journal of the Acoustical Society (...)
• 27  Jean-Claude Risset, “Paradoxes de hauteur,” IRCAM Research reports, no. 10, Paris, IRCAM-Centre Po (...)
• 28  David L. Wessel and Jean-Claude Risset, “Les illusions auditives,” in Claude Malric (ed.), Musique (...)
• 29  Douglas R. Hofstadter, Gödel, Escher, Bach. Les brins d’une guirlande éternelle [An Eternal Golden (...)

21In 1964, Shepard synthesizes a series of twelve sounds (consisting only of octave-separated components). Once set in a loop, they constitute an endless chromatic scale that seems to constantly ascend in pitch (or descend depending on the direction of the notes).26 In 1968, Risset performs an endless glissando by carefully proportioning the spectral envelope, so that the central components are strengthened, while the low and high components are gradually introduced or eliminated.27 This glissando is featured in his work Computer Suite for Little Boy (1968), the second movement of which entitled “Fall” symbolizes the dropping of atomic bomb Little Boy on Hiroshima by pilot Eatherly and the latter’s subsequent collapse. During the first part of this extract (fig. 4), up until the 70th second approximately, the drop is suggested by both an endless glissando going slowly in the low (intertwining of hyperbolic curves in the 0-1.000 Hz area) and a dozen of fast and isolated glissandi (almost vertically descending lines). In the second part, the slower endless glissando is superimposed on Shepard’s endless chromatic scale, the spectral components of which are mixed up with the lines of the glissando going low yet appear clearly in the form of dots beyond 500 Hz. As stated by Wessel and Risset, there is illusion when the perceived height is assimilated to the physical frequency of the sound : the latter gives the impression that it endlessly descends, which is impossible without the sound going beyond the hearing range.28 However, although we know that pitch depends in fact on two dimensions—pitch class and pitch height—rather than the physical frequency, the effect can still be regarded as an illusion since we are confronted with what Douglas Hofstadter calls a magical loop29: instead of perceiving a cyclic structure that constantly goes back to its starting point, we perceive a linear endlessly motion.

• 30  Diana Deutsch, “A musical paradox,” Music Perception, vol. 3, no. 3, 1986, p. 275-280.

22On the one hand, the term paradox refers to a thought or opinion that goes against the common view and on the other hand, it designates an antinomy, or in other words a contradiction between two ideas. The tritone paradox discovered and studied by Deutsch is aptly named, as it reflects the variability of perception depending on listeners rather than a misinterpretation. The same tritone, designed according to Shepard’s sound principle, is prone to contradictory interpretations. It seems ascending to some listeners and descending to others.30 By separately controlling pitch class and pitch height, Risset has managed to produce another kind of paradox: a pitch that simultaneously ascends and descends, the paradoxical nature of which makes it a particularly unusual effect. In the third movement of Computer Suite for Little Boy (1968) entitled “Contre-Apothéose,” the composer uses an ascending glissando that paradoxically ends up lower-pitched than it started. In Mutations (1969), he reverses the process (fig. 5). Once an essentially low-pitched endless ascending glissando performed (320-382 s), he extends it to the whole spectrum (382-450 s). At this moment, the endless glissando turns into a paradoxical glissando: the spectrogram clearly shows that the pitch class keeps on “ascending” (the spectral components still draw ascending lines), while the pitch height simultaneously “descends” (the energy is shifted to the low spectral components). As a result, the tone ascends in pitch; yet, at the end, it is lower than at the beginning.

• 31  Luigi Russolo, L’Art des bruits [1916], edition established and presented by Giovanni Lista, Lausa (...)
• 32  We are here especially referring to Jean-Claude Risset, John Chowning, James Tenney, György Ligeti (...)

24Unusual sound phenomena have long enthralled composers, who have sought to replicate them before some of them placed them, during the 20th and 21st centuries, at the heart of their composition process. Attempting to mystify listeners, like Rhys Chatham seems to do in Narcissus Descending, is a trend in many currents of music, be it composed or improvised, past or contemporary, art or popular music. “First make the senses vibrate and the heart will follow! Thrill people through the unexpected, the mysterious, the unknown, in order to truly, intensely and deeply move the soul!” wrote Luigi Russolo in 1916 in his manifesto entitled L’Art des bruits.31 Unusual phenomena should be thus considered, whether physical (acoustic phenomenon) or perceptive (auditory phenomenon), and whether they pertain to singularity or “special” effects. Then, it is no surprise that more and more artists strive to explore them, regardless of genres and aesthetics.32