On Sonic Art

Contemporary Music Studies

A series of books edited by Peter Nelson and Nigel Osborne, University of Edinburgh, UK

Volume 1
Charles Koechlin (1867-1950): His Life and Works
Robert Orledge

Volume 2
Pierre Boulez—A World of Harmony
Lev Koblyakov

Volume 3
Bruno Maderna
Raymond Fearn

Volume 4
What's the Matter with Today's Experimental Music? Organized Sound Too Rarely Heard
Leigh handy

Volume 5
Linguistics and Semiotics in Music
Raymond Monelle

Volume 6
Music, Myth and Nature, or The Dolphins of Arion
François-Bernard Mâche

Volume 7
The Tone Clock
Peter Schat

Volume 8
Edison Denisov
Yuri Kholopov and Valeria Tsenova

Volume 9
Hanns Eisler—A Miscellany
Edited by David Blake

Volume 10
Brian Ferneyhough—Collected Writings
Edited by James Bows and Richard Toop
 On Sonic Art

by
Trevor Wishart
A new and revised edition

Edited by
Simon Emmerson
Copyright © 1996 OPA (Overseas Publishers Association) N.V. Published
by license under the Harwood Academic Publishers imprint, part of The
Gordon and Breach Publishing Group.

All rights reserved.

First published 1996 by Gordon and Breach

Reprinted in 2002
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

Transferred to Digital Printing 2005

Routledge is an imprint of the Taylor & Francis Group

Transferred to Digital Printing 2002

No part of this book may be reproduced or utilized in any form or by any

means, electronic or mechanical, including photocopying and recording, or
by any information storage or retrieval system, without permission in
writing from the publisher.
British Library Cataloguing in Publication Data

Wishart, Trevor
On Sonic Art, – New and Rev. ed. –
(Contemporary music studies; V. 12)
1. Computer composition 2. Computer music
3. Music – Philosophy and aesthetics
I. Title II. Series III. Emmerson, Simon
781.3’4

ISBN 3-7186-5847-X (paperback)

ISBN 3-7186-5848-8 (CD)
 CONTENTS

Introduction to the Series

Editor's Introduction
Preface
Acknowledgements

PRELUDE
Chapter 1 What is Sonic Art?

PART 1: THE SONIC CONTINUUM

Chapter 2 Beyond the pitch/duration paradigm
Chapter 3 Pythagoras, Fourier, Helmholtz: towards a phenomenology of
sound
Chapter 4 The nature of sonic space
Chapter 5 Sound structures in the continuum
Chapter 6 Gesture and counterpoint

PART 2: LANDSCAPE
Chapter 7 Sound landscape
Chapter 8 Sound-image as metaphor: music and myth
Chapter 9 Is there a natural morphology of sounds?
Chapter 10 Spatial motion

PART 3: UTTERANCE
Chapter 11 Utterance
Chapter 12 The human repertoire
Chapter 13 Phonemic objects
Chapter 14 Language stream and paralanguage
Chapter 15 The group

CODA
Chapter 16 Beyond the instrument: sound models

Bibliography
Music Examples
Music References
Index
 INTRODUCTION TO THE SERIES

The rapid expansion and diversification of contemporary music is explored

in this international series of books for contemporary musicians. Leading
experts and practitioners present composition today in all aspects—its
techniques, aesthetics and technology, and its relationships with other
disciplines and currents of thought—as well as using the series to
communicate actual musical materials.
The series also features monographs on significant twentieth-century
composers not extensively documented in the existing literature.
Nigel Osborne
 EDITOR'S INTRODUCTION

It is a dilemma for Trevor Wishart that so many people appreciate and

demand his writings rather than just simply listen to the music. But luckily
for us, providing the music is not compromised, he has consistently agreed
to put his thoughts to paper. In this book there is an advocacy and an
evangelism which is characteristically direct and uncompromising. On
Sonic Art is a demand for a renewal, not just a personal credo. Also—
notwithstanding a forthright statement on the very first page that the book is
about the ‘why’ and not the ‘how’ of sonic art—there are clear insights into
methods and means which will help attain these aims.
It was sometime in 1993 that Trevor Wishart said in an aside that, due
to work on a new (self-published) book,1 he was having difficulty financing
and organizing a reprint of On Sonic Art. By good chance I was due to
discuss another project with Robert Robertson of Harwood Academic
Publishers and added the suggestion of a newly edited edition of this work
to the agenda. To my surprise and delight, Robert had known the book for
some years and accepted the project with enthusiasm. He was, in fact, one
of an enormous group throughout the world for whom this text had been an
inspiration, a stimulus and a constant reference.
The problems of editing were several. I wanted to preserve Trevor
Wishart's idiosyncratic style and have only altered to clarify or correct. I
decided that with a few minor exceptions the ‘datedness’ of some of the text
was a strength. There are elements of the book that remain ‘1985’; to have
updated these would have been difficult without major rewriting. By that
time, while PCs and mainframes may yet have remained substantially
separated and a great deal less powerful, the landscape we have today (of
music software tools, at least) was largely formed. But in fact, and more
importantly, many critical points which Wishart made remain unaddressed
more than a decade later. A very high proportion of the book remains as
relevant now as then, while the remainder may be seen in clear perspective
as a historical document.2
I intend the suggestions on the production of sound and music
examples to represent a challenge to the reader, the listener and the teacher.
The recording accompanying this book was produced by Trevor Wishart
and myself and is largely of otherwise unavailable material. But
increasingly the expanded ideal list of examples will be available through
the development of on-line facilities: On Sonic Art suggests—even foretells
—the development of this new resource. It is a key text for the aurality of
the network.
Simon Emmerson
London, 1995

1 Audible Design (A Plain and Easy Introduction to Practical Sound Composition). York: Orpheus
the Pantomime Ltd., 1995.
2 For this reason added footnotes only reference more recent developments where this helps clarify a
point.
 PREFACE1

This book was written in a period of six weeks whilst in residence as the
Queen's Quest Visiting Scholar in the Music Department of Queen's
University, Kingston, Ontario. It is a very much expanded version of a
series of lectures given in the Department on the subject of electronic
music, though in fact it ranges over a field much wider than that normally
encompassed by this term. The book grows out of my own musical
experience over the past twenty years. Some of the ideas which I had been
developing were fully confirmed during my experience of the course in
computer music at the Institut de Recherche et Coordination
Acoustique/Musique (IRCAM) in Paris during the summer of 1981.
I would particularly like to thank the Queen's University Music
Department for their generous invitation to me which made the writing of
this book possible. In addition I am particularly indebted to Professor Istvan
Anhalt and to Jean-Paul Curtay for the sharing of ideas and insights into the
field of human utterance. I am also indebted to the Yorkshire Electro-
Acoustic Composers’ Group (previously the York Electronic Studio
Composers’ Group) for the various debates and discussions on musical
aesthetics in which I have been involved over the years.
In a way this book has grown out of a profound disagreement with
my friend and fellow-composer, Tom Endrich, whose very thorough
aesthetic research is founded primarily upon the properties of pitch and
duration organization in different musical styles, both within the Western
tradition and from different musical cultures. I hope that this book will
present a rigorous complement to those ideas and look forward to further
intense debate.
In addition, I would particularly like to thank Richard Orton, Peter
Coleman, Philip Holmes, Simon Emmerson, David Keane, my wife Jackie
for her continuing support and Jane Allen who typed the [original edition of
this] book.
Trevor Wishart
York, 1985

1 The original edition of this book was produced entirely by the author. There were two additional
paragraphs in the Preface of 1985. The first included a pessimistic prediction about the development
of (open access) computer music facilities in Britain which Wishart's own participation in the
foundation and development of the Composers’ Desktop Project in subsequent years was at least
partly to prove wrong. A final paragraph apologized for some of the literary and editing problems
inherent in a self-produced publication which we trust this new edition has addressed. (Ed.)
 ACKNOWLEDGEMENTS

Permission for the use of copyright material is gratefully acknowledged for

the following:
Figure 4.10 Calycles of Campanularia
Figure 4.11 Patterns of growth compared
From D'Arcy Thompson On Growth and Form (pp. 68, 294, 299, 318,
319) © 1961 Cambridge University Press. Reprinted by permission of
the publisher.
Figure 4.12 Two examples of ‘catastrophes’
From René Thom Structural Stability and Morphogenesis (pp. 66, 67,
72) © 1989 by Addison-Wesley Publishing Company, Inc. Reprinted
by permission of the publisher.
Figure 7.3 Layout of the orchestra in Stockhausen's Trans
From Karlheinz Stockhausen Trans (score) (p. xv) © 1978
Stockhausen Verlag. Reprinted by permission of the publisher.
Figure 10.1 The Gmebaphone (schematic)
Courtesy of the Groupe de Musique Expérimentale de Bourges.
Figure 10.3 The spherical auditorium Osaka World's Fair
From Karlheinz Stockhausen Spiral (score) (p. ii) © 1973 Universal
Edition. Reprinted by permission of the publisher.
Figure 11.9 Spectrograms of rhesus monkey sounds
Reprinted by permission of the Zoological Society of London.
Figure 12.15 Examples of Jean-Paul Curtay's iconic notation. Reprinted by
permission of Dr Jean-Paul Curtay.
Figure 13.2 Example of notation from Roland Sabatier's aphonic poem
Histoire. Reprinted by permission of Dr Jean-Paul Curtay.

Attempts have been made to locate the present copyright-owners of the

following figures: 4.8, 4.9, 11.1, 11.2, 11.3, 11.4, 11.5, 11.6, 11.7. The
author and editor would be happy to receive any information to allow due
acknowledgement in future editions.

The Editor acknowledges with thanks the assistance of Hugh Davies in the
location of material by Jean-Paul Curtay, and of Bob Cobbing for the loan
of the rare Kostelanetz Text-Sound Texts; also the continued support of
Harwood Academic Publishers.
Prelude
 Chapter 1
WHAT IS SONIC ART?

This book is based on a series of lectures given whilst in residence at

Queen's University, Kingston, Ontario in the Autumn of 1983. The
inspiration for this series of lectures was the new control possibilities
opened up by digital analysis, synthesis and processing of sound-materials.
Whilst attending the IRCAM1 computer music induction course in the
summer of 1981, I had many of my ideas about the internal structure of
sounds confirmed through the research on psycho-acoustics then going on
at IRCAM and also discovered an instrument—the computer—through
which I could realise some of the concepts of musical transformation I had
been dreaming of for some years.
This book is not about how to do it, though I will be discussing at
some stage new techniques such as frequency modulation synthesis, cross-
synthesis and model-building. Rather it is about why. Faced with all the
new possibilities for structuring sound and sequences of sound-events
thrown up by digital synthesis, analysis and control, what might be the
effect of ordering sounds in one way rather than another, and what might be
fruitful avenues for exploration?
This book is also essentially speculative, though I should stress it is
based on my own experience of working with sounds over many years. I
intend to throw up various possible options which might prove fruitful in
the future, and discuss some of my attempted solutions to these problems.
In so doing I am attempting to draw together various theoretical threads
which have emerged from fifteen years experience of working in music,
music-theatre and electro-acoustic music. Some of the material in this book
has appeared elsewhere in similar form (Shepherd, Virden, Vulliamy and
Wishart (1977); Wishart (1979)), but the majority of it is entirely new.
My own experience as a composer is quite broad. I have worked in
the spheres of music-theatre and environmental music events and, of
particular relevance to this book, free improvisation, electro-acoustic music,
both live and in the studio, and with extended vocal techniques. I have also
spent a good deal of time listening to the world, observing natural
landscapes and events and their structure and interaction, observing speech
and verbal communications and working with the calls of animals and
birds. Although I will refer a great deal in the first part of the book to the
work done by the Groupe de Recherches Musicales,2 I should emphasise
that I read their work (Schaeffer, Reibel and Ferreyra (1983)) only a few
weeks before presenting the series of lectures which form the basis of this
book. I quote extensively from it as it is both very thorough and it confirms
my own experience up to a point. But, although I admire the French group's
general refusal to present any written down theory of sound organisation,
preferring to rely exclusively on the ears and direct aural feedback, I would
like to go a small step further here and attempt to theorise about these
matters. I would stress, however, that my theories remain heuristic tools,
and not a means to replace intuition by an overall logos such as the serial
principle (which I shall criticise in detail later).
One essential aim of this book is to widen the field of musical debate.
One problem I have had in my own musical career is the rejection by some
musicians and musicologists of my work on the grounds that ‘it is not
music’. To avoid getting into semantic quibbles, I have therefore entitled
this book On Sonic Art and wish to answer the question what is, and what is
not, ‘sonic art’. We can begin by saying that sonic art includes music and
electro-acoustic music. At the same time, however, it will cross over into
areas which have been categorised distinctly as text-sound and as sound-
effects. Nevertheless, focus will be upon the structure and structuring of
sounds themselves. I personally feel there is no longer any way to draw a
clear distinction between these areas. This is why I have chosen the title On
Sonic Art to encompass the arts of organising sound-events in time. This,
however, is merely a convenient fiction for those who cannot bear to see the
use of the word ‘music’ extended. For me, all these areas fall within the
category I call ‘music’.
Of the eight sound examples3 which illustrate this chapter, only the
fifth was originally presented not as music. Example 1.1, from Pentes by
Denis Smalley, is at first glance a piece of pure electro-acoustic music on
tape, into which live-performance instrumental material has been
integrated. The second (Example 1.2) from my own Menagerie (the section
called Musical Box), uses the accident of interruption of another piece of
music as a starting point for structuring an electro-acoustic piece. Example
1.3, from Concrète PH-II by Iannis Xenakis is the most persistently abstract
example on first hearing. Example 1.4, from Michael McNabb's Dreamsong
uses computer technology to transform representational and vocal material.
Example 1.5 is an extract from BBC Radio's Goon Show (Napoleon’s
Piano). Although the use of sound-effects here is essentially humorous, this
is an early example of the creative use of sound-effects to do more than
merely set a scene for an essentially verbal presentation. Example 1.6 is by
Bernard Parmegiani, Étude elastique from De Natura Sonorum. Although at
first hearing no two examples could seem more different than the previous
two, I will show in a later chapter of the book how certain aspects of these
two approaches to the organisation of sound are in fact very similar.
Example 1.7, from Thema—Omaggio a Joyce by Luciano Berio, could
hardly be rejected for inclusion in a text-sound collection, though it has
always been presented as music. The final example (Example 1.8) is from
an album of free improvisation by the English guitarist Richard Coldman.
From the final quarter of the twentieth century, it now seems clear
that the central watershed in changing our view of what constitutes music
has more to do with the invention of sound recording and then sound
processing and synthesis than with any specific development within the
language of music itself. These latter developments have vastly expanded
our knowledge of the nature of sounds and our perception of them and
contradicted many nineteenth century preconceptions about the nature of
pitch and its relationship to timbre. Computer technology, offering us the
most detailed control of the internal parameters of sounds, not only fulfils
the original dream of early electronic music—to be able to sculpt all aspects
of sound—but also (as evidenced by the McNabb piece) makes the original
categoric distinctions separating music from text-sound and landscape-
based art forms invalid. We can no longer draw these lines of division. In
future it might therefore be better if we referred to ourselves as sonic
designers or sonic engineers, rather than as composers, as the word
‘composer’ has come to be strongly associated with the organisation of
notes on paper.
Looking around for a more general definition of the task of the
composer, we are faced with the following definitions of what music might
be. John Cage, for example:
Music is sounds, sounds around us whether we're in or out of concert halls: cf.
Thoreau. (Personal communication to Murray Schafer (Schafer 1969: 1)).

This is certainly a good definition to open our minds to the new possibilities
but unfortunately it is much too wide to offer us any advice or sense of
direction in our approach to the vast new world of sounds at our disposal.
At the other extreme Lejaren Hiller remarks, perhaps inadvertently, in an
article in Computer Music Journal:
[…] computer-composed music involves composition, that is note-selection.
Hiller 1981: 7)

Clearly for the contemporary composer, this is a uselessly narrow definition

of composition. It does not even apply to the structuring of the extended
drone sound in our first example Pentes. In fact, long ago, with the advent
of the voltage control synthesiser, it was possible to generate a piece
consisting of a singly attacked event which then proceeded to transform
timbrally and perhaps split into a number of lines without ever re-attacking,
i.e. a piece involving musical evolution, but without any ‘notes’.
These narrow conceptions can equally be found on the other side of
the fence. Richard Kostelanetz, in his compilation Text-Sound Texts
(Kostelanetz 1980), makes the following distinction between text-sound and
music:
The first exclusionary distinction then is that words that have intentional pitches, or melodies,
are not text-sound art but song. To put it differently, text-sound art may include recognizable
words or phonetic fragments; but once musical pitches are introduced, or musical
instruments are added (and once words are tailored to a pre-existing melody or rhythm), the
results are music and are experienced as such.
(Kostelanetz 1980: 15)
This definition is too narrow from the opposite point of view, as a listening
to both the Berio and the McNabb examples will evidence.
A more sophisticated series of specifications for the boundaries
between music and other disciplines is provided by Boulez in the book
Boulez on Music Today (Boulez 1971). In some ways On Sonic Art can be
viewed as a reply to Boulez's proposed limitations on the sphere of what
constitutes music. Here is what Boulez has to say:
Pitch and duration seem to me to form the basis of a compositional dialectic, while intensity
and timbre belong to secondary categories. The history of universal musical practice bears
witness to this scale of decreasing importance, as is confirmed by the different stages of
notational development. Systems of notating both pitch and rhythm always appear highly
developed and coherent, while it is often difficult to find codified theories for dynamics or
timbre which are mostly left to pragmatism or ethics […].
(Boulez 1971: 37)

In this book I will suggest that the logic of this assertion is inverted. It is
notatability which determines the importance of pitch, rhythm and duration
and not vice versa and that much can be learned by looking at musical
cultures without a system of notation.
What is the series? The series is—in very general terms—the germ of a developing hierarchy
based on certain psycho-physiological acoustical properties, and endowed with a greater or
lesser selectivity, with a view to organising a FINITE ensemble of creative possibilities
connected by predominant affinities, in relation to a given character; […]. (Boulez 1971: 35)

In this book, I will suggest that we do not need to deal with a finite set of
possibilities. The idea that music has to be built upon a finite lattice and the
related idea that permutational procedures are a valid way to proceed will
be criticised here and a musical methodology developed for dealing with a
continuum using the concept of transformation.
When noise is used without any kind of hierarchic plan, this also leads, even involuntarily, to
the ‘anecdotal’, because of its reference to reality. […] Any sound which has too evident an
affinity with the noises of everyday life […], any sound of this kind, with its anecdotal
connotations, becomes completely isolated from its context; it could never be integrated,
since the hierarchy of composition demands materials supple enough to be bent to its own
ends, and neutral enough for the appearance of their characteristics to be adapted to each
new function which organises them. Any allusive element breaks up the dialectic of form and
morphology and its unyielding incompatibility makes the relating of partial to global
structures a problematical task.
(Boulez 1971: 22–23)
This is a rather eloquent example of the ideology of instrumental puritanism
—thou shalt not represent anything in music. In this book I will propose:

(1) that pitch-free materials can be structurally organised, though not in

the hierarchic fashion used in lattice pitch music;
(2) that anecdotal aspects of sound-material can also be organised
coherently and in a complex manner and even enter into our perception
of the most supposedly abstract pieces. We are not talking here about
the concept of association which is often used in reference to
nineteenth century programme music, but about much more concrete
things which I will describe as landscape and gesture.

As has already been pointed out, sound-art can no longer be confined to the
organisation of notes. Even this original conception had already been
broadened to include at least three areas:

(1) the instrumental approach where pitched sound-objects of short

duration and fixed timbre were organised into larger structures through
the medium of conventional notation;
(2) musique concrète, using instead a vocabulary of sound-objects of
various types categorised according to a phenomenological description
of their properties and organised using studio techniques without
(necessarily) any reference to the notated score;
(3) voltage control synthesis techniques, giving us the possibility of
sustained yet transforming streams of sound.

The power of the computer to help us construct the internal architecture of

sounds from first principles allows us to broaden the concept of composer
to include the notion of sonic sculpture. At the same time the use of sound-
materials whose source is apparent or materials which, however abstract
they may appear to the composer, suggest a source to the listeners, means
that we may concern ourselves as composers with a landscape of the sound-
world we are creating. The ability to capture and manipulate text or other
vocal utterance (whether it be of human beings or other living creatures)
brings into consideration other aspects of the presentation of sound-material
which overlap almost completely with the concerns of text-sound-artists
and in fact links us into the sphere of animal communication.
 An additional reason that this book is called On Sonic Art is that, as I
shall explain further, conventional music theories, dealing with the
organisation of pitch in finite sets, rhythms using summative notation and
most usually in fixed tempi, and sets of instruments grouped into clearly
differentiated timbre-classes, I shall call lattice sonics. Everything from
isorhythm through Rameau's theory of tonality to serialism comes under the
general heading of lattice sonics and is adequately dealt with in existing
musical text-books. I therefore intend to concentrate on areas that have
conventionally fallen outside the scope of these theories. Hence On Sonic
Art. I must stress, however, that I am not underrating the organisation of
pitch and duration parameters as discussed in conventional theories. I am
merely assuming that all this is by now common knowledge.
Also, one further important point, in contradistinction to what is
implied in Solfège de l’objet sonore (Schaeffer, Reibel and Ferreyra
(1983)4) this book assumes that there is no such thing as an unmusical
sound-object.

1 IRCAM, the Institut de Recherche et Coordination Acoustique/Musique, is part of the Centre

Pompidou in Paris and was directed by Pierre Boulez from 1974–1992.
2 The Groupe de Recherches Musicales (GRM), is the current name of the organisation (the
originators of musique concrète) originally founded by Pierre Schaeffer in 1948 within Radio France,
now part of the Institut National Audiovisuel.
3 See the introductory note to the sound examples list.
4 An earlier identical LP version was issued to accompany Schaeffer (1966).
 Part 1

The Sonic Continuum

 Chapter 2
BEYOND THE PITCH/DURATION PARADIGM

This chapter is an expansion and development of ideas first put forward in

my contribution to the book Whose Music? A Sociology of Musical
Languages (Shepherd, Virden, Vulliamy and Wishart (1977)). The principal
point I am going to develop is that the priorities of notation do not merely
reflect musical priorities—they actually create them. It is fundamentally
important to grasp this point if we are to understand an approach to music
based on our listening experience. In order to develop this particular point,
we shall begin with a digression into media sociology. Our aim will be to
draw a distinction between what our notation system puts an emphasis upon
and what truly contributes to sound experience.
Three fundamental perspectives will be developed in this chapter.
The first of these is that notation is lattice-oriented; there are fundamental
aspects of sound experience even in the most highly notation-structured
music, which are not conventionally notatable and therefore are not in the
score. In fact music does not need to be lattice-based at all. Secondly, pitch
and duration do not need to be the primary parameters in musical
organisation. Thirdly, a perception and conception of music focused
through notation can lead to an abstract formalist approach. What I am
looking for are experientially verifiable criteria for making music. A
preoccupation with conventional notation can lead us into formalism, a
situation where there is no longer any experiential verification of our
theories about how to compose music.

Writing, speaking

Since very ancient times, human thought and communication has been
inextricably bound up with the use of the written word. So much so that it
becomes almost impossible for us to disentangle ourselves for a moment
from the web of written wisdom and consider the problems of meaning and
communication in vitro, so to speak. Ever since the ancient Egyptians
developed pictures into a viable form of hieroglyphic notation, our world
has been dominated by a class of scribes, capable of mastering and hence
capable, or deemed capable, of controlling what was to be written down and
stored in the historical record. Although this function was often delimited or
occasionally usurped by illiterate or semi-literate political supremos, such
tyrants have usually succumbed to the literate scribehood's cultural web as
evidenced by the ‘barbarian’ invasions of the Roman and Chinese empires
and to some extent by the Moslem conquest of Persia and Byzantium which
generated a novel cultural epoch by throwing together the divergent
scribehoods of these two long-established cultures under the unifying
banner of Islam.
In the long era of scribery, all people regarding themselves as
‘cultured’ or ‘civilised’, as opposed to illiterate peasants or craftsmen, have
lived within the confines of an enormous library whose volumes have laid
down what was socially acceptable and, in effect, possible to know and to
mean. Whilst those lying on the margins of ‘civilisations’ retain some
subcultural independence—variously labelled as ‘ignorance’,
‘backwardness’, ‘superstition’, ‘folklore’ or ‘folkculture’—they equally had
no access to the pages of history, and hence whatever the significance of
their cultural world, it was devalued by default. The vast growth in literacy
in the last century, with its numerous undoubted social advantages, has,
however, further increased the dominance of our conception and perception
of the world through that which can be written down.
So here we are in a library, and I would like to convey to you what I
mean. If, for a moment, we could put all these volumes of words on one
side, if we could face each other across a table and engage in the immediate
dialectic of facial and bodily gestures which accompany face-to-face speech
communication, perhaps you could appreciate that what I intend to mean is
not necessarily reducible to the apparent meanings of the words I employ
during the interchange; perhaps you could reach through my words to my
meanings.
Writing, originally a clever mnemonic device for recording the verbal
part of important speech communications between real individuals, soon
grew to such a degree as to dominate, to become normative upon, what
might properly be said. Divorced from the immediate reality of face-to-face
communication, it became objectified, generalised, and above all, permitted
the new class of scribes (whether priests, bureaucrats or academics) to
define and control what might ‘objectively’ be meant. Max Weber's
conception of the advance of Western civilisation, spearheaded by a
specialist rational bureaucracy, is a natural outgrowth of this simple
development. In fact, Weber devoted a small volume to a discussion of the
‘rationalisation’ of musical systems embodied in the Western European
tempered scale (Weber 1958).
For Plato, the idea of the object, which took on a new historical
permanence in its notation in the written word, came to have more ‘reality’
than the object-as-experienced. The commonplace tables and chairs which
we experience in the course of our everyday life were mere pale reflections
of the ideal table and chair existing in some Platonic heaven. (This heaven
in fact was to be found between the covers of books.) This radically new
stance reflects a permanent tendency of scribe-dominated cultures towards
the reification of ideas and the undervaluing of immediate non-verbal
experience, which has special relevance to the history of music. Even for
the average literate individual it might at first sight appear that what we can
think is commensurate with what we can say, and hence to appear verbally
confused or elliptical is easily interpreted as a failure of clear thought,
rather than a difficulty of verbal formulation of a perfectly clear non-verbal
idea. For example, the idea of a good ‘break’ in improvised musical
performance is clearly understood by any practitioner but has never been
adequately reduced to a verbal description.
I am going to propose that words never ‘mean’ anything at all. Only
people ‘mean’ and words merely contribute towards signifying peoples’
meanings. For the scribe meaning appears to result as the product of a
combinatorial process; broadly speaking, various words with more or less
clearly defined reference or function are strung together in a linear
combination to form sentences, paragraphs, etc., which have a resultant
clearly specified meaning. For the individual speaker, however, meaning is
a synthetic activity. She or he means. Not merely the combination of words
but a choice from an infinitude of possible inflections, tones of voice and
accents for their delivery, together with possibilities of movement, gesture
and even song, enter into the synthesis of the speech-act which attempts to
convey what he or she means. In this way a speech act may uniquely
convey quantities of information about the state of mind of the speaker and
his relationship to what is said (for example irony and so on) which would
be entirely lost if merely the words used were transcribed, but is certainly
not lost on the person spoken to. It is clear that not meaning, but
signification, resides in the words and that the mode and context of use of
these significations all contribute towards the speaker's meaning. These two
quite different conceptions of the meaning of words contribute differently to
our experience. The idea of meaning as a synthetic activity is most
significant in direct communications with other human beings, which might
be mediated through musical instruments or recording. The idea of meaning
as a structural property of written words governed by rules of combination
is the basis for the operation of our system of law. Law codes are in a sense
seen as existing transcendentally and having a meaning independent of the
original creators of the legal documents—though of course this does in time
lead to difficulties of interpretation.
Now immediately we become aware of a problem, for all that
remains of what we or anyone else ever meant, once committed to
parchment or print, is these marks on the paper. Here in the library, we see
love, tragedy, joy, despair, lying silently on the shelves, the entire history of
the word. Occasionally, a gifted scholar does appear to question the very
basis of a writing-dominated world-view. Lao Tse, the Chinese philosopher,
resorted to extreme verbal ellipsis in a late attempt to notate his
philosophical stance. At the other extreme, Marx, whose principal
commitment lay outside the scholarly profession, still felt impelled to
justify his world-view before the international scribehood and committed to
paper the astonishing theory that the world is shaped by human activity,
whilst talking, writing and the resulting development of ideas, constitute
only one particular type of human activity, and this of secondary
importance to materially productive economic activity. What had usually
been regarded as history-as-such was, in his view, merely one particular
reified result of human activity. The enscribed verbalisations of certain
mortals with certain preconceptions, economic interests and systems of
relevance.
Unfortunately, Marx's great scholarly erudition won for his radical
works a more or less permanent place on the library shelves, but in so doing
it delivered his work into the hands of the scribehood, who would
promulgate his writings, but not very often their significance. The up-and-
coming would-be radical scholar would learn about ‘praxis’ as a concept in
‘Marxist epistemology’, his understanding of alienation or class-
consciousness would be understood by its verbal competence.

Music and social control

At the other extreme, we have music! Ever since the world library opened,
there have been problems in this department. Somehow it seemed that
music could mean something to people, judging by their reactions, but this
something rarely seemed reducible to any definite verbal equivalent. Music
as an alternative mode of communication, however, has always threatened
the hegemony of writing and the resultant dominance of the scribehood's
world-view. Therefore, from the earliest times, attempts have been made to
lay down what could and could not be accepted as ‘correct’ musical
practice. Both Plato and Confucius recognised the threat posed by
uncontrolled musical experience to the ‘moral fibre’ of the rationalistic
scribe state, and advised the exclusive adoption of forms of music which
seemed to them to be orderly in some kind of verbally explicable way. As,
for the moment, there was no way of capturing music in the same way as
speech—no notation procedure—it seemed safest to adhere absolutely to
previous musical practice, while often ensuring that the music itself was
subservient to an approved text. The codification and standardisation of
church chant by Pope Gregory in post-Roman Europe may be seen as but
one example of a tendency which is exemplified by the Chinese emperor's
great concern for the ‘correct’ tuning of the imperial pitch-pipes at the
beginning of his reign, the execution of performers who made mistakes
during ceremonial performances in the Aztec world and in many other
cultures, and so on.
With the appearance of musical notation, new factors came into play.
However, a rapid glance at the syllabuses of most Western universities
(centres of writing dominated culture) will reveal the tremendous emphasis
placed upon the study of composers who employed a clearly, rationally
codifiable (verbalisable) musical praxis, in particular the work of Palestrina
(the champion of the Council of Trent), J. S. Bach and, of course,
Schoenberg and his ‘12-tone technique’. Even so, music continued to
convey its alternative messages and holy men (like St. Augustin) were
obliged to admonish themselves before God for being seduced by the ‘mere
sensuous quality of musical sounds’. This feeling that attention to aspects of
sound beyond those which are capable of description, and hence
prescription, in writing (and later in musical notation), is lascivious or
morally harmful is a recurring theme of scribe-dominated societies.
Committed verbalists will not be convinced by anything I have to say
about the separation between ‘meaning’ and ‘signification’. For the
linguistic philosopher all problems are reducible to problems of
signification within language and such a philosopher will merely deny the
validity of our problem. However, if you are capable of imagining that
talking to your lover is not merely an exchange of syntactically-related
arbitrary signs and bodily gestures, but an essentially non-verbal
communion between two people, mediated and articulated through word
and gesture, but not constituted by them, then you may understand what I
have to say.
Firstly, if this communion exists, surely it can be named. This is
perfectly true; however, the point remains that its articulation is not the
articulation of signs. We must not assume that we can notate its articulation
by attaching signs to different parts of it and then articulating the signs.
Written language constitutes what I will call a discrete/combinatorial
system. Written words are strictly delimited, distinct and repeatable entities
which form the finite elements of a combinatorial process of structure-
building. Our internal ‘state’ (whether a ‘bio-emotional state’ or
‘intellectual-physiological state’—but let us not be deceived by a label)
constitutes a holistic/ processual system. The distinction between these two
systems can be hinted at by reference to analogies. First of all we have the
distinction between an analogue and a digital system. In an analogue system
the state of the system can be represented by continuously varying
parameters (corresponding to the holistic/processual system) whereas in the
digital system the state is broken up into discrete samples which have
discrete values (corresponding to a discrete/combinatorial system). Of
course, with modern digital technology, the discrete states can be made so
close together, particularly in terms of time that the distinction between a
discrete and a holistic representation ceases to be of importance. However,
on the grosser level of representation that we find in the
discrete/combinatorial system of language, the distinction is absolutely
crucial. A second, though more tenuous, analogy might be seen in the
distinction between particulate and wave descriptions of phenomena such as
the behaviour of light, though again these have a point of reconciliation in
modern quantum theories.
The distinction between these two systems is perhaps one reason why
our vocabulary for referring to internal states is so vague and ill-defined.
Furthermore, there is an important distinction between the experience (the
state) as the state of ourselves, and the mere notations of it, the arbitrary
labels assigned to bits of the ongoing process; or between the most
immediate reality of me, now, and the reality of socially interdefinable
name-plates and syntactic laws. We may reach some agreement on how to
use these name-plates, but that does not touch the heart of the matter. This
problem is with us as soon as we begin to speak. But it is writing, with the
consequent reification of ideas in written reportage and the scribal control
of world-view that forces the problem to the centre of civilisation. Very
soon we are beginning to deny the existence of any sub-label reality at all,
and such things that we have called ‘the emotions’, or the highly articulate
gestural response in improvised music which we may vaguely refer to as
‘spontaneity’, become as mysterious as Platonic ideals.
What the aural-tradition musician takes on faith is that music does
touch the heart of the matter. With language, the actual medium may not be
of special significance; it may be spoken (sound), written (visual), touched
(Braille) and so forth. In a certain sense, a significant part of the message
transcends the immediate concrete experience of the medium which carries
it. Music, however, cannot be divorced from the medium of sound1 and
enters into our experience as part of an immediate concrete reality; it
impinges on us and in so doing it effects our state. Furthermore, as Susanne
Langer remarks in Feeling and Form, in its articulation of the time-
continuum of concrete experience, it corresponds directly with the
continuum of our experiencing, the continuous flux of our response-state
(Langer 1953: chapter 7).
 Hence, our pre-notation musician takes on faith that the way his
musical articulation of sound impinges upon his own state is in many ways
similar to the way it impinges upon the state of others. He seeks no verbal
confirmation (except indirectly), understanding that there can be none. We
might say that there is no divorce between the syntax of musical activity
and the syntax of musical experience. Whatever is played is directly
monitored, by the ears, by the player's immediate response to it. There is an
immediate dialectic of musical action and experience by which music
reaches directly to us in a way which language can never do,
communicating powerful messages which are not refutable within the
socially-approved categorical systems of any scribe-culture. It is music's
intrinsic irrefutability, its going behind the back of language, which has
caused it to be viewed with so much suspicion and disdain by guardians of
socially-approved order.

Musical gesture

The essential feature of this direct musical communion is what I shall

describe as musical gesture. In a sense it would be more logical to drop the
qualifying adjective ‘musical’ as the concept of gesture has much more
universal application both to other art-forms and to human experience in
general. In Chapter 6 I will be discussing in greater detail this concept of
gesture. Here I will confine myself to a few important observations. Gesture
is essentially an articulation of the continuum. It is therefore of special
relevance to any art-form or approach to an art-form which attempts to deal
with the continuum. Conventional music theory (at least in the West) deals
almost exclusively with the properties of sounds on a lattice. We will
discuss this concept a little further on.
Secondly, musical gesture is evidenced in the internal morphology of
sound-objects and also in the overall shaping of groups, phrases, etc.. In
fact, the morphology of intellectual-physiological gestures (an aspect of
human behaviour) may be translated directly into the morphology of sound-
objects by the action of the larynx, or the musculature and an instrumental
transducer. The translation of performance-gesture into the gestural-
structure of the sound-object is most complete and convincing where the
technology of instrument construction does not present a barrier. Thus vocal
music where there is no socially-constructed mechanical intermediary—and
particularly where performance practice has not become dominated by a
notation-based system of theory—is the most sensitive carrier of gestural
information. This reaches down to the level of timbre modulation, as well
as amplitude and frequency modulation (vibrato and tremolo and
articulation of all these) and up to all higher levels of sound ordering. All
wind instruments having a direct and continuous connection with the
physiological breathing of the player are similarly gesturally-sensitive
transducers although technology and performance practice can get in the
way—compare, for example, typical contemporary performance practice on
the flute and the saxophone. Bowed instruments, similarly, where sound is
produced by a continuing physiological action, are also gesturally sensitive.
Percussive instruments (from drums to pianos) are not gesturally sensitive
at the level of the individual sound-event, except in the elementary sense
that more energy in the gestural input leads to a louder sound, but gestural
information may be carried by groupings of individual sound-objects.
It is this immediate dialectic, however, which is broken asunder by
the advent of musical notation, causing a fundamental reorientation of
musical conception and perception in the West, and rendering music
susceptible to new verbal definitions and hence subjecting it to increasing
interference from the ‘verballigentsia’. Gestural structure is the most
immediate and yet notationally the most elusive aspect of musical
communication. One important feature of this book will be to suggest
means whereby gestural structures may be both notated and harnessed to
contrapuntal musical ends. Furthermore, in music which attempts to deal
with the continuum (rather than the lattice), gestural structure becomes the
primary focus of organisational effort.

Ideograms and alphabets: neumes and notes

Undoubtedly, musical notation, like ‘speech-notation’, originated first as a

mnemonic device for already well-established musical practice, but, like
writing, it quickly grew to dominate that musical practice. Just as the
original form of writing, the ideogram (see Figure 2.1), did not attempt to
convey the sound of words (as with alphabetic writing) but the ideas which
were expressed through the word-sounds and hence demanded a familiarity
with, and an adherence to, the sphere of those ideas, so the neume did not
attempt to mark out what we have now come to regard as individual pitches
and units of rhythm but only shapes and contours of melodic line customary
in current practice, and hence also requiring a complete familiarity with that
practice, and an adherence to it, before becoming usable.2 In this way these
first notation procedures tended to stabilise, if not to atrophy, the pre-
existing ideological and musical praxes. A more significant breakthrough
occurs with the emergence of analytic notation systems (see Figures 2.2b
and 2.2c). Here the verbal or musical praxis is analysed into constituent
elements which are notated, and the notations combined to form the
meaningful or characteristic units of verbal or musical praxis. In terms of
language, the earliest examples were afforded by the syllabary, as in
Hebrew, where constituent, but meaningless, syllables are assigned separate
written signs, and these strung together to form the combined sounds of
meaningful words and utterances. However, the most significant form of
analytic notation for language was the alphabet, probably invented in the
Middle East but taken up by the Greeks as the foundation of the first
literate, critical culture.3 The alphabet takes the principle of the syllabary
one stage further, notating the (idealised) sound-constituents of the syllables
themselves, and in so doing achieving such a considerable economy of
means—for example 26 letters in the English version of the Roman
alphabet as compared to tens of thousands of Chinese ideograms—that
universal literacy became a practical possibility for the first time (see Figure
2.1).
 Figure 2.1 Forms of script.

Particularly in relation to the further development of ideas in this

book, it is important to bear in mind that even in almost entirely phonetic
languages, like Finnish, there is not a one-to-one correspondence between
the spoken sound-object and the notation of it. The distinction we have
made earlier between the sequence of combinatorial sound-units in speech
and the use of inflection, tone of voice, etc. in the conveying of meaning is
only one level at which this comment is true. This distinction has been
raised as an issue within the sphere of linguistics. The original theorists of
language seem to have been committed to the discrete/combinatorial view
of the subject, but the conflict between ‘discreteness’ and ‘gradience’ is
now an issue. At a deeper level, computer analyses of the sounds of speech
show that the individual sound constituents (phonemes) are not spliced onto
each other in a way one might achieve in an editing studio but in most cases
elide into one another very rapidly in the course of the speech-act. Even
more fundamental, as will be discussed later, many consonants are
characterised by their morphology—the way in which they change form—
rather than by their spectrum (their particular frequency or formant
characteristics). All this relates very strongly to what I shall be saying about
the architecture of music.
The ideogram-writer had attempted to write down what was meant by
the speaker in terms of the ideograms which were notations of
conventionalised and traditional ideas; by the intrinsic nature of the system,
novel ideas were extremely unlikely to be recorded, even if they did arise in
speech-discourse. With the alphabet, however, the notation of the
constituent sounds of language made possible the recording of what was
actually said and hence made possible the recording of conflicting
statements and the emergence of the critical tradition (see Goody and Watt
1963). Whilst this freed language from the domination of the tradition-
bound ideas of a tiny elite of priest-scribes, it vastly expanded the spread
and domination of writing as a vehicle for mediating and explaining human
experience, and hence led to the devaluation by default of all non-verbal
modes of action and communication and all non-notatable aspects of
discourse—the ultimate triumph of a newly-expanded secular scribehood.
 Figure 2.2a Tibetan neumes.

Figure 2.2b 10th century European neumes.

Figure 2.2c Modern European analytic notation.

The effect of analytic notation of music in the context of a writing-

dominated world was much more fundamental. Arising only in Western
Europe, it developed considerably later than alphabetic writing. The
fundamental thesis of this system is that music is ultimately reducible to a
small, finite number of elementary constituents with a finite number of
‘parameters’, out of which all sounds possibly required in musical praxis
can be notated by combination. It must be noted from the outset that this
finitistic thesis is a requirement of notation rather than fundamentally
necessary to conceivable musics. For a notation procedure to be of any use
it must use only a manageable (small) number of constituents which are
then permuted; notation of the continuum is necessarily approximate. This
is the same problem we have met with verbal categorisation of the internal
experiential state (and also in the discreteness/gradience issue) and is very
important in relation to my discussion of gesture.
The two features of sound used in tenth century Western musical
practice which appeared most accessible to analytic musical notation were
pitch-level and rhythm. Timbre was not tackled in this way, up until the
twentieth century being limited by the available instrument technology; the
continuum of possible dynamic levels has never been remotely accurately
categorised, despite attempts to give it a notational rationale in some
integral serial composition; while dynamic balance—remaining largely a
matter of unspoken convention—and acoustics—usually the accident of
performance location—have only come under accurate control with the
advent of electronic sound-recording techniques.
However, even pitch and rhythm could only be captured in a very
particular way, determined by the exigencies of analytic notation itself.
Thus, whereas aural rhythm takes place against the silent backdrop of
somatic rhythm, enabling the aural musician to indulge in the most intricate
articulations of time, notated rhythm is limited by the problem of notational
economy. We can divide time infinitely and in performance can judge
directly the effectiveness of the most subtle placements of sounds. But
analytic notation is a finitistic procedure. We must be able to count the
divisions in order to write them down—but not necessarily in order to judge
aurally what is effective. Hence, analytically notated music is bound within
the limitations of summative rhythm (see Figure 2.3).
Similarly, discrete fixed pitches are idealisations of acoustic reality.
In practice there are only sounds in their infinite variety of possible
frequency, spectrum, timbre, dynamic-envelope, and change (dynamic
morphology) and combinations of all these. Consider the irreducible
infinitude of tones of voice. But the infinite is not simply notatable. What
notation demands is a finite set of pitch-levels which we can permute and
combine. The refinement of instrument technology attempts to impose this
discrete permutational rationality upon the very production of sounds, and
our ears learn to approximate our acoustic experience to the discrete steps
of our imposed logic.

Lattice and continuum; on instrumental streaming

We are now in a position to describe the concept of a lattice and its bearing
on conventional music theory. For anyone who has ever heard a pitch
portamento or a tempo accelerando, both pitch and tempo can take on an
infinitude of possible values and may vary continuously over this
continuum. Notation, however, imposes a finite state logic upon the two
domains. The result is that music, at least as seen in the score, appears to
take place on a two-dimensional lattice (see Figure 2.4a). Two things should
be said about this lattice formulation. First of all it is our conception of what
constitutes a valid musical object which forces ‘musical sounds’ onto this
lattice; secondly, despite our intentions, the lattice only remains an
approximate representation of what takes place in actual sound experience
(except in the extremely controlled conditions of a synthesis studio).
The technology of instrument design underlines and reinforces this
lattice conception of musical architecture. First of all, on keyed, holed or
fretted instruments, the discrete logic of the pitch lattice is imposed on the
production mechanism of sound-objects. Secondly, the concept of the
instrument itself further expands the lattice notion. Conceptually, at least, an
instrument is a source of stable timbre, but variable pitch. The essential
function of an instrument is to hold timbre stable and to articulate the pitch
parameter. This conception contributes to the myth of the primacy of pitch
(and duration) in musical architecture. The grouping of instruments into
families of distinct timbral type and the development of music based upon
fixed-timbre (or instrumental) streaming develops the lattice one stage
further.
Figure 2.3 Summative rhythm: each note value can be expressed as the sum of smaller equal note
values.
 Figure 2.4a Music on a two-dimensional lattice (schematic representation).

Hence music can now be viewed as taking place on a three-

dimensional lattice (Figures 2.4b–2.4d). The three dimensions being made
up of discrete pitch-levels, discrete durational values, and discrete timbral
objects (or instrumental types). In fact, the concept of the instrumental
stream is perhaps the most persistent in conventional musical thought—the
lattice of both pitch and duration have been challenged by composers
working within conventional notation. Even in the classical voltage control
studio it was possible to conceive of a musical composition in which a
single sound stream evolved, possibly diverged into separate streams which
might be separately articulated, might reconverge and continue thus to the
end of the piece (see Figure 2.5). The evolving streams within such a piece
might be continually changing their timbral characteristics, even though
they were continuously connected to the opening event (i.e. the piece need
only have one attack—at its opening—and therefore in the conventional
musical sense contain only one ‘note’). The conception of music as
consisting of fixed-pitch, fixed-timbre entities called ‘notes’ is extremely
persistent.4 It even imposes conceptual limitations upon the design of
digital musical instruments (where such traditional conceptions are no
longer necessary). Computer music machines such as the Fairlight and
Synclavier with their keyboard input and instrument definition, and even
the more general Music 11 program, carry with them into the digital world
the concept of instrumental streaming from conventional musical practice.5
It is of course possible to subvert the various systems but it is a struggle
against the design concepts of the instrument or software.
 Figure 2.4b Music on a three-dimensional lattice (schematic representation).

Figure 2.4c A complex sound-object moving in the continuum (schematic representation).

Figure 2.4d Frequency/timbre cross-section of sound at start, mid-point and end.

 Figure 2.5 Evolving timbre-stream composition (schematic representation).

We can, however, perceive important distinctions within lattice-based

musical conventions. The major distinction—though this is usually a matter
of degree—is between music ‘hung around’ a lattice and music developed
‘on’ the lattice. Examples 2.1, 2.2, 2.3 and 2.4 illustrate Japanese joruri
singing (gidayu style used as part of the bunraku (puppet) and kabuki
theatres), North Indian singing, jazz singing and part of a Haydn mass. In
certain musical cultures, the pitch lattice (which might also be referred to as
a harmonic field so long as we do not connect this with Western harmonic
thinking) may be regarded as a framework around which the music takes
place, thus in the example of joruri singing, the pitch set (or rather the
pitches which can be identified as being on a lattice) is fairly limited and
does not change its general character. However, the focus of attention of the
performer (and of the listener) is heavily weighted towards aspects of the
sound articulation which cannot be related directly to the lattice. Thus there
is a complex articulation of portamento structures leading onto or away
from the lattice pitches, the control of articulation of vibrato (frequency
modulation)—in Western art music vibrato is usually a relatively constant
parameter of a particular type of vocal production rather than a parameter
which one articulates through time—and focusing on the evolution of
timbre within individual events (particularly the development of sub-
harmonic colouration). These aspects of musical articulation are carried
over into the more speech-like sections of gidayu style presentation. An
attempt to capture the essence of this music in conventional Western
notation would clearly fail miserably.
In North Indian music concepts of pitch (or more precisely pitch on
lattices), as opposed to the kind of subtle portamento articulation found in
the Japanese example, are more highly developed, although musical
development still takes place over a fixed pitch-set (harmonic field). Even
here, however, we find the use of subtle sliding inflections onto and away
from the lattice pitches and the internal articulation of the sound-objects
which make up the ululation-based runs. Again, these cannot be
approximated by conventional Western notation procedures but are clearly a
fundamental aspect of the musical structure. Jazz is clearly much more
strongly influenced by the lattice-based approach of Western harmony. Yet
typical jazz vocal and wind instrument production is very heavily
concerned with the internal articulation of sound-objects including sliding
ornamentations and careful control of vibrato and other timbral modulations
of the sound. All these features can now be clearly described—they are not
mysterious in any way—but again would be lost in conventional Western
notation.
Finally, the example from Haydn illustrates what happens to vocal
production when musical conception is focused upon the lattice itself. Vocal
production becomes conventionalised and aims at an idealised type of
production focusing on the lattice pitches. Idiosyncratic developments of
timbral and pitch articulation, which serve to identify and project particular
jazz singers, for example, are to be rejected in favour of a universally
stereotyped bel canto production. Vibrato is no longer a parameter for
musical articulation but a relatively fixed feature of the required sonority.
The latter example typifies music developed on a lattice where development
of the parameters of the lattice itself dominate all other types of musical
articulation. If we now turn to the instrumental Examples 2.5, 2.6 and 2.7
which are of Japanese shakuhachi playing, a classical chamber work (using
wind instruments) and a piece of jazz, we will hear a similar development
in the use of wind instruments. In fact, it is the combination of a conception
of music focusing on the parameters of the lattice and the developing
technology of instrument design going along with this developing
conception which leads us away from the multi-dimensionally rich
articulation of the shakuhachi towards the timbral uniformity of the present-
day Western keyed flute.6 Despite these developments, however,
articulation of the continuum is still present in performance practice. As we
have discussed earlier, the articulation of the continuum in intellectual-
physiological gesture is transferred directly to the sound-object by the
player of a wind instrument. Even in the most rationally developed notated
scores, aspects of performance gesture, often loosely referred to under the
term ‘interpretation’, still have an important bearing on our musical
experience. In certain types of music, articulation of the continuum plays a
much more significant role as can be discerned by comparing the typical
use of the trombone, trumpet and particularly the saxophone in jazz with the
typical use of the keyed flute in classical Western music.
In this continuing technological development, the voice and the
keyboard may be seen as occupying the two opposite ends of the musical
spectrum. Voice, the immediate source of intellectual-physiological gesture,
will be seen as an important focal model for music throughout this book.
The keyboard on the other hand represents the ultimate rationalisation of a
lattice-based view of music. Timbre is fixed; pitches are incapable of any
sort of inflection, physiology is only allowed one single point of contact
with the sound-object, at the initiation of the note, and thereafter can have
no impact on the internal morphology of the sound. This distance can
perhaps best be appreciated by comparing the ululated trill articulation of
the North Indian vocal music example with the typical trills and turns on a
keyboard instrument. Vocal articulations such as trills and turns are semi-
unified objects, the apparent pitch-elements of which are bonded by subtle
internal portamenti and timbre transitions. On the keyboard instrument the
individual notes of the trill or turn are as close as we can possibly
approximate to the individual notes on the page. It is interesting and ironic
in this respect that the computer, in some senses the ultimately definable
and controllable musical instrument, has for the first time begun to reveal to
us the subtle inner architecture—the continuum architecture—of sounds.
It is very important to understand that the lattice is a conceptual
construct. It is we who have decided to construct our musical architecture
on the lattice. Because we do, however, it is very easy to fall into the mental
trap of observing the world of sounds as if it divided up neatly on a three-
dimensional lattice. Thus for anyone with a conventional musical training—
and particularly for those with no studio experience—sound-objects appear
to be divisible into three distinct categories of pitch, duration and timbre.
This is of course true of most sound-objects appearing in conventional
music—they have been constructed on the lattice and are therefore divisible
in terms of that lattice. In fact as we proceed we shall see how the
conventional (Helmholtzian) view of acoustics tends to fall into the same
trap. At this stage we will merely note that lattice notation encourages the
following connections:

(1) instrumental streaming leads us to suppose that timbre is a simple

category like frequency;
(2) focus on pitch leads us to suppose that pitch itself is a simple category
(though it is in fact simpler than timbre);
(3) viewing duration through lattice notation leads some members of the
musical community to view Dave Brubeck's excursion into 5/4 metre
as a major breakthrough in jazz rhythm (rather than the minor
excursion on the lattice which it is), while entirely overlooking the
highly articulate development of phrase-structures against the lattice
(Charlie Parker) or placement of individual events or groups against
the lattice (the essence of ‘swing’).

Even where it is clear that the lattice is only an approximation to

musical reality, notation focuses our attention on the lattice. In the long run,
all ‘respectable’ theory is based on the lattice (see below).

Pitch versus timbre: primary and secondary qualities

In the West, the rationalisation of music on a lattice is taken to its extreme.

First from the infinitude of possible pitch levels which could give rise to
numerous subtly different musical scales, such as the scales of the ragas of
Indian music and probably those of Western medieval pre-notation chant—
though this we will never know—a small set of twelve clearly specified
pitch-levels is gradually selected. Then partly through the tendency—
intrinsic in the notation system and its realisation in the technology of
instrument (especially keyboard instrument) design —towards a rational
simplicity, a notational economy, the well-tempered scale arrives,
permitting a considerable opening up of the field of harmonic inter-relations
among a limited set of fixed pitches as Bach and composers through
Wagner and Schoenberg were to demonstrate.7
In similar ways to alphabetic writing, analytic notation is in many
ways a liberating invention. It frees composers from the established norms
of a musical tradition and permits him or her to explore new and unheard
possibilities. At the same time it is this very malleability of the notatable
parameters which enables and encourages the one-sided, two-dimensional
expansion of musical possibilities. This eventually leads to Boulez's
theoretical distinction between primary and secondary qualities in music.8
The primary qualities are those which have been accurately notated—in a
certain limited sense—the secondary qualities those which have not.
 There is a striking parallel here with the distinction made by
Descartes between primary and secondary qualities of perceived natural
phenomena. For Descartes, a phenomenon such as motion which could be
given a direct quantitative mathematical description was regarded as
primary, whereas qualitative phenomena such as colour were seen as
secondary qualities and ultimately reducible to descriptions in terms of
primary qualities. Exact mathematical representation, at least in theory, here
plays the same role as accurate score-notation plays in Boulezian music
theory. An interesting sidelight on this parallel is thrown up by the recent
development of ‘catastrophe theory’ which will be discussed more fully
below. Very briefly, physicists have tended to confine themselves to a study
of equilibrium situations where in most cases precise quantitative
mathematical formulation of the problem is possible. Note that most
musical objects may be considered as examples of stable equilibria (for
example after the initiation of a flute tone, a stable resonance is set up
within the body of the instrument which constitutes the sound-object which
we hear). Recently, however, attention has been focused on the study of
more complex regimes whose stability may vary along with small changes
in the parameters which define the situation. The study of such situations
has established the first essentially qualitative branch of mathematics—
differential topology. This branch of mathematics may give us some insight
into the structure of the continuum and therefore has a bearing on the study
of sound-objects in the continuum that we are pursuing in this book. In a
similar way, the Helmholtzian theory of timbre may be seen as an attempt
to reduce the qualitative (timbre) to the quantitative (frequency) which has
in fact proved untenable (see Helmholtz (1954 originally 1877)).
In its constant search for new modes of expression, the Western
classical music tradition was, however, constrained by its very
concentration upon relationships of a limited set of thus notatable ‘pitches’
to extend the notatable field of harmonic relationships to the limit. The final
step into a twelve-tone and thence ‘integral’ serial technique, rather than
being a liberation from this restricted-set tonality, should be seen in
historical perspective as the final capitulation to the finitistic permutational
dictates of a rationalised analytic notation system. Within this same
tradition, however, composers have made attempts to abandon the lattice-
dominated aesthetic.
 Consider now Examples 2.8, 2.9, 2.10 and 2.11. In Example 2.8, from
the Webern Symphonie, we hear the apotheosis of the rational extrapolation
of lattice aesthetics. In Example 2.9, from Penderecki's Polymorphia, we
have a fairly typical example of this composer's approach to composing
music which no longer conforms to the traditional lattice. In particular, he
uses thick groupings of pitches only a quarter-tone apart (thus destroying
the twelve-note chromatic lattice) and also textural aggregates of sounds
with no, or ambiguous, pitch content. The sonorities are very striking, but
the overall architecture does not seem so strong. The music seems to
develop monophonically and tends to fall into long blocks of a particular
sonority. We can say that the composer has broken free quite successfully
from the domination of the lattice but as yet no strong and sufficiently
articulate means of organising the new material has emerged.
In Example 2.10, from the end of Xenakis’ Pithoprakta, we have a
more interesting example of non-lattice-based musical organisation. The
written score for this piece is superficially impenetrable, but if we sketch
out the various notated pitch glissandi on a sheet of graph paper in which
pitch and time form the axes the architecture of this particular section is
quickly revealed (see Figure 2.6). Xenakis has grouped individual short
glissandi on the string instruments into larger arching glissandi (glissandi of
glissandi!).9 At the same time the sounds are grouped into three contrasting
string sonorities and the three resulting timbre streams arch up and down
independently. In this way a pitch-based counterpoint of timbre streams is
created which in no sense depends on the typical pitch lattice of
conventional music. At the end of the section, as will be seen clearly from
the figure, the glissandi of glissandi thicken out and unfold into a sustained
chord, a wonderful process of pitched evolution which has no real parallel
in typical lattice aesthetics. Although the processes of musical organisation
here seem more articulate and evolved than in the Penderecki example, they
have what Pierre Schaeffer has described as an architectural feel, that is to
say that the gestural unfolding of events is quite slow and controlled. There
is as yet not a moment-to-moment feeling for the gestural development of
musical form. This is partly due to the essentially cumbersome nature of the
orchestra when it comes to attempting to define non-lattice structures.
Inevitably such structures must be constructed from individual elements
which are notated on the lattice or in relation to the lattice and it becomes
difficult to notate a rapidly evolving event. Such events with rapidly
involving internal morphologies are much more easily accessible in the
electro-acoustic studio and it is here where the problem of their organisation
begins to confront traditional musical aesthetics.

Figure 2.6 Glissandi of glissandi in Iannis Xenakis's Pithoprakta.

Then finally in this group, Example 2.11, from Stockhausen's Carré,

illustrates yet another attempt to deal with the internal morphology of
sound-objects. Stockhausen's piece is largely concerned with relatively
sustained events in which there is internal motion, for example the slow
glissandoing of trombones and voices in the opening moments. The larger-
scale relationship between these individually articulated ‘moments’ is still
governed by serial permutational criteria which, in my view, are an
outgrowth of lattice-oriented thinking, and not on a gestural interaction
between the individual sound-events which would generate a truly dynamic
non-lattice-based musical form.
It is interesting in this respect that a composer like Boulez, who
seems so adamantly committed to a lattice-based view of musical aesthetics
produces music which is, from the listener's point of view, much more
clearly gesturally articulate. In Example 2.12 we may consider a section
from Don (from Pli selon Pli). The pitch and durational characteristics of
this section are no doubt exceedingly carefully worked out, but in practice
what one hears is its gestural structure. The music is dominated by
sustained but hovering—and by implication pregnant—events. The initial
loud attack and the mode of sustainment suggest that the events will burst
forth into something else. This feeling is underlined both by the fact that
separate events enter on different, though related, harmonic fields and
particularly by the brass event, which, after its initial attack, begins to die
away and then crescendos. This emergence verifies the pregnancy of the
other (sustained) gestures. I will not attempt to give a more detailed gestural
analysis of this particular passage here, but even this much serves to
underline the significance of musical gesture, even where a lattice-based
aesthetic appears to dominate through the score.

Musical values distorted; the emergence of formalism

As we have mentioned previously analytic notation is in many ways a

liberating phenomenon; it permits us to explore new possibilities for
musical expression. Using it we may, but need not, discover new modes of
musical experience. However, this begs the central question of what defines
a musical experience and this very concept has been fundamentally twisted
by the impact of musical notation itself, gradually forcing music to kow-
tow to the verbally definable. In fact, with the increasing domination of
notation, there has been a move towards Platonic idealism in our conception
of what music is. In the most extreme cases, music is viewed as an
essentially abstract phenomenon and the sound experience of essentially
secondary importance. More commonly the score is seen as normative on
the musical experience.
The split in conception between what are seen as primary and
secondary aspects of musical organisation leads to a split between composer
and performer, between composition and interpretation and the gradual
devaluation of non-notatable formations. This development leads directly to
the attitudes expressed by Boulez and to the intellectual devaluation of
forms of music (such as jazz improvisation) where non-notatable aspects of
musical form have greater importance than in conventional classical music.
At the same time, the spatialisation of the time-experience which takes
place when musical time is transferred to the flat surface of the score leads
to the emergence of musical formalism and to a kind of musical
composition which is entirely divorced from any relationship to intuitive
gestural experience.
What takes place is not merely a focusing of our perception upon the
notatable and the consequent feedback upon our musical praxis, but a
reorientation of our conception of music. Whereas previously verbal
discourse had little of permanence to grasp onto in music except the very
continuity and unity of established practice, which it could reinforce and
stabilise by verbal decree, now musical process appeared to reveal itself
concretely in the form of musical scores. A fleeting succession of musical
experiences in time appeared to be captured in a continuously present
spatial representation which could be studied at any time, at any speed and
in any order. Just as the immediate dialectic of speech had been
fundamentally subverted and devalued by the permanent monologues of the
written word, so an intuitive and unverbalisable knowledge of music as an
immediate dialectic of musical action and the fleeting, inscrutable musical
experience was to be fundamentally challenged by the permanence and
scrutability of the score. Permanently available and amenable to
rationalistic verbal explication the score rapidly usurps the sound
experience of music as the focus of verbal attention and becomes the
keystone of an eminently verbalisable conception of what ‘music’ is.
The most obvious consequence of the discovery of analytic notation
is the emergence of the composer, who is able to challenge and expand
existing musical praxis through creating notations of novel musical
activities, his original scores. The novel split which gradually emerges
between composer and performer, between a score and its ‘interpretation’ is
the concrete realisation in music praxis of the perceptual focusing upon
notatable ‘parameters’.
Interpretation, still a semi-intuitive discipline, remains of great
importance in the education of the musical performer, who remains
somewhat outside the sphere of intellectual respectability. For the music
scholar, however, raised in primarily verbally-based institutions, especially
the new European ‘university’, the focus of attention is on that musical
syntax which can be discovered in the score. At the same time, the
composer, whose musical tools are the notations at his disposal, will clearly
tend to develop a musical syntax based on the organisation of these
notatables. Hence, whilst ever the musical scholar concerns himself with
notation-composed works, there will be a congruence of attention upon
analytically notatable syntax, as scholar and composer have the same vested
interest in notatability. The concatenation of scribal domination,
compositional necessity and the limitations of analytic notation, however,
elevate the organisation of a certain limited range of musical variables to
the status of ‘music’ as such and leads to an inevitable clash of values when
the classically trained musician comes into contact with music from an alien
tradition.
We have already listened to and discussed music from the Japanese
joruri tradition and from jazz, musics which to some extent develop their
form outside the notatable lattice. A more radical example of such a music
can be found in free improvisation. In Example 2.13 (free improvisation)
we are faced with a musical experience where reference to the notational
lattice is completely useless. The reaction of classically trained musicians to
free improvisation, often on the basis of limited experience, can be quite
negative. In fact the lack of explicit criteria in the field of free
improvisation does lend itself to exploitation by mediocre performers.
‘Spontaneity’ is taken by some to mean self-indulgence, arbitrariness or
whimsicality and in a typical bad performance, the participants ignore each
other for most of the time, except at points where they all get louder and
then (hopefully) all get quieter again. In a performance, however, where the
participants have absorbed themselves in the technique of this kind of
improvisation and freed themselves from conventional response and
musical clichés, what we hear is a rapid and highly articulate gestural
interaction between the performers. The ebb and flow of musical tension
can be exceedingly rapid and the music highly articulate in its impact.
Because of its essentially highly gestural basis, free improvisation need not
confine itself to the use of conventional sound sources. Any sound-
producing object may be turned to musical advantage10—an interesting
parallel with musique concrète. The use of simple and non-prestructured
sound-sources allows, in fact demands, strong gestural input from the
performer. Experience from the free-improvisation forum can be extended
into the electro-acoustic studio, as is shown by Example 2.14 from my own
Anna's Magic Garden where the use of sounds of a piece of elastic is
extended by simple musique concrète techniques.

Figure 2.7 An example of ‘plus/minus’ notation (after Stockhausen).

Free improvisers in general make no attempt to notate the gestural

structures and interactions which underlie their musical activity.
Stockhausen, during his short digression into the sphere of improvised
music, did attempt to develop some kind of simple notation to give form to
otherwise freely-improvised music pieces. The ‘+/-’ notation of this period
(see Figure 2.7)11 is an interesting early attempt to impose some sort of
compositional rationale upon this basically intuitive discipline. Its
simplicity and rationality meant, however, that not very much of what
actually happens in the musical unfolding is really captured in the score and
in some ways it seems to function more as an enabling device for
permutational procedures to be imposed at least on the notations if not on
the actual musical experience!
I shall have much to say about the sphere of electro-acoustic music
elsewhere in this book, but in this particular context it is interesting to note
the relative failure of electro-acoustic music to achieve academic
respectability. This can to some extent be put down to the fact that no
adequate notation exists for it—in fact many composers have actively
avoided developing notational parallels to their musical events. But scholars
like to see the music! The arrival of vast reams of computer print-out will
no doubt put an end to this lacuna. Even the skills of the professional
mixing engineer in achieving balance and artificial acoustic in the typical
rock recording would not normally be considered ‘musical’ by the
traditionally trained musical scholar, though composers who have worked
in the electro-acoustic studio have come to see this rather differently.
A further, somewhat negative effect of the focus on lattice aesthetics
is the destruction of the lively non-lattice aspects of various folk music
cultures by concerned composers anxious to preserve these traditions in
conventionally scored arrangements. In the extreme case, the combination
of pitch, rhythm and timbre inflection in jazz and rock music is seen as
lascivious, sexually suggestive and ultimately a threat to social order. As we
can now see, this is more than a mere rejection of that which falls outside
the clearly-definable limits of a long-established notation (perception and
conception) procedure with its verbal-explicability and hence its social
controllability. In a narrow sense, this attitude is correct, for musical
experience, even where apparently constrained by clearly explicable
notation-based procedures, is ultimately irreducible to verbalisations and
hence beyond any direct social control.
The most radical impact of analytic notation on musical praxis is to
transfer the musical structure out of the uni-directional continuum of
experiential time, in which the musical dialectic takes place and in which
musical gestures unfold, into the spatialised, perfectly reversible
(Newtonian) time of the printed page.12 In sound, the musical experience
begins at the beginning and must be taken in the irreversible order and at
the rate at which it comes to the listener. Furthermore, our experience of
what arrives later is modified by our (perhaps inaccurate) memories of what
has passed and, in this sense, there can never be a clear-cut ‘recapitulation’;
everything is modified by the context of what went before. In the score,
however, the whole span of the music appears to exist in a timeless,
spatialised, present. We may peruse its contents at any rate, in any order. In
this way we may be able to see relationships, for example of recapitulation,
which, however, after repeated and thorough aural experience of the music
as sound we may never be able to hear. Can we thus treat such a
recapitulation as an element of musical structure? This, of course, begs the
central question of what constitutes music, what we experience in the
sounds, or what we might theoretically appreciate of the score through the
sounds, if our aural selectivity were more finely developed.
The best example of the split between a view of music based in
unidirectional experiential time, and one based on spatial reversibility of
time as represented in the score, is found in the concept of the retrograde as
found in serial music, and also in some medieval and renaissance
polyphony. Here the notational view is that by reversing the order of a
group of notatable pitches we arrive at a pitch-set which is merely a derived
form of the original. In the immediately present and spatially reversible
time of the printed page the relationship of the two sets may be abundantly
visually clear, but in the uni-directional and memory-dependent time of
musical experience, considerable aural retentivity and the performance of a
rapid feat of mental inversion is necessary to grasp this relationship. This
may be a simple matter if the sequence of pitches is quite short or very fast.
It is also true that if we conceive of the pitch set as a harmonic field the
relationship between the two sets may be easier to grasp. With long and
complex structures, however, the difference in perception between the
sound experience in time and the visual scanning of the score is extremely
marked. When we consider extended use of retrograde or cancrizans form,
such as the perfect arch-form of Der Mondfleck from Schoenberg's Pierrot
Lunaire, where the entire movement runs in reverse order of pitches and
durations from the centre point—except, to complicate matters, the voice
and piano, the latter having an elaborate fugato at the same time!—we must
declare that experiential structure has been sacrificed to notational
‘conceptual art’. The retrograde of a ‘duration series’ as used in ‘integral
serial’ composition, is even more experientially problematic. All this then
begs the question. What is music? The time-based experience in sound? Or
an essentially abstract entity existing outside time, a Platonic conception of
music?
Thus, just as the permanence of the written word ‘table’ appeared to
Plato to project something more permanent and more ‘real’ than the many
experienced tables of a concrete reality, so, to the Western musical scholar,
musical notation can appear to project something more permanent and more
‘real’ than the direct, but fleeting, experiences of the sound of a musical
performance. Music may hence be regarded as a phenomenon which
transcends immediate sense-experience. With the accompanying dominance
of composed music, ‘music’ and its ‘interpretation’ can hence be
distinguished from one another and notatable syntax, discussible in a verbal
space divorced from direct sense-experience, elevated to the position of
musical syntax itself.
Schoenberg, the originator of the serial method, was clearly not
unaware of the notational ideal/experiential dichotomy we have discussed.
In actual practice the ‘harmonic style’ of his later serial works is not greatly
dissimilar to that of his pre-serial ‘expressionist’ works, where the harmonic
tensions characteristic of tonal music can still be felt even though traditional
tonal progressions have disappeared. Having abandoned tonality as a basis
it would seem that Schoenberg still felt the need to rely on an intuitive feel
for harmonic relationships, and this approach is characteristic of his musical
language with or without serialism. This fact is underlined by the rejection
of Schoenberg's ‘backward-looking’ approach by proponents of the post-
Webern school of serial (and integral serial) composition.
The fundamental conflict between the two views of music is in fact
most clearly expressed in the symbology of Schoenberg's serial opera
Moses and Aaron. The conflict between Moses’ view of God as the all-
pervasive, yet ultimately intangible, idea and Aaron's desire to relate God to
tangible experience, is represented at a surface level by a verbal/musical
dichotomy—Moses’ part is confined to heightened speech (Sprechstimme),
while Aaron sings expressive melodic lines—and at a deeper level by a
notational/experiential dichotomy—the tone-row is all-pervasive in the
score as the structural material out of which the opera is built but is not
generally audible as such.
The second act (the end of the opera as it exists) ends with a dramatic
duologue in which Moses (the speaker) holding the tablets of the written
law, confronts Aaron (the singer) and finally breaks the tablets declaring “0
Wort, du Wort, das mir fehlt” (“Oh word, you word, which has failed me”).
In the archetypal ideology of Schoenberg's biographer,
[...] in contrast to Moses the thinking character who clings to eternity, metaphysics and real
values, Aaron is a materialist of everyday life who is impressed by the glitter of gold and the
successes of the moment.
(Stuckenschmidt 1959: 151, my emphasis)

However, the fact that Schoenberg felt unable to write the third act of the
opera, in which Moses’ view triumphs has, in the light of the present
discussion, far-reaching significance for an understanding of contemporary
avant-garde music.
Where concrete musical relationships—at least originally based on
their experiential success—are represented by their notations in the score,
and study and conception focuses upon this structure, divorced from the
experiential immediacy of the sound itself, these relationships, as
rediscovered in the score, may be mistaken for conventional relationships.
In other words, what to direct gestural experience may appear as a
necessary relationship—in that it is only through that particular musical
structure that a successful communication of the kind intended can take
place—can come to appear in the score as merely arbitrary permutations of
‘notes’ and ‘time-values’. On the timeless flat surface of the score the
visual-spatial relationships of the notes (used to represent real time) may be
changed at will to produce arbitrarily arrived-at visual-spatial structures, all
having equal validity in visual space, but not necessarily so in experiential
time.
Once, however, we demand that music be heard in terms of the score,
then it is no longer experiential success which justifies notational visual-
spatial arrangements, but notational arrangements become their own
justification. Hence, ‘musical form’ may become freed from any restriction
of direct experiential success in our original terms. This leads ultimately to
a rational formalism in music. The composer establishes certain visual
relationships between entities in his notation, the musical scholar is trained
to listen for these relationships, he hears them and a successful ‘musical’
communication is declared to have taken place.
This beautifully closed rationalist view of music is the ultimate in
scribal sophistication, it is complete and completely unassailable in its own
terms. Music is hence completely socially definable and musical success
may almost be measured with a slide rule. How much more tidy and
convenient such a norm-adherent view of music than one bringing in the
messy business of inter-personal, yet unverbalisable, gestural dialectics.
The rationalist view of music fits ideally into a technocratic age with its
linguistic and positivist ideologies. What we cannot talk about we cannot
know, only that which we can talk about is real—so much for music!
Thus, ultimately, the score becomes its own rationale. It is what it is,
and there is nothing more to say about it. The composer cannot be in
error.13 We see this spatial score-based focus in preoccupations with such
two-dimensional visual forms as the golden section in analytical articles,
but it is permutationalism which is the ultimate notation-abstracted
procedure. Because musical notation presents music to us outside of time in
an essentially two-dimensional scannable score, it does not seem
immediately unreasonable to extract various parameters of the sound and
arrange these into various other patterns. The most thorough-going way of
going about this is the technique of the permutation of parameters as used in
much serial composition. The technique of permuting objects is very
general and is in fact a principle of ordering which does not relate to the
materials being permuted directly. We may permute pitches, dynamic levels
or, for that matter, sizes of shoes, using exactly the same criteria.
Applications of the principle can be very sophisticated, based upon analysis
of the nature of sound-objects. The principle problem from our point of
view is that being an outside-time procedure, there is no reason why the
resulting sequences of sounds should have any dynamism. The parameters,
separated through a lattice-based conception of musical structure, cease to
have any meaningful linkage or gestural continuity and serve merely as
evidence that the permutational procedure has taken place. This abstract
architecture, therefore, reduces all objects which it touches to the same
rather empty non-dynamic experience. There is no rationale beyond the
arrangement of the bricks; the nature of the bricks becomes irrelevant so
long as they fit into the pattern. The committed permutationalist is the
musical (or artistic) equivalent of the linguistic philosopher. He or she
cannot understand that there is a problem inherent in this approach.
A much more sophisticated and satisfactory approach can be seen in
the work of Brian Ferneyhough. Ferneyhough is clearly (from my listening
to the music) concerned with musical gesture and in a piece such as his
Second String Quartet the interaction of musical gestures between the four
players is of primary importance in our apprehension of the music. In works
for a greater number of performers, however, (such as Time and Motion
Study III for sixteen amplified voices) the sheer density of musical gestures
leads to a process of self-cancellation. The individual details of each part
are extremely interesting but an overall sense of direction is lost in the
welter of interaction. In 1981 I had the pleasure of meeting Brian
Ferneyhough over dinner in Paris and the ensuing conversation may serve
as an interesting footnote to our discussions of idealism and materialism in
the conception of music. Ferneyhough and myself both declared that we
were anarchists but on further discussion it transpired that our conceptions
of anarchism could not have been more different. Ferneyhough's view was
that he could take the strongest stand against the system by not voting. In
this way he symbolically denied the relevance of the system and therefore
in some way negated it. My more pragmatic view was that it was important
to vote in order to keep out the worst possible contender. These conflicting
idealist and materialist views of anarchist action had an interesting parallel
in our discussions about musical structure. Ferneyhough noted that a
particular passage in one of his works sounded pretty well aleatoric and that
this was interesting because it was the result of a multi-layered process of
complex compositional decisions. He seemed to be saying that the
methodology of composition was the principle object, not the effect on the
listener. The composition is more like a document which evidences the
composer's methodology and it is evident in the particular case under
discussion that the methodology will only become apparent through
detached analytical study of the document, not directly through the effect of
the music. Thus a priori design, not directed pragmatically to some
practical sonorous end, has become the principal focus of the composer's
interest. The concept of musical experience has been redefined as
rediscovering the composer's methodology through the musical experience
(or rather through the score) rather than feeling the gestural structure in
time of the music in the listening experience, and hence directly
understanding, through the gestural experience, the composer's design.
For me, on the other hand, a musical experience which appears
aleatoric is aleatoric. The experience that the listener has is the music and
the composer's methodology, no matter how rational it may be, is a heuristic
device realising the source of that experience in sound. In Ferneyhough's
case it would seem that music is an idealist object defined essentially by the
composer's intention (just as the political stance is defined by the intention
of the act of not voting). In my case, music is a material entity which is
socially defined and judged by its results (similarly the political act must be
an action taken in the world that will be judged by its success there). This
being said, one must not confuse materialism with populism, but that is the
subject of another essay and I will not pursue it here.
A fundamental thesis of this book is that, in order to understand and
control the musical continuum, we will have to appeal to time-based
notions like gesture and not only at the level of the individual musical
event. Although a formalist, permutationalist approach can be applied to
literally anything, including a particular classification of gestural types, we
cannot ultimately learn anything from it because it is not open to any
experiential verification (except in the tautologous sense that it evidences
the permutations made). What I am searching for in this book are criteria
for composing music with non-lattice materials which ‘work’ in some
experientially verifiable sense that is not merely circular.
A final comment: it is clear that the separation of notation and
actuality is of great value for the purposes of scholarship, even though it
does lead to a distortion of our understanding of the object of study. The
advent of digital recording and analysis of sound opens up a wonderful new
opportunity for such scholarship. In one sense it can be very negative as this
is a heaven-sent opportunity for formalism to run riot with a new ultra-
powerful permutational tool. To date, however, computer technology seems
to have been used in a much more sensitive way in the exploration and
understanding of the inner details of sound architecture. The preliminary
results of this exploration have been a source of inspiration for this
particular book and the control which the computer will give us over this
inner architecture makes the control of the details of gestural structure a
compositional possibility for the first time. With musical sensitivity we may
allow the computer to do the number-crunching and with real-time, or at
least acoustic, feedback we can begin to make more refined aesthetic
decisions about the gestural structure of sound on its most refined levels.

1 Though various scribe-philosophers and aesthetes have attempted to declare that music is
essentially abstract—we shall return to this point later.
2 An interesting example of this is to be found in the neumic notation of Tibetan chant (see
Kaufmann 1967), in which a single curvilinear neume might indicate changes in pitch, duration and
timbre. See Figure 2.2a.
3 A fuller discussion of these issues is to be found in Goody and Watt (1963).
4 See the quote from Lejaren Hiller in Chapter 1 (Hiller 1981: 7).
5 This situation has only partly been alleviated in the years since this was written and was certainly
reinforced by the Midi protocol, relatively new at the time of writing (Ed.).
6 Contemporary instrumental composers have, of course, sought to counteract the stranglehold of
technological rationalisation by exploring non-conventional modes of sound production on the
Western instrument (flutter-tonguing, key-slapping, whistle-tones etc.).
7 Bach's Art of Fugue, arguably one of the finest achievements of the traditional art music of Europe,
illustrates our thesis in an interesting way. Bach confines himself to the notation of pitch and
‘summative’ rhythm, leaving unspecified dynamics and even timbre (instrumentation), both of which
are usually notated (or at least indicated in the score). Although this approach may appear to
approximate very closely to the ‘abstract’ view of music, we would argue that the work is, however,
not an illustration of ‘rational formalism’ as discussed below, as the score notates sets of relations
between sound-qualities which are experientially valid (see text), even if the range of possibilities is
necessarily restricted by the nature of the notation system itself.
8 See the quotation in Chapter 1 (Boulez 1971: 37).
9 Figure 2.6 is from Trevor Wishart's PhD Thesis (University of York, UK 1973).
10 Such as paper bags, soft trumpets (Martin Mayes, Trevor Wishart), amplified springs (Hugh
Davies), or long pieces of elastic (Paul Burwell) (see discography).
11 For an explanation of Stockhausen's ‘+/-’ notation system and some examples of its application
see the introduction to the score of Spiral (Universal Edition).
12 I am indebted to Jan Steele for the following line of argument concerning the problem of musical
retrogrades.
13 Although of course, not all composers, even today, accept this absurd view! There are often other
criteria involved in composition, even where composers refuse, in a strictly positivist way, to talk
about them.
 Chapter 3
PYTHAGORAS, FOURIER, HELMHOLTZ:
TOWARDS A PHENOMENOLOGY OF SOUND

Pythagoras, that grave and venerable personage, reproved all judgement of Musick which is
by the eare, for he said that the intelligence and vertue thereof was verie subtile and slender,
and therefore he judged thereof, not by hearing, but by proportionall harmonie: and he
thought it sufficient to proceed as farre as to Diapason, and there to stay the knowledge of
Musick.
[Plutarch]1

Pythagoras: stable vibrations and the Harmony of the Spheres

The preceding chapter sought to account for the ideology of musical praxis
which sees pitch and duration as primary musical qualities, timbre as a
distinct and secondary musical quality and takes instrumental streaming and
the generation of music on a lattice for granted. The philosophy of the
musical practice based upon establishing elementary relationships between
stable vibrating systems has a very ancient and respectable pedigree.
Pythagoras himself is believed to have first noted the fact that there is a
simple relationship between the lengths of vibrating strings and the
perceived quality of musical consonance between them. Given two strings2
of the same material at constant tension then if one is stopped exactly half-
way along its length it will produce a note sounding one octave above the
other string (which is not stopped). The octave itself is qualitatively
perceived to be the most stable or consonant of music intervals. With a
length ratio of 3:2 the musical interval of a fifth is perceived which is also
very stable and consonant. In general, the relative consonance of an interval
is seen to be directly relatable to the simplicity of the ratio of lengths of the
strings (or columns of air) which produce it (see Figure 3.1). This was not
only the first important contribution to music theory but also had a
significant role to play in the development of a scientific view of the world.
It was the first clear demonstration that qualitative aspects of nature could
be reduced (apparently) to simple numerical relationships. This general
view has been exceedingly fruitful but on occasions misleading. It led
indirectly to the concept of the Harmony of the Spheres, one of the most
persistent and misleading conceptions ever to animate the human mind: the
heavenly bodies were assumed to be transported around the earth on giant
spheres whose motion generated a heavenly music, governed in some way
by the Pythagorean laws of proportion. The desire to find ‘celestial
harmony’ in the world of nature persists even into the work of the
astronomer Kepler, who was obsessed by the desire to fit the (assumed)
spheres of planetary motion (Figure 3.2) around the Platonic solids (Figure
3.3).
 Figure 3.1 Relation between string length and ‘consonantly’ related tones.

It is important to point out that, important though these elementary

physical relationships are in the underpinning of various musical languages,
music in general is a cultural construct. Although the primacy of the octave
and the fifth is preserved in most musical cultures (though not everywhere),
we would have great difficulty in explaining away all the subtle
ramifications of the North Indian scale system in terms of Pythagorean
interval theory. More significantly, perhaps, for the subject of this book, the
Western tempered scale in fact preserves only the true octave in its
structure. All other apparently simple-ratio intervals (such as the fifth,
major third and minor third) are mere approximations to the simple
Pythagorean ratios, the actual ratios used being governed by the rationality
of tempering and the twelfth root of two! Although in Western tonal music
the fifth plays a central role (next to the octave), this cannot be put down
merely to its Pythagorean simplicity. The well-tempered ratio ((12√2)7 (=
1.48):1) is in no sense simple and it is difficult to see in what sense it
approximates ‘simplicity’—are we to say that it is simpler than the ratio 4:3
because it is closer to 2:3?—in that case, would not 2:π be simpler than 4:3?

Figure 3.2 The planetary and celestial spheres.

Figure 3.3 Kepler's attempt to fit the Platonic solids within the planetary orbits (assumed spherical).

One further point: Pythagoras’ theory essentially establishes

relationships between simple, stable vibrations. Current (and in fact ancient)
musical practice is not solely concerned with such sound phenomena and
the advent of computer analysis and synthesis permits us to understand and
control much more complex sound phenomena.

Helmholtz, Fourier and Standard Musical Practice

The next major breakthrough in our physical understanding of the nature of

sound came with an important discovery by the mathematician Fourier.
While attempting to solve various problems relating to the conduction of
heat in solids, Fourier discovered that it was possible to represent an
arbitrary mathematical function by a sum (possibly infinite) of simpler
functions. The simpler functions which he chose for this representation
were the elementary sine and cosine functions with which we are all now
familiar from work in acoustics or electronic music. It is possible to give an
approximate description of the Fourier method without going into
mathematical details to anyone familiar with the concept of vector.
A vector may be regarded as a line of a particular length and a
particular direction. In three-dimensional space we may affix to one end of
this line a set of three lines at right angles to one another, a system of
coordinates. The point where these lines meet is called the origin. If our
vector starts at the origin no matter where its other end point is, it is always
possible to reach that end point by proceeding from the origin a certain
distance parallel to one axis, a certain distance parallel to the second axis
and a certain distance parallel to the third axis (see Figure 3.4). These three
new vectors are called the components of the original vector. Roughly
speaking the fact that the three components are at right angles to one
another means that they are independent of one another (one cannot express
any of the components in terms of the others). The components are then
said to be orthogonal.
If we now take an arbitrary function (see Figure 3.5), this associates
with every point along the horizontal axis a point on the vertical axis. We
now need to make a leap of the imagination and imagine that every
infinitesimally small point along the horizontal axis corresponds to a
dimension in an infinite-dimensional space and that the corresponding value
on the vertical axis corresponds to a distance along that particular
dimension. We can now, at least conceptually, represent the entire function
by a single point in this infinite dimensional space. If now, just as in the
case of the vector in three-dimensional space, we can set up a system of
orthogonal coordinates and define a set of related components in the space
(which in this case will turn out to be other mathematical functions) then we
can make a representation of the original arbitrary function. The set of sine
and cosine functions used by Fourier, can be shown to fulfil the criterion of
orthogonality. Hence we have discovered a very powerful mathematical
tool. It soon became clear that Fourier's method was especially applicable to
the description of sound phenomena. As is now well-known, any sound
phenomenon, no matter how complex, is carried by variations of pressure
within the air and can be represented as a function of air-pressure against
time. Such functions are, at least in principle, directly amenable to Fourier's
method of analysis.
 Figure 3.4 Representation of vectors in a coordinate system.

Figure 3.5 Representing an arbitrary function as a point in infinite-dimensional space.

Helmholtz and others, working with what they called ‘musical’ tones,
i.e. the sounds of conventional musical instruments, proposed a simple
theory of pitch and timbre perception. A sound perceived as a single pitch
was found to be made up of various sine wave components (through Fourier
analysis). These bore a simple harmonic relationship to one another. The
frequency of the higher sine tones were integral multiples of the frequency
of the lowest (which for the moment we will assume to be the fundamental)
frequency. The pitch of a sound corresponded directly with the frequency of
the fundamental, the timbre was the result of the presence (relative
amplitude) or absence of the other sine tones (the partials). Before going on
to criticise and comment upon this theory, we should note that it seemed to
absolutely confirm the ruling musical ideology that pitch was primary and
timbre secondary. Pitch could be seen as fundamentally related to frequency
and timbre as merely a secondary phenomenon arising from the
combination of the frequencies of the constituent sine tones. Timbre
appeared to be thus almost a fused chord over a fundamental pitch.
However, the fact that Helmholtz's theory appeared to confirm the ruling
musical ideology should come as no surprise. It was framed within a culture
which took that system of musical thinking for granted. Helmholtz confined
himself to the analysis of what he arbitrarily defined to be ‘musical’ tones,
i.e. sounds forced onto the pitch-timbre-duration lattice by preconceptions
of the musical and their realisation in instrument technology. Furthermore,
the assumptions firstly that timbre was a unitary phenomenon and secondly
that pitch and timbre were clearly separable qualities, were taken for
granted directly from preconceptions of music tradition.

Walsh functions, indeterminacy and the missing fundamental

The first question we must ask about Fourier analysis is, although it is
clearly a very powerful mathematical tool, does it bear any relationship at
all to our perception of sonic reality? Is the Fourier analysis of sonic events
into sine tones unique, or is there any other alternative analytic breakdown?
It turns out that other systems of orthogonal functions can be defined and
used to represent arbitrary mathematical functions. One such system is that
of Walsh functions illustrated in Figure 3.6. With the advent of digital
technology some programs have already been developed for the synthesis
of sounds using Walsh functions rather than the more usual sine tones.
However, whereas the Walsh function analysis of a sound-object seems to
bear no clear intuitive relationship to our aural experience, Fourier analysis
relates very clearly to what we hear. It has been shown in fact that the
human ear is a kind of Fourier analyser so that we may assume that up to a
point the mathematics of Fourier analysis has some direct relationship with
our perceptual experience.
The result of Fourier analysis is what is called a Fourier transform.
The mathematics of the Fourier analysis convert information about the
variation of amplitude with time (time-domain information) into
information about the variation of amplitude with frequency (frequency-
domain information). Simply put, we start off with a graph of amplitude
against time and we end up with a graph of amplitude against frequency
(see Figure 3.7). The inverse Fourier transform performs the opposite
function, turning information about frequency and amplitude into
information about amplitude and time.

Figure 3.6 The first eight harmonics (sine waves) and the first eight Walsh functions.
 It must be said immediately that the notion that somehow frequency
(periodicity) is more physically real than spectral information is hard to
justify. In a simple instrumental tone, periodicity can certainly be more
easily seen from a graph of amplitude against time than can any spectral
information. However, this is partly the nature of the beast being analysed
and when we consider more complex musical objects (see the section on
noise below) we will find that the graph of frequency against amplitude (i.e.
the spectrum of the sound) is far more lucid than the amplitude against time
graph (which may be totally aperiodic). In fact, to be entirely reductionist
for a moment, all that really exists is the amplitude of displacement of the
air or the ear-drum and its variation in time. Both periodicity and spectral
information are higher-level derived entities.
A more important problem is simply that the ear is unable to function
as a Fourier analyser above frequencies of around 4,000 Hz although
frequency information above this threshold can be very important in our
perception of timbre. In relation to these I quote from Schouten, Ritsma and
Cardozo:

Figure 3.7 The Fourier transform and its inverse.

[...] there may exist one or more percepts (residues) not corresponding with any
individual Fourier component. These do [however] correspond with a group of Fourier
components. Such a percept (residue) has an impure, sharp timbre and, if tonal, a pitch
corresponding to the periodicity of the time pattern of the unresolved spectral
 components.
(Schouten, Ritsma and Cardozo 1962: 1419, emphasis added)

Put simply, above 4,000 Hz we hear timbre simply as timbre!

A more serious problem with the ideologically natural view which
equates sine tones with ‘pure pitches’ and their combination with timbre is
related to the principle of indeterminacy. Towards the end of the nineteenth
century a serious problem arose in the theory of the absorption and
emission of radiation by heated bodies. Put simply, conventional theories
seemed to predict that any body in thermal equilibrium should radiate
infinite amounts of energy in the ultraviolet region of the spectrum and this
was consequently known as the ultraviolet catastrophe. The problem was
eventually solved by Planck's introduction of the quantum of action. The
Quantum Theory is now central to contemporary physics and its central
assumption is that energy can only be emitted in small finite packets
(known as quanta). The size of these energy quanta (E) is directly related to
the frequency (f) of the oscillator emitting or absorbing radiation by the
formula E = hf/2π where h is Planck's constant.
An important cornerstone of quantum mechanics is Heisenberg's
Uncertainty Principle. This states simply that the position and momentum
of a particle cannot both simultaneously be known exactly. In fact, the more
accurately one is known the less accurately the other must be known. A
similar relationship holds between energy and time, i.e. it is possible to
know the energy of a particle with great accuracy only if one does not know
exactly at what time it has this energy. The Uncertainty Principle is often
presented in elementary books about modern physics as a result merely of
the interference of the instruments of observation with what is being
observed. It is after all more natural for us to assume that the particle does
have a definite energy at a particular time and it is merely a problem of
getting to know what that is. The conventional view among physicists,
however, is that the Uncertainty Principle is intrinsic to the nature of reality
rather than a mere accident of experimental design. It can be shown that the
mathematics resulting from an assumption of the truth of the Uncertainty
Principle has observable physical consequences which would not be
expected if the uncertainty were not an intrinsic part of nature (for example,
so-called exchange phenomena between electrons in chemical bonding).
 Energy/time indeterminacy (which has a direct relationship to our
discussion of the structure of sounds) can be understood fairly simply. First
of all, we must remember that in Quantum Physics the energy of a system is
related to its frequency. How can we therefore measure the frequency of a
system at a particular instant in time? The answer is simply that we cannot
because frequency is a property of the system dependent on its actual
evolution through time. We can say that a system vibrates five times in a
second but we cannot talk about how many vibrations it undergoes at an
instant in time. Hence the instantaneous energy of a system is not definable.
Fourier analysis of a signal into its spectral components is similarly limited.
The mathematics of Fourier analysis assumes that the signal persists for an
infinite length of time. If it does not, even if it appears in the time domain to
be a pure sine tone, in the frequency domain it will be found to have some
spectral colouration. The reader may intuitively grasp why this is so by
looking at Figure 3.8. Thus when only a very small part of the curved edge
of a sine wave is present we have no way of predicting that this will
continue as a sine wave or that it will prove to be the leading edge of an
essentially stepped signal with sinusoidal rounding (for example). Even
when we have a complete cycle of a sine wave we do not know that this is
one complete cycle of the wave. It may be only the opening formation of a
more complex pattern. Even when we have two cycles of a sine wave,
though now we can perceive a regularity in the structure, it is still not
certain that we have the complete picture. As more and more cycles of the
sine wave are taken into our sample, the uncertainty in the nature of the
signal reduces rapidly but it only reduces to zero when the sine wave
persists to infinity.
 Figure 3.8 Indeterminacy in the analysis of brief signals.

The physical consequence of this is that if we produce a single cycle

of a sine wave (no matter how ‘pure’ it may be) what we will in fact hear is
a click, a sonic impulse whose frequency is maximally indeterminate. This
is not as some of the Groupe de Recherches Musicales writings seem to
suggest just a limitation of the ear. It is intrinsic to Fourier analysis itself
and therefore any physical instrument performing a spectral analysis on the
signal would register a similar indeterminate result. Hence a sine tone, no
matter how ‘pure’, if sufficiently brief has no definite pitch! This
phenomenon is illustrated in Examples 3.1 and 3.2. The same melody is
played in both examples but in the first the individual elements are less than
five milliseconds long, whereas in the second example they are ten
milliseconds long. In the first example no sensation of pitch is conveyed.
For similar reasons, if we suddenly switch on a pure sine tone and
then suddenly switch it off we will experience (and analysis will confirm) a
spreading of the spectrum at the start and end of the sound. In this case we
are not talking about switching transients in the apparatus which generates
the sine tone (which we may assume we have eliminated) but the inevitable
results of Fourier analysis of a finite signal. This frequency spreading can
be reduced by using a cosinusoidal envelope on the attack and decay of the
sample (see Figure 3.9).
 It will no doubt have occurred to the reader that no real sounds last
for an infinite length of time! Furthermore, few sounds exhibit constant
periodicity for any length of time. Also, a sound does not have to be
random to have a constantly changing wave-shape. A simple example
would be a portamento on any instrument. Any practical Fourier analyser
(including the human ear) must do a kind of piece-wise analysis of sound
and link up the results in order to gain any reasonable conception of the
nature of the complex sound-world surrounding us.
Having discovered that ‘pure’ sine tones may have no definable pitch,
we may also discover that clear pitches may not be associated with the
appropriate sine tone! The simplest example of this is illustrated by first
recording a low note on the piano (Example 3.3) and then filtering out all
frequencies below the second harmonic,3 i.e. filtering out what we assume
to be the fundamental pitch (Example 3.4). Amazingly the filtering has no
effect on the sound whatsoever and we are led to conclude that there is no
single component in the sound corresponding to what we hear as the
fundamental pitch! The fundamental is in fact a mental construct
extrapolated from the information contained in the higher partials. This
effect is not peculiar to piano tones and can be demonstrated for other types
of sound.

Figure 3.9 Spectral spreading in finite signals.

 Most spectacular of all (Example 3.5), using extremely precise
control of the overtone structure of a sound through computer synthesis, we
can trade off the increase in perceived ‘fundamental frequency’ against the
weighting of high and low partials in the sound in such a way that the pitch
of a sound glissandos upwards at the same time as the sound descends
gradually to the lower limit of the perception of pitch! The sound appears to
ascend and descend simultaneously. Such experiences lead us inexorably to
the conclusion that pitch is an aspect of the perception of timbre, and not
vice versa. At the very least, we should become aware that the strict
separation between pitch and timbre is an artefact of the way we have
constructed our musical reality.

The inharmonic and the non-periodic: a dual conception of pitch

So far we have confined ourselves largely to Helmholtz's material, i.e.

sounds of stable tessitura in which the partials are integral multiples of the
frequency of the real (or imaginary) fundamental. When the relationships
between the partials of a sound are not of this kind (i.e. when they are
inharmonic) the Fourier analytical attempt to extract a single pitch
characteristic from the sound breaks down. The object appears to perception
as an aggregate of various pitches, more fused than a typical chord in
instrumental music but definitely not a singly-pitched note. Many bell and
bell-like sounds are of this type. Composers working in electro-acoustic
music seem to be particularly attracted to these kinds of sounds and one
speculates whether it is not due to their similarity to chord structures in
conventional music. The sounds, in fact, are not radically different from the
normal sound-objects found in conventional musical practice. They lend
themselves to similar (lattice-based) modes of organisation and do not
challenge our conception of what is and is not a musical object (as sounds
with dynamic or unstable morphologies will do). At another perceptual
extreme we have entirely non-periodic signals. In the architecture of typical
analogue synthesisers (and in much discussion of electronic music) such
sounds, usually referred to as ‘noise’, are often treated as entirely separate
entities from materials with clearly defined spectra usually generated from
simple oscillators. In fact, as we shall discuss, there is no simple dividing
line between periodic and non-periodic signals, but in fact a
multidimensional array of complex possibilities between the two extremes.
Noise is not something to be treated separately from other materials, either
compositionally or conceptually, but an alternative way of perceiving and
relating to sound phenomena which we shall now discuss.
As discussed earlier, the result of Fourier analysis is to transform
information about a sound from the time domain into the frequency domain.
The most immediately striking thing about noise-type sounds is that the
time domain criteria apparently cease to apply. For example, if we take a
recording of a typical periodic pitched sound or even an inharmonic sound
and we play back the recording at double speed the time domain
information passes us at twice the rate and we hear the sound transposed up
an octave. If, however, we try the same experiment with white noise, we
experience no change in the frequency domain information (Example 3.6).
More complex signals may yield even more startling results (for instance in
Example 3.7 where a complex sound is played first at normal speed and
then at double speed. The sound appears to shift by about a third).
The explanation for this lies in the characteristics of noise-type
sounds and the way we perceive them. For noise-type sounds the time
domain information is essentially random. The amplitude of the signal
varies randomly with time and, for example, for Gaussian noise, the
probability that the amplitude will have any particular value at any
particular time is maximum for zero amplitude and dies away smoothly to a
probability of zero as the amplitude increases. A Fourier analysis of a small
sample of a noise signal will yield a somewhat arbitrary array of
frequencies and amplitudes. Analysis of another similar sample will yield a
quite different but equally arbitrary array of frequencies and amplitudes.
The only way that we can achieve a coherent analysis of this signal is
through a statistical averaging process. We can show the average amplitude
of the various frequencies in the signal over a long period of time. For white
noise the frequency domain analysis reveals a completely flat spectrum. All
frequencies are present with equal probability. This explains the
phenomenon illustrated in the example. When the tape is played at double
speed, all the frequencies are shifted up an equal amount and those that
were lost near the bottom of the spectrum are replaced by even lower
frequencies (which, due to the physiology of hearing we might not even
have been able to perceive previously).
There are two things we must bear in mind when considering the
nature of noise-based signals. First of all by declaring that a typical noise-
based signal contains all possible frequencies at random amplitudes
distributed randomly in time, the typical text book on acoustics tends to
imply that noise is essentially an aggregate of more elementary sounds.
This conception is an artefact of our perception of the analysis. As anyone
who has switched on a synthesiser will know, noise is a perfectly coherent
source, no less coherent than a typical sine wave oscillator and need have
no particular granular characteristics. What we are saying here applies
particularly to white noise but the noise concept (involving time-averaging
of a spectrum) may be applied to a great many sound-objects, and noise-like
objects may certainly be produced through the aggregation of simpler
objects. For example a dense cluster in the lower register of the piano has a
fused noise-like quality quite different from a similar cluster played in the
high register. Secondly, and most important, noise involves a different mode
of listening. In fact, it might be more accurate to suggest that the noise
conception is a property of the way we hear rather than of the object itself.
This distinction might seem arbitrary when we consider only oscillators and
noise-sources on a typical synthesiser but when we begin to consider
sounds with extremely complex evolving and semi-irregular spectra, the
distinction comes strongly into play.
With sounds therefore that we describe as noise, or noise-based we
appear to respond directly to information in the frequency domain as there
is little information to be gained from the time domain. This fact may be
illustrated in two ways. In Figure 3.10 we see time domain representations
of Gaussian and binary noise. The Gaussian noise has the random
amplitude fluctuations we might expect. The binary noise, on the other
hand, flips randomly backwards and forwards between two very definite
amplitudes. However, the frequency domain information can be the same in
both cases and we can perceive both as the familiar ‘white’ noise. In
Examples 3.8 and 3.9 we hear first of all the effect of transposing white
noise by an octave (no effect on the frequency domain information) and
secondly a ‘melody’ created by filtering variously the white noise to
produce bands of different ‘coloured’ noises. Again the effect of the latter
on the time domain information would be barely perceptible but can be
clearly heard in the frequency domain. Perhaps the most striking example
of the independence of frequency domain hearing from time domain
information in noise perception is the experience of comb-filtering. In this
particular case a brief delay is imposed on the noise signal and the result
mixed with the original signal. The time domain representation presents us
with no perceivable patterning but in the frequency domain we discover that
various regular peaks and troughs appear in the spectrum, the spacing of
these being related to the duration of the time-delay. Perceptually speaking,
a pitch or particular spectral characteristic is imposed upon the previously
undifferentiated noise source (see Figure 3.11).

Figure 3.10 Random and binary noise (independence of frequency-domain representation).

Filtering noise can be used to produce broad bands which can be

perceived to be higher or lower than one another but are not perceptibly
pitched or to produce narrower bands which are perceived as being more or
less clearly pitched. This suggests that there are at least two conceptions of
pitch involved in our perception of sound-objects. The first type of pitch is
related to periodicity and arises from the real or implied fundamental
frequency of vibration of the source (or several of these in the case of an
inharmonic sound). The second results from the imposition of a spectral
envelope through some sort of filtering or resonance procedure acting on a
noise-averaged spectrum. Just as the perceived fundamental pitch may be
multiple (in an inharmonic spectrum) the spectral envelope itself may be
quite complicated and we will normally refer to this as a formant structure.
These two conceptions of pitch (in a sense one related more strongly to the
time domain and the other more strongly to the spectral domain) meet most
strikingly in the sound production of the human voice where the pitch of
standard musical practice is defined by the fundamental vibrations of the
larynx while the articulation of vowels in speech (and in song) is defined by
varying formant structures.
We may now imagine very complex tones in which the two concepts
of pitch are traded off against each other. For example, an inharmonic tone
in which the various partials are in constant and rapid motion might be
given a focused pitch sense by the strong accentuation of a narrow formant
band using a filter. Conversely and more commonly heard, the time domain
pitch of a sung tone is clearly maintained through the most extreme
variations of formant structure details. The question of what happens when
both of these change at once will be dealt with in the following chapters!
 Figure 3.11 Comb-filtered noise (perception of pitch direct from frequency domain).

Spectral evolution: harmonic and dynamic timbre

A much more fundamental break with Helmholtz may be made. Simple

experiments with sound-objects demonstrate that timbre is not merely
dependent on spectral information but also upon the way that information
evolves through time. The most striking illustration of this fact is given in
Examples 3.10 and 3.11. In Example 3.10 we hear first the sound of a piano
note whose envelope has been smoothed and flattened. Following this is the
sound of a flute playing the same note. The two are virtually identical. In
the second example (3.11), we first hear the sound of a flute on which has
been imposed the amplitude envelope of a piano note. This is followed by a
piano note at the same pitch. Again, the two sounds are indistinguishable.
The fact is that in this particular register the spectral content of piano and
flute sounds are very similar. What differentiates these sonic objects is the
temporal evolution of the amplitude of the event. The flute remaining
relatively constant whilst the piano decays linearly4 from a sharp attack.
Thus dynamic aspects of the spectrum enter into our perception of timbre,
so we see that timbre is at least a two-dimensional entity contrary to the
conventional wisdom.
Furthermore, the characteristics of the evolution of the amplitude
envelope are of fundamental importance. If we take a sound of relatively
constant envelope and edit away various parts of it, we experience no
noticeable change in timbre (Example 3.12). Extending this notion we need
not keep the amplitude constant but merely maintain a constancy in its rate
of change. (A constant amplitude is simply an amplitude envelope with
constancy of change, the rate of change being zero.) With a piano, the
sound actually dies away but the rate of this decay is constant (linear
decay). If we therefore edit off various amounts of time from the beginning
of the piano note, its timbre will not be noticeably changed (Example 3.13).
If we try the same experiment with an instrument whose amplitude
envelope varies non-linearly, we discover that editing off the beginning of
the sound changes the timbre distinctly (for instance, Example 3.14 using a
bell). This mode of argument applies equally well to unpitched sound and
may be heard by listening to Example 3.15 which uses a cymbal having a
linear decay with noise-type material.
To be even more precise we must take into account the evolution of
amplitude of each component of the spectrum in our sound. If we repeat the
same experiment using a vibraphone (Example 3.16), we will find that
when we lose the very start of the sound it is altered significantly, but
editing off more of the sound's beginning has no further appreciable effect.
This is simply because there is a rapid spectral change during the initiating
moment of a vibraphone note, due to the metallic sound of the hammer
striking the key but then the subsequent resonance dies away linearly (as
with a piano note) because of the material and design of the instrument.
Even a conventional instrumental sound may contain significant
noise components. In Examples 3.17 and 3.18 we first of all hear a flute
sound without its initial 50 milliseconds (followed by the original
recording) and the same experiment repeated with a trumpet. The effect is
very much more marked with the flute because the flute sound is partly
characterised by a noise-based breath-sound which initiates the resonance
of the tube. Thus, even with conventional instruments, we begin to see a
sequential breaking up of the characteristics of the sound-object which we
will explore more fully in the section on multiplexing and which is more
typically characteristic of speech-streamed sound-objects. The discovery
that timbre itself is partly dependent upon the evolution of spectral
characteristics is our first real link with sounds of dynamic morphology, i.e.
sounds in which the perceived pitch spectrum, amplitude envelope etc., all
evolve through time.

Coherence of sound-objects; aural imaging

Having discovered that sound-objects may be exceedingly complex and that

our perception of them may involve processes of averaging and attention to
spectral evolution, an obvious question presents itself; how are we ever able
to differentiate one sound-source from another? As all sounds enter our
auditory apparatus via a single complex pressure wave generated in the air
why do we not just constantly hear a single source with more or less
complex characteristics? We might ask the same question in reverse: how is
it that a complex sound does not dissociate into a number of separate aural
images? Much research has already been done and much is still being
carried out on the problem of aural imaging. The following observations
are drawn from the work of Steven McAdams at IRCAM in Paris (see
McAdams 1982).
To phrase the question a little more technically, once our auditory
mechanism has dissociated the incoming sound into its constituent sine
wave components, or at least generated some kind of spectral analysis, how
can it then group the various components according to the separate sources
from which they emanated? There appear to be at least four mechanisms in
operation here. These are:

(1) components having the same (or very similar) overall amplitude
envelope, and, in particular, components whose onset characteristics
coincide will tend to be grouped together;
(2) components having parallel frequency modulation (either regular in the
 form of vibrato or irregular in the form of jitter) will be grouped
together;
(3) sounds having the same formant characteristics will be grouped
together;
(4) sounds having the same apparent spatial location will be grouped
together.

Any or all of these factors may enter into the process of separating one
aural image from another. The importance of onset synchrony is
demonstrated in Example 3.19 where the various constituents of a sound are
separated by increasing time-intervals and then the time-intervals
successively reduced until there is again complete synchrony. The sound-
image will be heard to split into its component parts and then recohere. The
importance of frequency-modulation information has been most eloquently
demonstrated in work by Roger Reynolds5 (Example 3.20). Data from a
phase vocoder analysis was used to resynthesize an oboe tone, elongating it
as well. The regenerated oboe tone was projected from two loudspeakers,
the odd harmonics on one side, the even on the other. These two groups of
partials were each coherently, but differently, frequency modulated.
Because the even set was modulated at a rate corresponding to vocal
vibrato, and the odd necessarily had a clarinet-like sound, the listener
experiences a distinctive composite as the amplitude of modulation
increases: clarinet on one side, voice on the other at the octave, and the
sum, an oboe sound, in the centre. In this way we can contemplate playing
with the aural imaging process and not merely destroying the convention of
instrumental streaming.
Conversely, we may use these aural imaging factors compositionally
to integrate sound materials which might otherwise not cohere into objects.
Thus, by imposing artificial attack and amplitude characteristics on a
complex of sounds (e.g. a small group of people laughing), by artificially
synchronising the onset of two or more normally quite separate sound-
objects, by artificially synchronising the vibrato and jitter on two or more
normally quite separate sound-objects we may create coherent composite
sound-objects. A recently popular example of this approach is the use of the
vocoder where the evolution of the formant characteristics of a speaking
voice is imposed on an otherwise non-coherent source (e.g. the sounds of a
large group of people speaking before a concert as in Michael McNabb's
Dreamsong). This further opens up our conception of what might be
considered a coherent musical object.
With all these potential sound-materials at our disposal a further
problem arises. Music is normally concerned with establishing relationships
between various kinds of material. The question is what determines whether
we perceive a particular piece of sound-material as related to another.
Speaking of sound organisation in the broadest possible sense, the answer to
this question will clearly depend partly on context (upon which aspects of
sonic organisation are being focused upon—pitch, spectral type, formant
streaming etc.). But whichever approach we take there will be a point
beyond which the manipulated sound-material will cease to have any
audible relation to its source. This can already be perceived in conventional
music where in some types of complex serial organisation, the concept of
the derivation of material from a source set becomes meaningless. In the
studio it is seductive to assume that a sound derived from another by some
technical process is hence derived from the original in a musical sense. A
simple example of this may be given as follows. Suppose that we start with
a sustained orchestral sound, the sound of a large crowd and the sound of a
stable sine tone. Let us now take each sound, put it on a tape recorder and
switch the tape recorder onto fast wind so that the sound accelerates from
speed zero to very fast and as the tape recorder reaches its maximum speed
fade out the sound to nothing.6 Having done this with all the sounds, let us
now speed them up to at least sixteen times their original speed. In what
sense are the resultant sounds related to the originals? What we perceive in
each case is a brief, high frequency glissando. Furthermore, and most
striking, all three sounds now appear very closely related whereas the
sounds from which they originated have no relationship whatsoever. At this
distance of derivation it is the overall morphology of the sound-structures
which predominates. We may learn two lessons from this. First of all, with
sound-objects having a dynamic morphology, it is this morphology that
dominates our perception of relatedness- unrelatedness, rather than spectral
or even more general timbral considerations. Secondly, if the organisation
of our music is to be based on the audible reality of the listening experience,
the music must be organised according to perceived relationships of
materials or perceived processes of derivation of materials. In order to
accomplish the former we need an analysis of sound-materials based upon
their perceived properties, a phenomenological analysis of sounds.
The phenomenology of sound-objects

The pioneering work on the development of a phenomenological

description of sound-objects and an aesthetic based upon it was done by
Pierre Schaeffer and the Groupe de Recherches Musicales. Certain aurally-
perceived characteristics of sound-objects have already been discussed
above and this section is intended to complement what has already been
said. It draws largely upon the GRM research incorporating ideas drawn
from the writings of Robert Erickson and Iannis Xenakis. Perhaps the most
important concept advanced by Pierre Schaeffer was that of the acousmatic.
This term was originally applied to initiates in the Pythagorean cult who
spent five years listening to lectures from the master, delivered from behind
a screen (so that the lecturer could not be seen) while sitting in total silence.
Acousmatic listening may therefore be defined as the apprehension or
appreciation of a sound-object independent of, and detached from, a
knowledge or appreciation of its source. This means not only that we will
ignore the social or environmental origins or intention of the sound (from a
bird, from a car, from a musical instrument, as language, as distress signal,
as music, as accident) but also its instrumental origin (voice, musical
instrument, metal sheet, machine, animal, larynx, stretched string, air
column etc.). The idea of acousmatic listening is easily appreciated by
anyone who has worked with recorded sound-materials in the electro-
acoustic music studio. When working with large numbers of sounds from
different sources and particularly when this material has been transformed,
if only slightly, it becomes difficult to remember from where the various
sounds originated and from a compositional point of view such origins need
have no special significance. The transformation of the flute tone into the
sound of a piano (above) illustrated this thesis, though a truly acousmatic
approach would demand that we forget not merely that the sound derived
from a flute but also that after its transformation it appeared to derive from
a piano! We should concern ourselves solely with its objective
characteristics as a sound-object. Example 3.21 illustrates the need for the
separation of sound-object description from any reference to the source.
The various sounds here are all derived from the same source-object (a tam-
tam excited by a variety of objects).
From our discussion in the previous section we became aware of a
distinction between sounds which are transposed by a change of replay-
speed (time-domain transformation) and sounds (like white noise) which do
not change their pitch, or change pitch in some unpredictable way under the
same transformation. We also discovered that various properties of a sound
may be altered by filtering. With certain sounds, however, though
qualitative changes can be perceived as a result of filtering, we do not feel
that the underlying sound-object has been fundamentally altered. The
instrumental tones of conventional musical practice are typical examples of
sounds which can be transposed by time-domain transformation and are
resistant to filtering. They are not the only ones, however, and it is
necessary to define a more general characteristic of sound-objects having
this property. Following the French terminology, we will refer to this as
mass. An example of a complex sound having a definite mass and
illustration of its resistance to filtering and its transposition are given in
Example 3.22. In actual practice, of course, there will be a large, grey area
where time-domain-based perception of complex timbres and spectral
perception of formants (see previous section) meet.
A second perceived characteristic of sound-objects is grain. If we
take a slowly repeating pulse and gradually speed it up beyond about 20 Hz,
we begin to hear a definite pitch and as the speed increases further we begin
to lose any sense of the original individual impulses. In between the
extremes of impulse perception and pitch perception we perceive a pitched
object with a certain amount of ‘grittiness’ as the individual impulses are
still apparent in some sense to our perception of the sound-object. This
internal ‘grittiness’ is the grain of the sound-object and is illustrated for an
electronic impulse (Example 3.23) and for a bassoon note (Example 3.24)
in the sound examples. In this particular case we are talking about a regular,
periodic grain, but it is possible to define grain in a more general manner. If
we have a sound made up of a large aggregate of brief impulses occurring
in a random or semi-random manner, we can talk about the statistical
average rate of occurrence of these impulses and this particular parameter
will have an effect on the perceived characteristic of a sound. Erickson
(1975)) refers to this characteristic as rustle time and might also be thought
of as aperiodic grain. It has an important role in our perception of particular
types of percussion sounds. For example, we may note the perceptual
difference between sounds of types of rattles filled variously with sand,
seeds and shot. Note, however, that the spectral characteristics of the
individual impulses in all these cases will also contribute (probably in a
statistically averaged way) to the perceived character of the resulting sound.
Aperiodic grain also has a bearing on the particular sonority of sizzle-
cymbals, snare-drums, drum-rolls, the lion's roar, and even the quality of
string sound through the influence of different weights and widths of bows
and different types of hair and rosin on the nature of bowed excitation.
It is important to realise that there is a perceptual threshold at which
we cease to perceive individual events as individual events and begin to
experience them as contributing to the grain of a larger event. As a result, at
sufficiently high speed, any sequence of sound-objects may become fused
into a larger object with grain. Example 3.25 illustrates this first for a
descending scale and then for an irregular melodic pattern. Incidentally, if
we applied the same process to a string of speech sounds (Example 3.26)
we approach the conception of a multiplex to be discussed in the next
chapter.
Combining the concept of rapid rustle-time with multiplicities of
brief sound-objects of various spectra we begin to define another huge class
of possible sound-objects (e.g. the sound of rain, of poured pebbles, etc.)
hinted at in Xenakis (1971/1992). This book is a very interesting early
analysis of a generalised notion of sound-objects as evolving groupings of
elementary particulate sounds. Unfortunately, the musico-descriptive
potential of the approach gets rather lost in Xenakis’ absorption in the
particular mathematical methodology (stochastic processes and Markov
chains) leading the composer off in a rather specialised aesthetic direction.
In the book, however, Xenakis does deal with the grain structure and
filament structure of dense sound-objects. Just as grain structure may be
thought to apply to sounds made up of elementary impulses, filament
structure applies to sounds made up of elementary sustained units. Figure
3.12 expanding on Xenakis’ descriptions, illustrates various possible
filament structures. If we now begin to discuss the temporal evolution of
such concepts as aperiodic grain and filament structure a whole new world
of complex sound-objects begins to open up before us. The GRM
classification goes on to discuss the concepts of complex-note, web and
eccentric sounds but I will discuss sounds of this and related types in a
somewhat different manner in the following chapters. A more complete
description of the GRM methodology can be found in the Solfege de l'objet
sonore (Schaeffer, Reibel and Ferreyra (1983)7).
 Figure 3.12 Examples of filament structure.

1 The author can no longer recall where he came upon this quaint translation into Elizabethan
English (Ed.).
2 The argument applies equally well to the lengths of columns of air.
3 Strictly ‘second partial’ as the piano spectrum is substantially inharmonic (Ed.).
4 ‘Linearly’ in terms of the perception of loudness (roughly speaking dBs); not amplitude which is
decaying exponentially with time. (Ed.)
5 Working with Steven McAdams and Thierry Lancino at IRCAM. This example is based on band 2e
on the IRCAM LP 0001 which is described somewhat differently on the sleeve note. It is from his
work Archipelago. (Ed.)
6 Unedited from 1983, this phenomenon may be simulated in digital systems (Ed.).
7 The Solfège is still available on three cassettes from the INA/GRM (Paris) at time of press but the
trilingual printout of the recorded French commentary which accompanied the LP version appears to
have been discontinued (Ed.).
 Chapter 4
THE NATURE OF SONIC SPACE

Structure of pitch-space; harmonicity and adjacency

It might be assumed, wrongly, from Chapter 3 that one only runs into new,
non-lattice-based, conceptions of musical ordering when dealing with
sounds having complex mass or noise characteristics. This is not the case,
as we shall explain in the next chapter, but we must begin exploration with
a closer analysis of the Pythagorean concept of consonance or harmonicity.
A simple reading of Pythagoras’ theory would seem to imply that an
interval is more consonant the simpler the ratio of the frequencies of its
components. In fact, we can define a measure for this simplicity as follows:
calculate the ratio of the frequencies, reducing it to the simplest fractional
representation (thus 600/350 becomes 12/7) now add together the
numerator and denominator of the fraction (for this example, 19). Confining
ourselves to intervals contained within a single octave this procedure gives
us a reasonable measure of the simplicity of a given frequency ratio.
It might now seem merely a simple matter to plot the simplicity of the
ratio (corresponding to the degree of consonance) against the interval.
However, anyone familiar with the difference between rational and
irrational numbers will be immediately aware of the following paradox. If,
for example, we take the interval of a fifth with the Pythagorean ratio 3:2
and hence a consonance value of 5 and we shift the upper tone by an
infinitesimal amount either upwards or downwards in frequency, the ratio of
the two frequencies making up the new interval immediately becomes non-
simple. In fact, in general, the ratio of the two frequencies will not be
expressible as the ratio of two finite integers. The simplicity value will in
fact be infinite. If we confine ourselves merely to ratios along the line
which are expressible with non-infinite integers, we will discover that the
simplicity-value leaps around in an extremely erratic fashion as we proceed
along the interval axis. This behaviour is illustrated for a limited number of
points in Figure 4.1. Both the rigorous and the approximate description of
the situation lead us to the bewildering conclusion that if we play a constant
tone and make an upward portamento on a second simultaneous tone
starting at the precise interval of a fifth from the first tone we should
experience a sense of very rapid shifts in consonance between the two tones
no matter how slow the portamento is made. This prediction is, of course,
only true if we stick rigorously to the Pythagorean view.
Figure 4.1 Harmonic ‘simplicity’ of intervals as predicted from Pythagorean theory and as
perceived.

Clearly, our actual aural experience is quite different. The ear is

aware that a particular interval is at, or close to, the fifth (ratio 3:2), that it is
moving away from the area of this ratio, and soon that it is approaching
another recognisably different simple ratio (e.g. 8/5, a ‘natural’ minor
sixth). We must therefore conclude that a second principle of relatedness
enters our perception of pitch. We will refer to the Pythagorean criteria as
the principle of harmonicity. The new criteria we shall refer to as the
principle of adjacency. This new principle states merely that two pitches
which are sufficiently close together will be heard as being related.
We can now give a new analysis of the original experiment of the
portamento of a tone against a fixed tone which corresponds more closely
to our aural experience. Whenever the frequency ratio between the two
tones approaches one of the simple (Pythagorean) ratios, we perceive a
harmonic relation between the two tones. As the sliding tone moves away
from this tone we perceive the interval between the two tones as being
closely related to the previously perceived interval. A sense of distance
from the interval increases until we experience moving close to a new
simple frequency ratio. Otherwise our experience of any continuous pitch
motion would be hopelessly atomised and incoherent.
The principle of adjacency helps to explain how it is possible to
construct systems of scales which do not use the precise Pythagorean ratios
and yet function as if they do! For example, the interval of a fifth in the
tempered scale does not correspond to the simple ratio 3:2, yet the whole
system of tonality founded upon this scale system assumes that this
approximate fifth is one of the simplest, most consonant, intervals in the
system, thus establishing the dominant-tonic relationship (though other
factors, particularly the overlap of membership of the asymmetric sets of
pitches which make up the scales in this system, play an important role
here). The principle of adjacency can also be seen to be of paramount
importance in all those articulatory and gestural characteristics of musical
practice which fall outside the lattice-based description which we have
discussed at length earlier. Melodic practice, especially outside the modern
Western tradition, is heavily dependent on the co-existence of the two
principles of harmonicity and adjacency. What we have discovered is that in
the continuum of intervals between the unison and the octave, the ear is
attracted towards certain nodes of perception, defined somewhat loosely by
the Pythagorean ratios and otherwise perceives adjacent intervals as being
closely related. But we may ask, when is a frequency ratio perceived as a
node and when is it not? How simple does the frequency ratio have to be for
us to perceive that ratio as a node within the system? Or more simply, how
many nodes are there within the octave? Here, culture takes over from
nature. It is clearly a matter of cultural practice how many intervals within
the octave are perceived as being harmonically distinct. In many cultures
only five nodes are recognised, the pentatonic scale resulting. In a
pentatonic culture intervals not falling on the pentatonic nodes are still used
as ornament or portamento decoration but are related by the listener to the
pentatonic nodes through the principle of adjacency and not heard as
separate harmonic entities in themselves. Seven, eight or nine nodes are
common in many musical cultures and these may arise from an underlying
theoretical framework positing an even greater number of nodes (twelve in
the Western chromatic scale or the twenty-two or more srutis of Indian
music).

Figure 4.2 Various tempered scales and their relationship to Pythagorean ratios.

It is interesting to ask just how rational the Western tempered scale is.
Why should it have twelve equal intervals rather than seven or twenty-
three? If we plot various possible equal-tempered scales against the simple
Pythagorean ratios (see Figure 4.2) we will see that a scale of nineteen
equally-spaced elements would have generated a set of intervals more
closely approximating the Pythagorean ideal. Having made these
observations we may now plot a graph of consonance against interval which
corresponds more clearly to our aural experience. In the graph I have
chosen a set of nodes which perhaps corresponds most closely to what the
typical Western listener might experience. An experienced Indian musician
might want to include several more nodal points in such a graph.
Metric structure of the Pitch-Continuum

The most important result of the perceived nodal structure of the pitch
continuum is in giving us a means of measuring distance in the dimension
of pitch. We will say that the pitch dimension has an audible metric. To
explain what this means let us consider two separate sound-systems. In the
first we deal with pairs of stable pitches. If we put the first pitch on a fixed
note and then vary the register of the other pitch, it is always possible to say
of the interval between these two pitches that it is smaller or larger than
another interval. If we now repeat this experiment with two noise-band
sources filtered so as to be of particular colours, but sufficiently wide so as
not to present any aural experience of a definite pitch, we can produce the
same result. Keeping one band fixed, while changing the register of the
second band we can always judge whether the interval between the two
bands is smaller or larger than another interval. If, however, we now change
the register of both sounds in the two sets, our experience is quite different.
First of all, we play two different pitch-sounds; and then we move both
pitches to different registers and listen again. In this case, we can still say
which interval is larger or smaller. Repeating the experiment with the noise-
bands, however—here it is very important that the noise-bands are
sufficiently wide not to present any pitch characteristic—there is no way in
which we can judge which ‘interval’ is larger because we have no frame
against which to measure the distance between the bands.
It should be stressed here that we are talking about our aural
experience. It would of course be possible to make physical measurements
with appropriate instruments and establish the frequency separation
between the central frequencies of the bands of noise and from this
establish a ratio of these frequencies which we could then compare between
different experiences. The problem is that aurally we are not able to do this.
The reason for this we can now see is simply that the dimension of pitch has
a nodal structure. Given two pitches sounding together we do not have to
rely merely on the linear separation of the two sound-objects along the
dimension of pitch (the criterion of adjacency) but we can relate them to
adjacent nodes and thus, via the principle of harmonicity, establish their
intervallic distance. In the case of the noise-bands, the dimension of ‘noise
colouration’ has no perceivable nodal structure and therefore we can only
have a sense of linear distance (principle of adjacency) between the objects.
This does not suffice for comparing intervals originating from different base
lines. We will express this difference between the two systems by saying, at
least in our aural experience, the dimension of ‘noise colouration’ has no
metric.
It is the existence of this underlying nodal structure and the resultant
ability to define an audible metric on the dimension of pitch which permits
us to establish subtly different nodal scales (as, for example, in Indian
music). It might at first seem that a mode might be exhaustively described
in terms of the intervallic distance between successive notes as one ascends
the scale. The question is, however, how does one know that a particular
interval is larger or smaller than another, especially on a very small-scale?
What accounts for our ability to perceive the subtle intervallic differences
between different modes in Indian music? The answer is that we do not
relate merely to the frequency distance between the notes but to the
underlying nodal structure of the pitch dimension. We are able to tell where
the individual notes of the mode are in relation to the nodes in the pitch
dimension. We can tell with a fair degree of subtlety whether a particular
note is very close or not quite so close to a particular nodal point. It is this
which gives us a sense of measure and enables us to distinguish subtle
differences between modal structures.
We might now consider the question, is it possible to establish nodal
structures in ‘noise-colouration’ space? Can we have modes made up out of
(unpitched) bands of coloured noise? We can, of course, artificially define
such a mode; but if we define two such modes, each with very slight
differences between the placement of certain bands, can we distinguish the
two as different musical entities? The answer to the question is probably in
most cases no, and that even where we can we will experience no
qualitative difference in the nature of the music based upon the two
different modes. This is because, as there is no underlying nodal structure to
the dimension of ‘noise colouration’ then there is no qualitative point of
reference enabling us to experience the two structures in a musically
different way. It should be said that in the world of serialist
permutationalism where the nodal structure of the pitch dimension is often
ignored and pitch levels treated as abstract permutatable entities (like sizes
of shoes, having no intrinsic relationships among themselves, only the
extrinsic relationship of ordering in a set), then this distinction may be
difficult to comprehend. We are assuming, however, that formalists will
have abandoned this book after reading Chapter 1.
Two other features are of great importance in the conception of music
on the lattice. The first is that the set of nodes is finite and closed.1 In the
sense that once we reach the octave the set of nodal values is in a clearly
definable sense reproduced over the ensuing octaves. Music based
exclusively on the lattice is thus a finite closed system with a metric. This is
a more precise exposition of Boulez's conception of a music of hierarchic
relationships upon a finite set of possibilities.
This conception can be extended to the harmonic system of Western
tonal music. First of all let us note that one feature we have not discussed in
the definition of a mode is the ability to recognise the root (or dominant
tone). Where we have an entirely symmetrical intervallic structure (such as
in a chromatic scale or the whole-tone scale) a root can only be defined by
emphasis. Normally, however, the scales used in typical musical systems
are asymmetrical in intervallic structure. This allows us to define where we
are in the scale in relation to any particular note. If a root has been
established we can therefore relate where we are in the scale to that root,
even if the root itself has not been sounded for a very long time. The
asymmetry allows us to tell where we are in relation to an absolute point of
reference.
If our asymmetric scales are selected from an underlying symmetric
set (for example the chromatic scale) then we can define scales having
identical asymmetric structures (for example the major scale, T T S T T T
S, where T is the interval of a tone and S of a semi-tone) but with different
roots. If we now compare the members of these various scales, we will find
that the scales on certain roots will have more notes in common with the
scale on a particular root than others. This again allows us to define a
concept of harmonic distance between keys. Note that the major scale is
chosen so as to establish the closest relationship between scales whose roots
are a fifth apart. Experimentation with modal structures will reveal that it is
possible to construct scale systems where the closest relationship between
roots, as defined by numbers of notes in the scale in common, is the interval
of, for example, a minor third.
We can measure harmonic distance in relation to the cycle of fifths
(in fact the cycle of approximate fifths used in the tempered scale system)
and we can aurally perceive the measure of distance between two keys. It is
interesting to note that in the relationship between major keys we might
presuppose that the simple Pythagorean relationship of a fifth between the
roots of the scales of keys was the predominant factor but when we look at
relationships between major keys and their relative minor keys, we see that
in fact common set membership is the predominant perceptual force. There
are twelve notes in the cycle of fifths, which, being a cycle, is of course
closed, hence we can see that the Western harmonic system is also a finite
closed system with a metric (see Figure 4.3).

Structure of timbre-space; multi-dimensionality and non-metricity

Can we expand any of the insights we have gained from our analysis of the
structure of the pitch dimension to an understanding of the world of timbre?
Some crude attempts have of course been made to expand the ideology of
lattice-based music to the organisation of timbre but this is, I feel, merely an
a priori imposition upon the object of musical study. We can, in fact, draw
upon the insights we have already gained but the conclusions we will reach
will be radically different from the formalists. The area of timbre will be
seen to have a radically different structure from the dimension of pitch. This
does not mean that we should abandon it or regard it as essentially
secondary in musical practice, but merely that we should investigate what
criteria of sonic organisation would be appropriate to this particular area.
The first obvious remark we should make about timbre is that it does not
have one single dimension, as does the pitch continuum. This finding often
surprises musicians brought up exclusively in the tradition of Western
instrumental music where timbre has been streamed in specially
acoustically-refined instruments and adapted to the logic of pitch/duration
lattice architecture. It is obvious, however, from our discussion in the
previous section that timbre is a multi-dimensional phenomenon.
David Wessel conducted some preliminary psycho-acoustic
experiments to establish whether any structure exists in this timbre space. In
one experiment timbre has been plotted in a two-dimensional space in
which one dimension relates to the quality of the ‘bite’ in the attack, the
other the placement of energy in the spectrum of the sound (its ‘brightness’)
(Wessel 1979: 49). By this means Wessel has demonstrated that there is in
fact a continuum of values existing within this space which can be
perceived by the listener (Example 4.1).2 I also recall a brief discussion at
IRCAM (undocumented) on this topic between David Wessel and Tod
Machover in which two contrary views were expressed: roughly speaking
that, on the one hand, there is the possibility that the timbre domain will be
discovered to have a structure which we can relate in some way to the
structure of the pitch dimension, and on the other that the timbre domain is
quite distinct in its structure from the pitch dimension. My musical
experience leads me to favour the latter conclusion. But not, therefore, to
come to the Boulezian conclusion that timbre is essentially secondary, of
necessity, in any conceivable musical practice.
Figure 4.3 Representation of Western harmonic system as finite and closed in which harmonic
distance can be measured.

As we have already remarked, timbre is multi-dimensional. In fact,

the two dimensions of Wessel's experimental model need to be expanded to
include factors such as grain, noise characteristics, inharmonicity and
various morphological characteristics. Given this, even assuming that we
are able to discover nodes in the space, it is difficult to see how these might
be ordered in any way similarly to the one-dimensional and finite set of
nodes in the pitch dimension. If we could separate out each of these
dimensions and discover a nodal structure in each then there might be some
hope of success in this direction. It would seem, however, that, just as with
‘noise colouration’, there are no perceivable nodes in any (or at least most)
of the independent dimensions of timbre space.
But if we look at timbre space as a whole it is true that we can
recognise or define particular sub-sets which we might define analogically
as ‘plucked’, ‘struck’, ‘sustained’, or even ‘scraped’, ‘broken’ etc.. I will
return to such (morphological) descriptions in later chapters but for the
moment it is merely important to note that these sub-sets are not orderable
in the sense that the nodes of the pitch continuum form an ordered (and also
closed) set. On the other hand, the existence of clearly-distinguishable
timbre archetypes does mean that we can apply the concept of a field—as in
the usage harmonic field—to groups of timbres and work with a timbre
field. I will return to this idea in a later chapter. The important factor about
the Wessel examples is that although we can certainly define a concept of
distance in the two dimensions which he demonstrates, we cannot,
however, define a metric; there are no nodes in the particular dimensions
illustrated. In fact, we will now declare that timbral space as well as being
multi-dimensional is also not finite, not closed and does not have a metric.
Does this mean, therefore, that it has no structure and that we cannot
organise it in any way?
Referring back to our analysis of the structure of the pitch continuum
we remember that we uncovered two complementary principles, those of
harmonicity and adjacency. The structure of timbre space means that the
principle of harmonicity is not applicable and therefore simply that all the
principal assumptions of lattice-based musical practice cannot be applied to
our thinking about timbre (except in an entirely formalist, a priori
preconception). The principle of adjacency, however, remains. As
demonstrated by the Wessel example, timbral objects which are close to
each other in their multi-dimensional space of timbre are perceived as being
related. It is therefore conceivable to establish a feeling of progression
through timbre space moving via the principle of adjacency from one
distinguishable timbre area to another. This type of progression may be
thought of as in some ways similar to tonal progression in Western
harmonic music and I have in fact used it as the basis for compositional
work (as in for example the piece Anticredos). It does, however, differ from
tonal progression in very important respects. In particular, the system is not
closed and has no metric so, if we make a progression, this will involve a
different sense of distance from that experienced in tonal music. As we
progress, we will be able to sense that we have moved a noticeable distance
from the previous timbral area. We will not, however, be able to measure
how far we are from a different timbral area occurring earlier in the piece.
In this way, our sense of causality or necessity in the musical progression is
confined to the short term but breaks down when we attempt to refer it to
longer stretches of time. This need not be a disadvantage and, depending on
one's philosophical viewpoint, may be considered a distinct advantage!3
Modulation (in the sense of clear progression from one field of sound-
objects to another field) can be clearly demonstrated and utilised in timbre-
space. Modulation between different timbral sets could clearly be used as a
basis for the large-scale architecture of a work, though this does not define
how or why we should work on the small scale. I will discuss this problem
in a later chapter.

Does the continuum have any structure?

If we accept that timbral space is a multi-dimensional continuum, does this

mean that it can have no structure, or that there are no existing structural
models which we can apply to it? Once one becomes locked into lattice-
based or permutational thinking, the apparently natural way to deal with
any quality is to chop it up into a number of distinct and distinguishable
steps and then apply various well-established criteria for the organisation of
a countable set of objects. If, however, we are to deal with the continuum as
the continuum we must break this habit of thought.
We must ask whether this multi-dimensional continuum can have any
structure of its own. Our natural habit of thinking from the parameterisation
of lattice-based music is to assume that all parameters extend indefinitely in
all directions (or at least to the limits of audibility) and that the space is
hence somehow entirely uniform. Continuous space will then be seen as
some kind of endless fog extending in every direction. Continua do,
however, exhibit different structures and this is the subject of the
mathematical discipline of topology. Topology studies the properties of
objects (or spaces) which are not changed by continuous deformations.
Roughly speaking, what properties of a rubber object are retained if it is
stretched in any conceivable way but not broken, torn or pierced? If we
look at Figure 4.4 we see that, topologically speaking, the blob is equivalent
to the sphere and to the cube as one may be deformed into the other by
suitable stretching. None of these, however, is equivalent to the objects with
a hole in them (as it will be necessary to pierce the former objects in order
to obtain this hole), whereas all the objects with a single hole (including the
cube with a handle) are topologically equivalent to one another. Similarly,
objects with two holes are topologically equivalent to one another and
topologically distinct from the other objects we have discussed.
Does timbral space have a topology? When working with existing
musical instruments we may construct a map of the timbral possibilities of
the instrument. To do this, rather than merely listing all the possible sound-
types which an instrument such as a violin might produce, we would
attempt to place these on a map (which might be multi-dimensional) on
which similar sound-objects would be placed close to each other and sound-
objects which are quite different from one another would be placed at a
greater distance. A rough map of the sounds available from the string
section of an orchestra is illustrated in Figure 4.5. An examination of this
map shows that, due to the physical limitations of the instrument itself (and
sometimes of the player), the space does, in fact, have a distinct topology.
At least it is relatively easy to get from normal arco sounds to multiphonics
played arco sul ponticello by infinitesimal motion in the timbre space
(adjacency) but relatively difficult to get from normal arco to percussive
effects on the wooden body of the instrument. In fact, to make a
‘modulation’ in the timbre space from arco sounds to percussion on the
wooden body sounds, it is almost essential to go through col legno
production or through pizzicato production. This means that timbral space
viewed as a space in which timbral progressions (modulations) will be
made has a distinct structure which, although neither closed nor having a
metric, imposes specific limitations on our musical options.
 Figure 4.4 Topological comparisons of some three-dimensional objects.

It will be interesting to see if the flexibility of the computer will allow

us to overcome all such topological restrictions in timbre space (in some
ways this would be a pity) or will we discover that timbre space has an
intrinsic and insurmountable topological structure quite different from the
infinite-coloured fog it is usually taken to be. A second and related question
is, can there be any qualitative distinctions between the ways we move
through this multi-dimensional continuum? Can motion itself in the
continuum have any structure? Let us say first of all that the study of
motion in the continuum has been a central topic of study in science since
the days of Galileo and Newton. Motion in the continuum is usually
described in terms of differential equations. The principal concern of
physical science has been with systems which are structurally stable; for
example, if we are studying the motion of a ball down some sort of incline
we must assume that a small variation in the starting position of the ball
will result in a small variation in its finishing position. If this were not the
case, then the slightest error in the measurement of the initial position of the
ball would lead to totally false results about its final position. As there is
always some degree of error involved in physical measurements, the idea of
physical prediction would become untenable if the system were not stable
in this sense. Stable systems may have different kinds of structures. For
example, see Figure 4.6; in the diagram on the left the ball rolls down a
slope onto a concave surface and wherever we start the ball from at the top
of the slope it will always end up at the bottom of the cup. In the system on
the right, however, the ball rolls down the slope onto a convex surface.
There are now two possible final positions for the ball. It will arrive either
to the left or to the right. Both systems are entirely predictable yet different.
Let us now study what happens when we gradually deform the lower
surface from a concave shape into a convex shape. At a certain point (a
point of catastrophe) the system slips over from having one stable state to
having two stable states and at the exactly intermediate state where the
lower surface is completely flat there are an infinite number of possible
finishing positions for the ball.
Figure 4.5 Timbre map for strings.
 Figure 4.6 One- and two-outcome systems.

Let us consider a wind instrument in this light. If we have a column

of air and we excite it with a breath, a stable resonance will be set up, the
fundamental pitch of the pipe. If we now slowly increase the breath
pressure it will reach a point where the pipe suddenly responds differently
(producing the second harmonic). The system of breath and pipe may
therefore be seen as a physical system having a number of distinct stable
states, the fundamental and various harmonics (a little like the ball rolling
onto the convex surface but with many more than two possible outcomes).
If we imagine that we could now alter the physical characteristics of the
instrument (perhaps here we should imagine that we have a software model
of the system on a computer and we begin to alter the defining parameters)
we may well reach a point where a different structure of stable states is
established or even a state in which no stability can be achieved. It is in fact
possible to conceive of systems in which the smallest variation in the initial
conditions will lead to totally different final results. Such systems are
described in the theory of Strange Attractors. Any system governed by such
a regime, although in principle completely mathematically predictable is in
practice quite indeterminate in its result because the slightest error in the
measurement in the initial conditions of the system will lead to entirely the
wrong result in the prediction of the outcome (see illustration). It seems
possible that such unstable structures may be responsible for the
phenomenon of turbulence in fluid flow (see Figures 4.7 and 4.8).
From a musical point of view the evolution of systems through time
is of particular interest. A study of the differentiation of regimes in time, the
first rigorously mathematical yet qualitative discipline to have physical
implications, is known as Catastrophe Theory. It can be applied to the
formation of crown droplets and bubbles when a drop of water falls onto a
water surface (see Figure 4.9) or to the breaking of waves. One of its most
significant scientific applications seems to be in the study of the structure of
the form of organisms. Organic structures may be looked upon as the end
product of continuous growth processes. The form of an organism may be
viewed as the various instants of a process of continuous evolution
successively frozen in time to create a structure existing in space. Studies of
such phenomena were conducted by D'Arcy Thompson (1961) and Figure
4.10 illustrates Thompson's view of the parallel between the formation of a
‘crown’ when a droplet falls into water and the shape of various cup-like
structures in many minute organisms. Thompson's work also illustrates how
various rates of growth influence the final form of various organic
structures from the shells of sea-living creatures to the skulls of primates
and men (see Figure 4.11).

Figure 4.7 Stable and unstable systems.

Figure 4.8 Turbulent water jet.
 Figure 4.9 Instantaneous photograph of a splash of milk.

Figure 4.10 Calycles of Campanularia (after D'Arcy Thompson (1961)).

Figure 4.11 Patterns of growth compared: carapaces of crabs; shapes of fish; skulls of human,
chimpanzee and baboon (after D'Arcy Thompson (1961))

The theoretical study of such forms of evolution in the continuum

(see Thom 1975) suggested that there were only seven fundamentally
different structures of evolution, or catastrophes in three-dimensional space
and time. Two of these, the ‘swallow-tail’ and ‘butterfly’ are illustrated in
Figure 4.12. More recent research shows that it may not be possible to
generalise this conclusion to high numbers of dimensions. The important
point, however, for anyone concerned with structures evolving in the
continuum (for example, musicians working in timbre-space) is that there
are clearly definable and distinct structures. As Catastrophe Theory is
generally applicable to both the behaviour of physical objects through time
(for example, instruments, electronic sources, the voice or musculature) and
to the description of time-based phenomena (such as acoustic phenomena)
we may reasonably assume that a perceptually valid categorisation of time-
varying acoustic phenomena (either the structure of sound-objects with
dynamic morphologies or the description of the formal properties of sound-
structures evolving through the continuum) is a feasible proposition. It also
strongly suggests why there might be a link between the morphology of
sound-objects or streams of sound, even where these are not intentionally
produced by human gesture, and the quality of human response to these
events.
Figure 4.12a The ‘swallow's tail catastrophe’ and its associated ‘lip’ and ‘beak-to-beak’ singularities
(after Thom (1975)).

Figure 4.12b The ‘butterfly catastrophe’ and its associated shock wave (after Thom (1975)).

I shall not, however, succumb to the formalist temptation to

categorise all sound structures in the continuum in advance with reference
to existing catastrophe theory but will, in later chapters, point to some
perceptible and differentiable archetypes of sound morphology. The
important point to be made here is that the continuum is not an
undifferentiated seamless fog, opaque to human intellectual control but
rather a wonderful new area for exploration provided we have the tools to
control the phenomenon (the computer) and the right conceptual categories
to approach the material.
1 We are assuming here that nodal structures reproduce themselves at the octave as is assumed in
tonal music theory. If objections are raised to this conceptualisation then we will say merely that the
set of nodes is countable and therefore has a structure unlike the continuum which is mathematically
described as an uncountable infinity. The distinction between countable and uncountable infinities is
clearly established in the mathematical literature.
2 Wessel's examples consist of the same melody articulated in two ways: in the first successive notes
are constructed from timbres remote from each other in his two-dimensional timbre space (‘quality of
attack’, ‘brilliance’) while in the second the timbres are close forming a continuous path. The first is
perceived as a number of counterpointed lines, the second to a much greater extent as a single gestalt.
3 Contrary to the commonly-held conception of a deterministic, if not determinable, world, I hold the
view that the world cannot be shown to be, and therefore is not, deterministic and that it is
determinable in the short run but not in the long run. The type of musical architecture in timbre space
described above corresponds very happily with my own philosophical viewpoint!
 Chapter 5
SOUND STRUCTURES IN THE CONTINUUM

Lattice-free objects; dynamic morphologies

A first reading of Chapter 3 may have seemed to imply that we only run
into the area of non-metric adjacency-based organisation when we leave
sounds with simple stable harmonic spectra and deal with sounds of
complex mass or significant noise structure. This would be a misreading.
Although we can define a metric on the dimension of pitch, we do not have
to do so. Let us define some sound-objects based on elementary spectra
which are not amenable to the pitch-lattice description. The simplest object
will be a sine tone with portamento. In lattice-based music such portamento
events are perceived to centre on the pitch of the start or the end of the
portamento. However, we can design a glissando in such a way that it is
very smooth and has such an envelope that the beginning and end do not
significantly stand out from the rest. A music made up entirely of such
sound-objects would fail to draw our attention to the nodal structure of the
pitch-dimension because, without imposing some very special means of
organisation upon the music, nothing in the musical structure would lead us
to focus our attention upon a point of reference which would enable us to
define nodes in the pitch dimension and hence relate sound-events to these.
Continuing with the same material we may imagine a dense texture of such
portamentoed sine tones constructed with such an average density that no
particular pitch centre was predominant. Finally, we may imagine sweeping
a filter across this texture in an arch form (see Figure 5.1). This final object
has a clearly-defined structure of pitch motion imposed upon the texture of
elementary pitches-in-motion but nowhere can we define the sense of a
pitch in its traditional lattice-based sense.
Let us now define the concept of dynamic morphology. An object
will be said to have a dynamic morphology if all, or most, of its properties
are in a state of change—I use the word properties rather than parameters
here, because I feel at this stage that it is important to view sound-objects as
totalities, or gestalts, with various properties, rather than as collections of
parameters. The concept of a musical event as a concatenation of
parameters arises directly from lattice-based musical thought and is
singularly inappropriate to the musical structures we are about to discuss. In
general, sound-objects with dynamic morphology can only be
comprehended in their totality and the qualities of the processes of change
will predominate in our perception over the nature of individual properties.

Figure 5.1 Assembling a non-classical pitch-motion structure.

In his book Sound Structures in Music, (1975), Robert Erickson

discusses the concept of spectral glide, essentially the evolution of spectral
characteristics over a sustained pitch while Pierre Schaeffer (1966)
discusses the concept of allure, subtle variations in dynamic and spectral
envelope over a sound of otherwise constant mass. Here these will both be
regarded as special restricted-case examples of dynamic morphology. A
more typical sound will be one in which the spectrum, dynamic and pitch
level all change through the continuum in the course of the sound. This kind
of transition is illustrated on a more expanded time-scale in the two
Examples 5.1 and 5.2. The first (5.1), synthesised using the Chant
programme at IRCAM moves from pure bell-like sonorities into a vocal
sound through a combination of an acceleration of the individual impulses
and simultaneous widening of the formant structures involved. The second
example (5.2), taken from my work Red Bird, uses a combination of vocal
performance and classical studio techniques to transform the syllable ‘sss’
of “Lis(ten)” into birdsong. We may imagine similar total transformational
processes taking place within sound-objects of much shorter duration and
distinctive sound-objects of this type can be found both in the natural world
and in the realm of speech phonemes and animal sounds.
We may illustrate the relation of these sound-objects to the classical
lattice by attempting to draw them in the three-dimensional pitch-duration-
timbre space of conventional musical thought.1 In Figure 5.2 we see on the
left a representation of a sound-object with spectral glide and on the right a
sound-object of dynamic morphology. The difference between the two is
immediately apparent from the figure. The spectral glide object remaining
within a simple two-dimensional plane, whereas the object of dynamic
morphology winds around freely in all dimensions.

Figure 5.2 Schematic representations of spectral glide and dynamic morphology.

Unstable morphologies: recognition by morphology

From here onwards we will assume that sounds with spectral glide are a
special sub-category of sounds with dynamic morphology. There is,
however, a further class of sounds to be considered: sounds of unstable
morphology. These may be conceived of as sounds which flip rapidly back
and forth between a number of distinct states. In my own writing I often
refer to these as multiplexes. Such sounds are coherent in the sense that the
overall field of possibilities remains constant but the immediate state of the
object is constantly changing in a discontinuous fashion. Example 5.3
illustrates a typical vocally-produced multiplex. To complicate matters
further, multiplexes themselves may have a dynamic morphology! In this
case, the nature of the individual components of the multiplex undergo a
process of gradual change through the timbre space so that the general field
characteristic of the multiplex changes with time. This is illustrated in
Example 5.4.
At this stage, anyone thoroughly enmeshed in the lattice-based mode
of musical thinking may feel that such objects are essentially formless and
incapable of any coherent musical organisation. In fact, however, the
morphology of such objects is a significant recognition indicator in our
everyday experience. To take two simple examples: first of all the sound of
ducks which is normally imitated in the English language by the word
“quack”; the most striking feature of the duck call and the only real feature
which is paralleled in the word “quack” is the spectral glide characteristic
as the formant structure moves from a stressing of the lower formants to a
stressing of the higher formants (caused in the human, and presumably also
in the duck, by the gradual but rapid opening of the vocal cavity).
More significantly, morphology appears to be an essential
characteristic of recognition for certain consonant sounds in speech
discourse. In the Chant programme, developed by Xavier Rodet and
colleagues at IRCAM (See Rodet, Potard and Barriere 1984), vowels have
been successfully modelled by defining their spectral (formant)
characteristics. These models can be used without great difficulty to model
strings of vowels which imitate vocal production very precisely. The
attempt to model consonants has however met with difficulty as consonant
structures have turned out to be extremely context-dependent. Although
spectral characteristics (including noise-based aspects) are important in our
recognition of consonants, they are not sufficient. What does appear to be
preserved, however, from case to case is the shape of the motion, or the
morphology of the consonantal sound-object.

Notation procedures for the continuum

In my work with extended vocal technique and the extended use of

instruments, I have developed a number of notation conventions which are
particularly useful for dealing with continuum phenomena and unstable
acoustic objects. These notation conventions may be applied both to long-
term transitions in the timbre (or in fact in any) field and also to the
detailed, inner articulation of brief sound-objects. The illustrations in Figure
5.3 are taken from Tuba Mirum (1979).
The basic convention for this notation is that objects and processes
are separated. Objects may be thought of as referring to specific delimited
sound-objects or to specifically definable qualities of a sound-object at any
particular time. Thus for example on the tuba we might define a normally
produced tuba note to be one type of sound-object and a tongue-slap
resonance through the instrument to be another. Similarly, various types of
vocal production projected through the tuba may be described as objects.
These objects may then become the material of multiplexes or enter into
transformation procedures (see below). Alternatively, the object notation
may be used merely to define particular states of a system which is in fact
changing continuously. The simplest example of this would be the use of
the standard phonetic symbols for the various noise-based consonants (s, f,
h, and ∫ etc.) to indicate various aspects of a field which is in fact changing
in a noise-spectrum continuum which might pass through any of these
points (see below and also Figure 5.4).
We may now apply to these objects the process notation illustrated in
Figure 5.3. A symbol with three arrows emanating from a point indicates
that the performer is to move in a rapid disjunct fashion between the various
elements indicated in the boxes (the sound-objects). This notation is
particularly useful for describing sounds of unstable morphology
(multiplexes). The symbol with three small circles on a larger circle works
in a similar fashion, except that we are to move in a continuous fashion (i.e.
through the continuum, without discontinuity) between the elements in the
boxes. Looking again at Figure 5.4 we see the use of this sign in
combination with various noise-phonetics. The sound intended here is thus
a seamlessly-evolving noise-band approaching and passing through the
various indicated states, and not a disjunct motion between these states.
Further symbols help us to refine the description of these complex or
unstable sound-objects, for example, we may give the various constituents
different weightings (see various types of boxes in Figure 5.3).
Figure 5.3 Process notation (from Introduction to Tuba Mirum).
 Figure 5.4 From the score of Tuba Mirum.

The transformation signs indicate changes in the quality of the

complex sound-objects through time. In the example from Tuba Mirum (see
Figure 5.4) these symbols are used in an elementary fashion to indicate
slow changes of emphasis over extended sound-objects. In Figure 5.5 they
are used to describe global evolution in the structure of a multiplex stream
and a series of transformations (with qualifications, the equivalent in timbre
space of modulation in the space of tonal harmony) within a sound-object.2
In the final example, see Figure 5.6, a complex series of transitions in
texture, spectrum, pitch-stability, articulation etc. is defined by a set of
nested transformations.
Incidentally, in all these examples the expanded vocabulary of sound-
objects presented by extended vocal techniques is presented through a set of
special symbols extending certain standard phonetic notations to
incorporate features more appropriate to musical development than to
simple linguistic description of phonemic characteristics. Similarly, a fairly
detailed graphic symbolism for notating the timbre/pitch continuum is used
in the example from the vocal pieces, the expanded phonetic and
transformational vocabulary being employed in parallel as a detailed
commentary upon the evolution of the graphic symbols. Once a work of this
type has been learnt it is possible to perform it simply from the graphic
stream. This ‘2-stave’ procedure appears to be a satisfactory reconciliation
of the needs of clear timbral and morphological description of sound-events
with performance legibility. The simple convention of the (horizontal)
transformation arrows has made possible the development of the concept of
‘modulation’ or transformation from one timbral area to another as a
structural principle in the organisation of such works as Anticredos and
several of the Vox series.
Figure 5.5 From the score of Anticredos.
 Figure 5.6 From the score of Vox-I.

The structure of dynamic morphologies

We now have both a notation procedure and appropriate instruments

(particularly the computer) for describing both short-term sound-objects
and long-term musical structures exhibiting a dynamic morphology. How
are we therefore to approach these new areas? Our experience of sound
phenomena indicates that there are definite recognisable gestalts among
sound-objects and processes having a dynamic morphology. Our previous
discussion of catastrophe theory would lead us to believe that our mental
categorisation of such gestalts may not be merely arbitrary.
From a practical point of view, we may divide a discussion of these
matters into two parts. The first will define the structure of certain
morphologies in terms of gesture, the articulation of the continuum by the
agent which instigates the event, the second will classify morphologies in
relation to perceived natural phenomena. Catastrophe theory would lead us
to believe that, in fact, the distinction between these two perspectives is
purely one of convenience. Both the physical structure of living organisms
and the development of their actions through time are just as much ‘natural
events’ and describable in a catastrophe theory framework as any other
naturally-occurring phenomenon. We may expect, however, that the
category of gesture is in some ways more restricted than that of natural
phenomena structures—we would not normally expect to discover
phenomena such as breakage and turbulence in a viable organism, though
this may be open to dispute. It is in other respects more extensive than the
category of natural phenomena—higher organisms are capable of very
subtle articulations of the continuum, which we should only expect to find
by chance in the structure of natural phenomena. The interface of these two
types of description may be seen in the relationship between vocal and
instrumental music. A musical instrument is a device used to stabilise,
through its resonance-structures, the pitch and timbral dimensions of a
sound-event. The morphological structure of the sound-event is thus
dominated by the characteristics of the natural phenomenon of resonance.
The typical instrumental musical object is of the attack-resonance type. The
gestural input from the performer, especially in the highly technologically
developed Western instrumentarium, is thus subordinated to this particular
natural phenomenon structure.
The sound production mechanisms of certain creatures are also
similarly limited. For example, the stridulation of cicadas is heavily
determined by the physical structure of body, wings and legs which produce
the individual impulses and either the natural vibration frequency or a fairly
constant nervous-system clocking-pulse. Gestural input is here confined to
slight speed variations and alteration of grouping characteristics of the
individual pulses (for more detailed discussion see Chapter 11). The human
larynx and vocal tract, as well as the birds’ syrinx, however, are altogether
more malleable systems, allowing control through the pitch continuum (as
with unfretted strings), the continuum adjustment of formant spectrum3 and
transformations in the direction of noise, grain, grit or of harmonic and
multiplexed textures. Subtle muscular control of the physical apparatuses
which govern the emission of sounds ensures that intentional gestural input
may be exceedingly subtle and multi-dimensional. Hence the voice lies at
the farthest extreme from the typical musical instrument developed in the
West.
This, however, need not be the case. The dominant musical ideology
may demand that the voice be essentially an imitation of a typically Western
instrument. Alternatively, instruments may be developed which are not
totally determined by a single, natural phenomenon structure (that of
resonance). The ability of the computer to model physical sound-producing
systems (for example IRCAM's Chant program) means that we may define
instruments which may be controlled in a refined, multidimensional
manner, like the voice itself, or which are governed by different
morphological characteristics (morphologies of motion, turbulence,
unstable oscillations amongst a set of timbral objects, etc.).
It should be noted that there are two levels at which gestural structure
enters into performance practice. The most obvious level is the conscious
articulation of sound-materials involved in interpretation or free
improvisation; to say that this is conscious is not to say that it happens at a
rate which consciousness could precisely describe in real-time.
Consciousness perhaps allows us to rapidly articulate a set of triggers to
activities which are much too rapid for the train of consciousness to grasp in
the actual real-time of performance. This passes over into a more
unconscious use of articulation of the fine details of sounds. In an
established and received performance practice, certain articulations become,
either unconsciously or consciously through practice, ‘second nature’ to
performers, just as to adults walking may seem an entirely natural and
unconscious activity, it is in fact painfully learned by children. Once,
however, the appropriate gestural structures have been learned they become
internalised and do not normally require conscious intentional input, except
perhaps sometimes in the form of a trigger. The same may be said of
various aspects of musical performance practice. At the same time, these
‘second nature’ aspects of our behaviour may in fact be brought to the level
of consciousness through self-examination and hence brought under
conscious control and altered, developed or further articulated. The new
gestural procedures may themselves be once more internalised as ‘second
nature’. This process of externalisation and reinternalisation of ‘second
nature’ is an essential feature of learning to be a free improvisation
musician. It is necessary to become aware of ingrown habits and musical
clichés. Otherwise so-called ‘spontaneity’ reveals only mental habits and
the clichés of one's musical milieu.
Composers have adopted several different approaches to the
description and control of dynamic morphology. Stockhausen's
Mikrophonie I offers an interesting example in this respect. The piece uses
as its sound-source a large tam-tam. This instrument has a complex resonant
structure and is capable of excitation and articulation in a number of ways.
This may result in sound-objects with primarily noise-based spectra, or
objects of fixed or varying pitch, or fixed or variable multiphonics of
various kinds. The morphology of the sounds thus produced is governed
partly by the physical structure of the instrument (the sounds’ long duration
and stable complex resonance) and partly by the means of excitation, many
of which are open to strong gestural input from the performer. In the score
of Mikrophonie I Stockhausen attempts to describe the various sound-
objects required. His prescriptions may be described as analogical and
suggest the morphology of various natural processes (as might be expected
from our discussion) such as ‘cracking’, ‘grating’, etc. or certain definable
gestural types uttered by living creatures such as ‘groaning’, ‘baying’,
‘whimpering’, ‘shrieking’ etc.. At least sixty-eight such analogical
descriptions appear in the score. This approach may at first sight seem
wildly approximate, compared with the apparent clear specificity of
traditional music notation. It is clear, however, that we are capable of
categorising sound morphology into definable archetypes such as those
suggested by Stockhausen's descriptive terminology developed for
Mikrophonie I.4
Morton Subotnick approached this same problem from a different
perspective. Musical improvisation with various found objects led him to
appreciate the importance of intellectual-physiological gesture in musical
articulation. When working in the sphere of voltage-control synthesis,
however, this kind of articulate gestural input was not so directly
achievable. Subotnick worked on various ways of adapting the voltage-
control technology to the articulation of gestures in the music. The final
refinement of this approach can be seen in the so-called ghost box. Gestural
information is converted into various control voltages which can control
such parameters of the sound as speed and depth of amplitude modulation
and frequency modulation or characteristics of motion in space. Such
gestural information may be precomposed and stored on a tape5 later to be
input to a live performance, which it articulates using the ghost box
technology. The control voltage information is carried very simply on a
single mono audio channel. High frequency pulses activate gates which
initiate sound-events. Mid-frequency and low frequency signals are
separately modulated in amplitude according to some gestural criteria. The
three frequency bands are separated out by filters in the ghost box and the
amplitude modulations of the lower signals may be translated into
fluctuating voltages used to control any aspect of the performance material.
The amplitude modulations of the stored signals on the tape may be input
either by painstaking definition and control of voltages at a synthesiser or
by directly monitoring manual or other gestures (e.g. control voltages
generated by the manual manipulation of a joy-stick control) (see Figure
5.7).
The transfer of gestural information from one medium (e.g. manual
control) to another (e.g. control of motion and position in space) is one
important aspect of the development of electronic instruments. It permits
articulate gestural information to be transferred, without the need for
distanced intellectual conceptualisation, from a medium in which a
musician is most fluent (operation on a keyboard or bowing technique on a
stringed instrument), to areas over which he normally has less subtle or, in
fact, no control. A more recent and powerful tool in this direction is the
vocoder which permits the transfer of the characteristics of voice
articulation to the control of the parameters of other kinds of sound-
material.
On the other hand Brian Ferneyhough, working through traditional
conventions of notation (and their extensions) has evolved a complex
vocabulary of musical gesture on conventional musical instruments through
a very careful analysis of the various parameters of performance practice.
Aurally speaking,—I am now speaking of my aural experience of the music
and not of the composer's theories—Ferneyhough's music stands at an
interesting meeting-point of gesturally conceived musical practice and more
traditional serialist thought. To my ear, a work such as his Second String
Quartet presents an interesting approach to the counterpoint of musical
gestures, which I will be discussing in the next chapter, whereas some of his
works for larger ensemble appear to bury the articulation of gesture in the
density and complexity of their serial architecture; the music ceases to
breathe gesturally.

Figure 5.7 Schematic representation of Morton Subotnick's ‘Ghost-Box’ system.

Postscript: the integration of pitched and unpitched architectures

The principal problem facing music which focuses upon sound-objects of
dynamic morphology—and here I am talking particularly of objects whose
pitch is not stable—is how one achieves some kind of coherent connection
with more traditional pitch architecture. Much music, of course, starts out
from the basis of traditional pitch architecture and articulations of the
continuum are conceived within this framework. Music adopting a more
radical approach to the use of sound-material, however, has to find a
solution to this problem. When the musical objects used in an architecture
do not in themselves imply any particular lattice structure it becomes
difficult to justify any steady-state pitch material which enters into the
musical discourse. Often a pitch or pitch lattice may be introduced as a
drone, either on a single pitch or over a harmonic field emerging out of a
noise-based structure in which pitch is not clearly differentiable. This
process takes a certain time to unfold convincingly, hence the appearance of
such drones.
There are clearly three principal approaches to this problem. The first
lies in relating harmonic field structures to the internal structures of
inharmonic spectra. This approach of course only applies where there are
stable spectra present in the music already. The second approach depends
on defining a harmonic field (or a series of such fields) such that any
occurrences of stable or temporarily stable pitch objects fall on the lattice
defined by the field. In this way a lattice-based architecture may coexist
with an architecture of dynamic morphologies. This approach is in fact
similar to the idea of music hung on a lattice (see Chapter 2). It may be
thought of as a lattice hung on a morphological music, a change in
emphasis. Finally, the imposition of stable resonant structures on complex
(and noise-based) objects through, for example, filtering techniques may be
a way of achieving a mediation between a morphologically-based and a
lattice-based architecture in the musical structure.
It is interesting to note in this respect that musicians schooled entirely
in lattice-based musical thinking will tend to latch on to aspects of stable
pitch or resonance in any object (unfortunately, in many cases, they will see
this as carrying the primary musical substance of the object). The way one
relates to a complex object having some degree of resonance or time-
domain pitch-structure depends entirely on the contextualising mode of
organisation of the materials. At least in our culture it seems extremely
difficult to achieve a balance between morphological and lattice-based
architecture. For many musicians, lattice structure is what differentiates
music from non-music and morphological architecture will be perceived as
either chaos and no architecture at all or at least of no concern to the
musician. Hopefully as we approach the end of the twentieth century we are
growing out of this culturally ingrained habit.

1 It should be said that in order to represent timbre accurately we will need a number of dimensions
for timbre alone so that our representation should be in at least four, if not six, dimensions. We
maintain the fiction that timbre has only a single dimension here merely in order to be able to draw a
diagram.
2 Note that the three voices involved in the latter process form a single stream to the listener, as they
are projected via a loudspeaker system as emanating from a single, moving point in space—see the
comments on aural imaging in Chapter 3—the rests in the three parts are therefore not intended to
break up the perceived stream.
3 The only parallel in the Western instrumentarium is perhaps the crude one of muting with brass
instruments or strings, an all or nothing procedure in many cases.
4 The score gives 68 verbal descriptions of the sound qualities aimed at by the performers (playing a
large tam-tam with an enormous variety of implements of many materials, picked up actively with
microphones and processed with filters and potentiometers). These range from groaning, hissing,
yelling, rustling to quacking, fluting, whimpering, and murmuring (Ed.).
5 Later digital versions of the system utilised a portable EPROM instead of the tape, hence without
the need for a filter stage (see below) to extract the control voltage (Ed.).
 Chapter 6
GESTURE AND COUNTERPOINT

Gesture and pitch-structure

Up until now it seemed reasonable to assume that the gestural

characteristics of a musical performance were conveyed by dimensions of
sound-space other than pitch. This, however, need not be the case. In
lattice-based musical practice, pitch is constrained to manifest itself on the
steps of the lattice defined by the system. In natural processes, however, this
need not be the case. Thus, in a simple instrument such as a siren, the sense
of pitch is generated by the rotation of a physical system, itself set in
motion by the flow of breath. Due to the natural inertia of the system, when
we blow hard into a siren the pitch rises initially from zero cycles to its
maximum level in a rapid portamento and when blowing ceases it falls
slowly back to zero. Pitch here is a measure of the degree of motion or
activity of the system which is itself related to the gesture of blowing. The
gesture of blowing (see Figure 6.1) may have an almost square envelope,
but the physical inertia of the siren system will smooth this out into the
shape shown which also expresses the pitch-contour experienced. Hence we
obtain a somewhat fluid mapping of the energy input of the gesture of
blowing.
 Pitch-contour can also be an indication of energy input in such
natural phenomena as the whistling of wind through telegraph wires, higher
pitch being associated with greater wind velocity and therefore more
energy. Similarly, the vocalisations of terror of most creatures are emitted
with great energy and are of high pitch (more will be said on this subject in
Chapter 11). Inversely energy contours may be suggested by pitch contours.
In Example 6.1 from my work Anna’s Magic Garden the vibration of a
piece of elastic is used as a sound-source. When the piece of elastic is
physically vibrated with more energy, the perceived pitch rises. In addition,
due to the inertia of the system, when a switch-on, switch-off gesture is
inputted to the system (like the blowing gesture through the siren) it is
smoothed to give the characteristic pitch-contour perceived in the case of
the siren, though in this case the decay time is much shorter. If this sound is
recorded and speeded up one senses that there is more energy in the system.
This is partly to do with the physical characteristics of the sounding object.
Note that in the case of the voice a speeding up above a certain point moves
the characteristic formant bands and the voice begins to appear unvoice-like
(typically, a real voice may change its time-domain pitch but its spectral
domain pitch or formants remain fixed for a given vowel). The vibrating
piece of elastic clearly has no fixed formant spectrum and thus we accept its
speeding up on playback as if it were a physically realisable phenomenon.
Due to other characteristics of the sound (particularly its internal motion
and its spatial movement) an aural metaphor of flying is suggested or, at
least to the less poetically inclined, an intensification of energy and perhaps
a sense that this energy is ‘thrown’ to a high point and falls back as the
pitch falls towards the end of the signal.
 Figure 6.1 Relation between breath pressure and pitch in a rotary siren.

In the next example (Example 6.2) a related effect is achieved with

material from a quite different source. In this segment from Parmegiani, it
might be that the sound-objects are made by fast-winding a continuous
texture past the heads of a tape-recorder (though clearly many other
processes are involved ). Again, however, the formant structure permits us
to link pitch with energy input and the sense of short gestures rising to a
plateau and rapidly falling away helps to produce a sense of something
being thrown and returning. This segment is in fact from a piece entitled
Etude élastique. In both these cases I am not seeking to poetically elaborate
the sound experience, but merely to attempt to offer some coherent
explanation of our response to the sound in terms of the physical
characteristics of the real world and our own experience of gesture.
We may now extend such considerations to melody itself. Although
musical practice constrains melodic gestalts to notation on lattice pitches,
performance practice (particularly the articulation of sustains with, possibly
variable, vibrato and the closing of leaps by portamenti in some cases)
would suggest that melodic contour has something to do with the
expression of gestural energy through pitch-motion. Obviously in a musical
practice dominated by lattice-based notation, considerations deriving from
harmonic structures and motivic development will play an important part—
in some cases an exclusive part—in the structuring of melody. However, I
would suggest that in many cases we can perceive some gestural core to the
structure of melodies. One observation in favour of this proposition is the
tendency of melodies to cohere into a single object of perception, a
structure essentially destroyed by dissection and permutation. This point of
view will, of course, seem strange in the twentieth century climate of
motivic or serial thinking. But in music where melodic thinking (as opposed
to motivic thinking) is still of primary importance, this statement will be
clearly understood. In modern Western practice, of course, the dominance
of the notation system and the associated system of theory has permitted a
gradual breakdown of the concept of melody, first into that of motif and
latterly its complete atomisation into the concept of the (contour-
independent) series. Productive though these may be, I would suggest that
the perception of a true melody as a coherent whole has something to do
with its relationship to a coherently articulated gesture—the codification of
motivic practice, starting first with neumic notation, is part of a certain
puritan thrust apparent in Western Christian civilisation!
Taking a simple example, the melody illustrated in Figure 6.2 is
based on a motivic device with clear gestural import. The excitement
generated by the melody is to a large extent generated by the sense of
sudden accelerated energy as a pitch is thrown upwards and then falls back
slightly in this gesture. I am not suggesting for one moment that melody is
reducible purely to a gestural description but mean merely to indicate that
gestural thinking is not confined solely to aspects of sound experience
which are not normally notated. The important thing about gesture or
dynamic morphology in general, is that it is essentially a time-varying
property of a whole sonic object and cannot be atomised in the same way
that pitch-lattice components can be separated through their discrete
notation. Conversely, this property of gesture is one reason why it can be
applied to the analysis or control of sound-objects which are varying in a
continuous manner in many dimensions of the continuum. It does not need
to be atomised or broken into dimensions, though of course, gestures
articulated in independent ways in several different dimensions can carry
more information than a gesture whose evolution takes place in the same
way in all dimensions. In the same way, gesture is not reversible. To take an
elementary example, becoming more intense is in no sense a close relation
of becoming less intense! (Except of course if we insist on perceiving
gesture out of time, a habit ingrained in Western art music practice through
centuries of using spatialised time in notation.)

Figure 6.2 Gestural contour in a melodic shape.

Gesture in language, popular music, dance

As we have remarked earlier, language utterances do not consist exclusively

of a combination of phonemes. Many other components (tone of voice, state
of breathing, intonation, use of stress etc.) enter into the conveying of
meaning through language. Some of these factors are in fact culturally
based, particularly, for example, the stress patterning in particular
languages, and are correctly termed paralanguage, needing some structural
explanation in terms of linguistic categories. Other aspects of language
(intonation, voice quality in relation to sex recognition, loudness etc.)
appear to inhabit an expressive sphere independent of language itself. These
matters will be discussed more fully in the sections on animal
communication and human communication in Chapter 11. Here we may
note that a piece like Berio's Visage depends on the existence of
translinguistic expressive components in vocal utterances for its effect
(Example 6.3). Berio describes Visage as —
Based on the sound symbolism of vocal gestures and inflections with their accompanying
‘shadow of meanings’ and their associative tendencies, [...] [including] [...] vocal events
from inarticulated or articulated ‘speech’, [...] from patterns of inflections modelled on
specific languages [...].
(Berio 1967)

The languages which Berio mentions include Hebrew, Neapolitan,

Armenian, English and French (Emmerson 1976: 28). In many cases in this
piece the phoneme-like components merely act as carriers of vocal gestural
articulation which is carried through articulation of pitch, dynamics, timbre
inflection within and over different sound-objects, rhythmic grouping and
especially the integrated articulation of all these dimensions of gesture. It
might also be suggested that the observed stress-patterns and patterns of
inflection of specific languages have some more primordial source and that
some remnants of this quality remain despite the evolution of culturally
specific stress patterns in particular languages. It may also be the case that
when conversing within a language we interpret articulations of stress and
inflection as conventional but in listening to a language we do not
understand we tend to interpret these in a more direct gestural manner.
Language use therefore suppresses the gestural implications of the stress
and inflection patterns used. A very typical gestural structure, ‘pushing to a
stable articulated position’, the archetypal expressive gesture of the ballad
singer in popular Western music, is the sustained note which is attacked
very slightly flat and then portamentoed towards the standard pitch while
gradually bringing in vibrato. This gesture is, loosely speaking, a parallel, in
the domain of the continuum, of harmonic resolution.
 Figure 6.3 Laban notation for dance movements.

Gestural analysis may in fact be extended to realms of human activity

which do not involve the production of sound. Many aspects of dance are
self-evidently gestural. A fairly complex analytic notation for dance
movement was developed by Rudolf Laban. Laban himself attempted to
notate the stress patterns of movements and more recent developments of
the notation have attempted to describe the internal flow of ‘strength’ in a
movement (see Figure 6.3 in which black commas indicate particularly
strong movements, white commas particularly light movements). These and
some other aspects of Laban notation may be seen as relating to the
articulation of gesture in sonic utterance. It might be argued in fact that the
very attempt to establish a relationship between dance movements and
music (apart from timing of movements and rhythmic pulse) relies on the
unwritten assumption of a common gestural substrate at some level. This
does not preclude dance (like music) having other concerns, such as the use
of mime-imagery in disco dancing.

Contrapuntal structure
Can we establish a truly contrapuntal method of working in the continuum?
To answer this question we will need to analyse the concept of counterpoint
in lattice-based music and attempt to generalise the conception so that it is
no longer dependent on the existence of a lattice structure. We should also
examine some existing approaches to contrapuntal structure in the
continuum.
The example from Pithoprakta by Xenakis, quoted in Chapter 2,
illustrates a form of rudimentary counterpoint. The three lines of string
sounds, none of which is based on the pitch lattice (glissandi of glissandi),
are in fact streamed in terms of a timbre lattice. We differentiate the three
streams because of their different and consistent timbral qualities. The three
lines certainly coexist in the same musical space and are heard as distinct
entities, but we can hardly describe them as contrapuntal as there is no real
interaction between the parts except in their coming together in the high
register before the sustained chord material is revealed.
In my own piece, Anticredos, for six voices using extended vocal
technique, (listen to Example 6.4), a simple example of stream-divergence
is set up. When we are dealing with continuously evolving streams, we can
imagine that a single stream evolves through the continuum into two
distinctly separated streams. These may be separated in pitch or pitch area,
timbral characteristics or space (in the example given, all three). The
important point is that this division of one stream into two may take place
quite seamlessly, an effect which would be, to say the least, extremely
difficult to achieve in lattice-based instrumentally streamed music. Also, in
this particular example, the two independent streams are themselves
undergoing continuous timbral transformation as they independently circle
the audience (the live work is projected on four loudspeakers surrounding
the audience). In this case, therefore, we have streaming without
instrumental streaming of timbre. However, stream divergence is the only
‘interaction’ between the two existing musical threads.
A more complex and highly-articulated development of the concept
of seamless divergence and merging of sonic streams may be imagined (see
Figure 5.2) and this would certainly be an entirely new realm of musical
development dependent on our acceptance of the multi-dimensional
continuum as a valid substrate for physical composition. The gradual
separation or reintegration of the streams might be emphasised by the
imaginative use of spatial movement (another continuum parameter).
 We may, however, develop a conception of counterpoint closer to
what we understand by that term in lattice-based music theory. Let us first
of all attempt to make a generalisable analysis of the substance of the
contrapuntal experience. For us to describe the musical experience as
contrapuntal in a more conventional sense, it is not sufficient for us to
experience the mere coexistence of a number of musical streams. These
streams must be felt to relate to one another or interact in some way during
the course of their separate development. In lattice-based counterpoint this
will involve the ebb and flow of rhythmic co-ordination and harmonic
consonance (or ‘normality’ defined in some other sense) in the relationship
between the parts. Ideally, in tonal counterpoint, we should expect to feel
that the overall musical texture is ‘going somewhere’.
Thus, in addition to the streaming of individual parts, we can
establish two criteria for our recognition of a contrapuntal structure. First of
all, an architectural principle which supplies points of reference in the
overall progression of the musical material. This corresponds to the key
structure in tonal counterpoint. Secondly, a dynamic principle which
determines the nature of the motion. In tonal note-against-note counterpoint
this is related to the ebb and flow of rhythmic co-ordination and the ebb and
flow of harmonic consonance-dissonance that we have discussed
previously, both of which arise from the way in which notes in individual
parts are placed relative to notes in other parts. The lattice structure of tonal
music allows us to develop a detailed and elaborate sense of contrapuntal
development.

Figure 6.4 Relational notation from Stockhausen's Mikrophonie I.

Few explicit attempts have been made to achieve the same

contrapuntal results when working in the continuum. An interesting
exception is the relational notation used by Stockhausen in Mikrophonie I
(see Figure 6.4 ). Here the development and interaction of the individual
parts is organised according to a simple set of relational and
transformational symbols. This notation, however, seems to be essentially
enabling and tells us little about the details of interaction between the
internal structures of sonic objects which occur in the actual performance.
In fact, the experience of the counterpoint of the performance is quite
separate from the rationale of the score, an approach we would like to leave
behind.
I would therefore like to suggest a more detailed model for
counterpoint in the continuum, deriving from my generalised analysis of
counterpoint above. In the architectural dimension we will replace the
progression from one key to another, which is the architectural base-line of
tonal counterpoint, by the concept of transformation from one timbral or
sound-morphological area to another (this is discussed in Chapter 5). In the
dynamic dimension, we will replace the interplay of consonance-dissonance
in harmonic progressions by the idea of gestural evolution and interaction
between the separate streams.
A fundamental characteristic of our original analysis of tonal
counterpoint is that the two dimensions, the architectural and the dynamic,
are independent of one another. Thus the tonal contrapuntal development is
articulated upon the key structure and is not a mechanical manifestation of
the tonal progression implied by the architecture. The problem with our
earlier example from Anticredos is that there was no separation between the
architectural and the dynamic. The two streams of sound evolved timbrally,
but their content was entirely determined by this timbral evolution. There
was no separate dimension in which a dynamic could be developed. What
we are now suggesting is that the evolving stream may be gesturally
articulated and we may coordinate (or not) the gestures between the various
parts in such a way as to create a viable contrapuntal structure. I will
attempt to illustrate how this approach might be realised with reference to
the piece Vox I.
In Vox I, four vocalists, using extended vocal techniques, perform
against an ‘orchestra’ of natural-event sounds on four-channel tape. The
principal timbral progression of the piece is from pitch-free, unarticulated
multiplexes towards pitched language-like unison utterance. The line of
development is not direct (the piece ‘modulates’ back to unarticulated
multiplexes in its central section) and passes through a number of distinct
timbral-morphological areas. Although the piece uses four voices as its
sound-sources, these are not necessarily projected as four independent
streams. Using mixing and spatial motion facilities the listener may be
presented with a single stream, two equally-weighted, or unequally-
weighted streams, three unequally-weighted streams or four streams of
sound, and in any of these cases, the perceived streams may move around
the quadraphonic space in the auditorium. The timbral-morphological
progression in the piece can be partly analysed in terms of standard
(spectral and dynamic) timbral properties, but another important feature is
the mode of articulation (an aspect of gestural structure) itself. Roughly
speaking, this develops from continuous unarticulated multiplexing through
continuous articulated multiplexing,1 distinct articulated multiplex units to
distinct speech-articulated units.2
As mentioned before, this progression is not a simple motion from A
to D. Furthermore, the development of ‘articulateness’ is further
complicated by the division (and sometimes merging) of the sound-streams.
For example, at the commencement of the piece, the continuous,
unarticulated multiplex we hear is produced by a mixing of all four voices.
This slowly evolves (as in the Anticredos example) into two distinct
continuous streams which then begin to be articulated and at this point the
articulations begin to interact. At a later point, this stream divides again,
this time into four distinct vocal streams. Conversely, towards the end of the
piece, four distinct and separately-articulated vocal streams gradually
converge both in content and articulation to form the unison stream of
coordinated, speech-like articulation which ends the piece. The tape part,
which forms another layer in the composition, is also related to this
development. The initial stream of the four voices emerges seamlessly out
of the tape material whilst, at the end, the tape for the first time settles on
fixed pitches (as do the voices) and through vocoding procedures slowly
begins to parallel the articulation of the unison speech-like material in the
voices. There is thus a clear, architectural sense of progression through the
piece.
Once, however, more than one vocal stream appears, it is necessary to
develop a contrapuntal articulation between the streams. This becomes
particularly significant towards the end of the piece, where the four voices,
clearly separated on the four separate loudspeakers, each perform highly
complex sound-objects gradually evolving the material of the speech-like
utterances. I will give here a simplified description of the procedure
employed to develop a contrapuntal feeling in the music. I should stress, in
line with my comments earlier in the book, that in the end I am governed by
my aural experience of what actually functions as counterpoint. I am not
attempting to set up a formal procedure which can be followed blindly (or
rather deafly, notationally) by anyone else. However, this particular
problem is hardly likely to arise as, to be explained below, the gestural
procedures do not appear explicitly in the notation.
My approach to defining a gestural structure would be simplified, not
because, as some may think, gesture itself is a simplistic parameter, but
quite the opposite. To make a short digression, let us consider the gestural
articulation of a standard, stable musical note. In fact, let us consider just
one single dimension of this, the use of vibrato (that is the iterative
fluctuation of frequency around the mean). First of all, note that vibrato has
two different dimensions, at least: the rate of iteration and the depth of the
frequency variation (we might also have considered iterative spectral
variations). For the purposes of this discussion, we will define a number of
morphological archetypes for the gestural articulation of vibrato, bearing in
mind that these are merely poles in a continuum of possibilities. Referring
to Figure 6.5, we might classify the possible gestures in three ways:

(1) According to magnitude. Hence the depth of the vibrato might be

shallow, medium or deep (note also that the speed of the vibrato may
also be sluggish, normal or rapid). We consider here only three levels
for the sake of simplicity. In the case of vibrato which changes, we
should consider the three ranges narrow-to-normal, normal-to-wide
and narrow-to-wide.
(2) The morphology of the vibrato. In fact this is really only the first-order
morphology, as we will explain below. Let us choose just eight
distinguishable archetypes: stable, increasing, decreasing, increasing-
decreasing, decreasing- increasing, unstable, stable-becoming unstable,
unstable-becoming stable. Unstable should be understood to mean
non-periodic or semi-random fluctuation in frequency widths (or
iteration-rate) of the vibrato. Clearly this list might be extended, but
these seem to me to be eight perceptually quite distinct gestural
morphologies.
 Figure 6.5 Morphological archetypes for gestural articulation.

(3) Second-order morphology. In the case of changing morphologies, such

as an increase or a decrease in the depth of the vibrato or a change in
its stability of range, it is clearly possible to perceive a second-order
structure. Thus a change may be quite smooth or it may accelerate as it
gathers momentum, or begin quickly and slacken in pace. Thus we
might imagine a note which, from the moment it is attacked, begins to
vibrato minutely and this vibrato grows wider, smoothly with time.
Alternatively, the note is attacked and appears to remain stable in
tessitura for a short time after which shallow vibrato becomes apparent
but, as soon as we become aware of the vibrato, it rapidly increases to
a maximum. Alternatively, a note may be attacked and very rapidly
increase its depth of vibrato, reaching a stable plateau of wide vibrato
for its end portion. With the increasing-decreasing morphology, these
second-order effects would alter the temporal position of the maximum
point. With a steady rate of change the maximum point would be
reached at the centre of the note. With an accelerating change, the
maximum point would be reached towards the end of the note,
followed by a more rapid decrease and with the slowing shape the
vibrato would increase to a maximum quite rapidly and then die away
in range slowly. These three second-order morphologies may be
referred to as direct, accelerating and decelerating.
 Each of the second order morphologies may be associated with the
first-order morphologies which exhibit change. This gives us twenty-two
perceptible different morphologies and each of these may be associated
with the three magnitudes, giving a total of sixty-six gestural archetypes. If
we now remember that these articulations may be applied to both the
frequency width and the rate of iteration of the vibrato, we now have 3,756
ways of articulating vibrato! We may now apply the same gestural criteria
to the overall dynamic envelope of the note and the tremolo characteristics.
We can hence describe 14,106,536 possible articulations for a standard
musical note. If we now enter the field of the true continuum and consider
portamento motions of the pitch and timbral transformation of the pitch
through time, we discover 50,000,000,000 perceptible distinguishable
sound-objects. At this point, serial methodology loses its charm.
Let us therefore assume for the moment that there are only four
distinguishable gestural types which we shall apply to the gross
characterisation of a sound-event (we are making no attempt at parametric
separation). We will describe these as stable, unstable, leading-to and
leading-from; each of our four lines of counterpoint will consist of musical
gestures which we will describe in terms of these four archetypes. (Note
here that the description is a convenient fiction, in practice we will
articulate each musical object with greater subtlety. The four-pole
characterisation merely enables us to get a handle on the gestural structure.)
The sequencing of gestures in a single line will be strongly determined by
the particular type of expressive coherence (or lack of it) in that line. This
will determine both the type of gesture used, the sequence of individual
gestures and the average rate of gestural activity.
We may now look at this structure ‘vertically’. From this point of
view, the overall rate of occurrence of gestures from moment to moment
will be an important parameter. This may be defined, for example, by
marking off blocks of time in which equal numbers of gestures occur,
counting all four parts at once—although clearly from a perceptual point of
view if the gestures in two parts move exactly synchronously the overall
rate of gestural activity will appear to be half that we would obtain by
counting up the gestures separately in each part. We now have a matrix in
which horizontal and vertical considerations of the timing of gestures
interact.
 We may go a stage further, however. We may consider the gestures in
the various parts over a short period of time and consider (a) if the gestures
in different parts are similar to one another (homogeneous) or different
from one another (heterogeneous)—note that this is independent of whether
the gestures in an individual part are homogeneous, or heterogeneous—and
(b) whether the gestures appear to interact with one another or appear to
behave independently of one another. The first criterion requires no
qualification, the second, however relies on musical judgement in the act of
composition (in the act of listening the distinction is quite clear).
From this analysis, we may derive six archetypes for the vertical
ordering of gestures (see Figure 6.6). Gestures which are the same in all
parts may be organised completely in parallel,3 semi-parallel (in which the
parts follow the same gestural logic, but not in a synchronous way) and
homogeneous independence (where the parts appear to be behaving
independently of one another). When the vertical organisation of gestures is
heterogeneous these may also be independent (heterogeneous
independence), they may appear to be interactive (for example through the
relative placement of stress accents between parts which may suggest
causal or imitative links between events in different parts) or triggering, in
which a gesture in one part appears to initiate an event or change in another
part quite clearly. For example, one part sustains some material, a second
part enters quietly and becomes louder in an accelerating fashion, cutting
off at the point of maximum loudness. At that same instant the other part
begins to articulate rapidly. The event in the second part appears to trigger
the change in the first one.
 Figure 6.6 Six archetypes for the vertical ordering of gesture.

The independent sets of horizontal and vertical criteria for the

organisation of gesture allow one to develop a subtle architecture for the
contrapuntal development of the music. At the same time it is important to
realise that gestural structure is independent of the timbral characteristics of
the sound-objects themselves. In Vox I, therefore, the evolution from
multiplexes towards speech-articulated, stable-pitched sound-objects is
articulated through the counterpoint of gestures of the individual sound-
objects in the vocal streams (see Figures 6.7 and 6.8). Although, of course,
it would be possible to set up a serial net of possibilities and thereby
permute the various conceivable gestures, this would seem to me to defeat
the object of contrapuntal development. Hence the articulation of gesture
must underline the sense of timbral transformation and help carry it
forward. In actual practice, therefore, various nodal points in the musical
architecture are established where, for example, crucial sound
transformations materialise or a particular gestural structure (for example,
complete parallelism or a strong, trigger device) is decided upon. I would
then decide on the positioning of other minor nodes and construct the
remaining architecture of the music, both horizontally and vertically, in
such a way as to move the music towards and away from these points with a
sense of purposive development. Nodal and sub-nodal points are also
provided by the evolution of the materials on tape, all of which have a
gestural structure which can be related to the gestural counterpoint in the
voices themselves. Spatial placement and movement is also used to
articulate the material and I will discuss this aspect in Chapter 10.
I found this approach to be a powerful heuristic tool for composing
with this kind of material. It was possible to lay out the structure of the
overall density of events on the score, then compose the gestural structure
of a section using elementary symbols like those in Figure 6.7 and then,
working from the overall plan of timbral and articulatory development,
score in the details of the individual sound-events in each voice. (In
practice, of course, the general nature of the sound-events in any section is
already in one's mind when the gestural structure is being composed.) This
means, however, that the underlying gestural structure does not appear
explicitly in the score (for example, it is not a simple correlate of any
specific parameter notated in the final score) and could only be teased out
by a sympathetic music analyst, prepared to listen to the music itself.
Finally, it should be said that I have only been able to attempt a theoretical
description of gestural counterpoint given here, because of my attempt to
write down music using complex sonic objects. When working entirely in
the electro-acoustic music studio, gestural structure can be finely tuned by
the experience of aural feedback and need not involve any separate process
of conceptualisation. Having initially rejected the medium of music
notation some years ago and after working in the electro-acoustic music
studio had brought me face-to-face with its implicit musical ideology and
intrinsic limitations, I have found the attempt to evolve new and appropriate
notational procedures to be very fruitful. I would still insist, however, on
the ultimate validation of any musical procedure through the unmediated
and unprejudiced listening experience. No notational logos can in itself
justify a musical procedure. A rationale based on listening to and working
with acoustic materials—whether or not it is rigorous or elegant (and if it is
either of these we should already be suspicious)—will be needed when we
approach the multi-dimensional continuum opened up by digital sonics.
Figure 6.7 Some working parameters for the composition of Vox-I.
 Figure 6.8 Section of four-part gestural counterpoint from Vox-I.

1 This articulation is achieved partly through using the cupped hands as a variable filter on the
continuous vocal source.
2 It should be emphasised that the speech is speech-like articulation, the language is imaginary.
3 Akin to tutti so long as we bear in mind that we are talking here about gestural structure and not
spectral type.
 Part 2

Landscape
 Chapter 7
SOUND LANDSCAPE

Any sound which has too evident an affinity with the noises of everyday life, [...] any sound of
this kind, with its anecdotal connotations, becomes completely isolated from its context; it
could never be integrated, [...] Any allusive element breaks up the dialectic of form and
morphology and its unyielding incompatibility makes the relating of partial to global
structures a problematical task.
(Boulez 1971: 22-23)

I thought it had to be possible to retain absolutely the structural qualities of the old musique
concrète without throwing away the content of reality of the material which it had originally.
It had to be possible to make music and to bring into relation together the shreds of reality in
order to tell stories.
(Luc Ferrari interviewed in Pauli 1971: 41)

Chapter 4 of Pierre Schaeffer's Traité des Objets Musicaux (Schaeffer 1966)

is entitled The Acousmatic. According to the definition in Larousse, the
Acousmatics were initiates in the Pythagorean brotherhood, who were
required to listen, in silence, to lectures delivered from behind a curtain
such that the lecturer could not be seen. The adjective acousmatic thus
refers to the apprehension of a sound without relation to its source. It is
important in Schaeffer's development of the concept of the sound-object
that it be detached from any association with its source or cause. The
sound-object is to be analysed for its intrinsic acoustic properties and not in
relation to the instrument or physical cause which brought it into being.
However, in our common experience, we are more often aware of the
source of a sound than not and studies of behaviour and aural physiology
would suggest that our mental apparatus is predisposed to allocate sounds to
their source. We can see in a very crude way how this ability was essential
for our survival in the period before our species came to dominate the entire
planet. One needed to be able to differentiate between harmless herbivores
and dangerous carnivores, predator and prey, friend and foe. Even in the
cultured detachment of today's world, however, when we are listening to a
concert of instrumental music, except where the texture is very dense or
produced in a way which is novel to our ears, we are always very aware of
the instrumental source of the sounds we hear. We might, in fact, go a stage
further and notice that in the tradition of virtuoso performance our
awareness of the source and the performer's physiological, balletic and
dramatic relation to the source can become part and parcel of our aesthetic
reaction to the concert experience.
The formalisation of musical parameters, the lattice of the tempered
scale, the rhythmic co-ordination required by harmonic structuration, the
subordination of timbre to pitch and its streaming in separate instrumental
layers, is in many ways an attempt to negate the impact of the recognition
of the source (human beings articulating mechanical sound-sources) and
focus our attention upon the lattice logic of the music. Part of our
enjoyment of the music, however, remains an appreciation of the human
source of the sounds themselves (this is in a sense distinct from the
articulation of non-notated parameters of the sound through performance
gesture which we have commented on in earlier chapters).
In some contemporary instrumental music it has been possible for a
composer to specify a type of architecture and a mode of sound production
which limits the possible impact of gestural characteristics upon the
acoustic result, and when such music is heard on loudspeakers a large
degree of detachment from recognition of the source may be achieved for
some listeners. With music for voice, however, it is doubtful if we can ever
banish from our apprehension of the sound the recognition of a human
source for those sounds. Furthermore. that recognition often plays a
significant role in our perception of the music itself. For example, in the
Trois Poèmes d'Henri Michaux of Lutoslawski, our perception of a mass of
human ‘utterers’ is important in itself and becomes especially so in the
crowd-like sequence, the recognition of ‘crowd’ or ‘mob’ contributes
significantly to our aesthetic appreciation of the work (Example 7.1).
 At this stage, let us place these various characteristics of the sound
experience related to our recognition of the source of the sounds under the
general heading of landscape. It is important at this stage to differentiate the
idea of landscape from that of association as it is frequently used in
reference to programmatic music. Thus in our listening to the final
movement of Tchaikovsky's Manfred symphony we may be led (by the
programme note or otherwise) to associate the acoustic events with the idea
or image of the distraught Manfred wandering through the forest. It may in
fact be that there is some analogical relationship between our supposed
experience of the events described in the programme and our experience of
the acoustic events, perhaps related to the structure of musical gestures and
their relation to more universal human gestural experience (I would not like
to press this point too far!). The landscape of Tchaikovsky's Manfred is
however musicians playing instruments. Landscape, at least at this level, is
then a reasonably objective criterion, related to our recognition of the
source of the sounds. However, in the traditional Western repertoire there
are instances more difficult to classify. Generally speaking, these are
situations in which some degree of mimicry of non-instrumental or non-
human vocal sounds is attempted by the musical instrument or voice. In the
first example, from Janequin's Le Chant des Oyseaulx, human voices are
used to mimic the sounds of birds. This mimicry is however confined
within very specific limitations (Example 7.2 and Figure 7.1). In particular
the typical pitch ordering of Renaissance counterpoint is not contravened
and hence the typical pitch architecture of birdsong is not represented.
Rather an attempt is made to mimic the perceived articulation of bird calls
through the morphology of the syllables used for vocalisation. To the
modern ear, apart from the cuckoo, the effect is more like the setting of
sound-poetry to music than of an imitation of nature.
In the second example (Example 7.3 and Figure 7.2) from
Beethoven's Pastorale symphony, the same kind of bird imitation is
attempted on woodwind instruments. Here at least the spectral and attack
characteristics of the acoustic sources are more strongly related to the vocal
emissions of birds themselves. However, the mimicry is formalised to meet
the constraints of the rhythmic and pitch structures of the musical idiom.
We do not really hear birds. At this point the limitations of the concept of
association become apparent. We are aware that the landscape of these
sounds is musicians playing instruments, but we are also aware that there is
some attempt being made to mimic the natural sounds of birdsong. Do we
then ‘associate’ these instrumental sounds with birdsong, and if so in what
sense is this similar to the association of a progression of purely orchestral
sonorities with the programme note of the Manfred symphony discussed
above?
As a third example, let us consider Respighi's The Pines of Rome. In
this otherwise purely orchestral work, the composer overcomes the
problems inherent in attempting to imitate birdsong on traditional
instruments by introducing instead a gramophone recording of a bird into
the orchestral texture. In this case, then, there can be no confusion between
the source of the sound as instrument and the source of the sound the
instrument is attempting to mimic. We recognise the sound as coming from
a bird (it can of course be argued that the sound comes from a gramophone
record; this point will be discussed more fully below). Here, in fact, an
interaction of associationism with landscape takes place. The harmonic
melodic structures developed in the orchestra are meant to suggest, through
association, some kind of natural scene. The birdsong, however, presents it
quite directly.
Figure 7.1 Vocal bird imitations in Janequin's Chant des Oyseaulx.
 Figure 7.2 Instrumental bird imitations in Beethoven's Pastorale Symphony.

In our final example from Stockhausen's Trans, the orchestra is used

to suggest an alternative landscape (to that merely of people playing
instruments) in a different manner (Example 7.4). In his comments upon the
piece, Stockhausen has talked of a dream in which he perceived various
animated events as if through a veil or curtain. The piece begins with the
gradual unfolding of a cluster on the stringed instruments quietly from the
high register to a broad band cluster filling the entire range and played loud.
This sound persists almost without change throughout the entire piece and
‘rises’ again at the end in an approximate retrograde of the opening
sequence. In some senses, this composite sound-object may be considered
as an aural metaphor for ‘curtain’. Aurally speaking the effect is heightened
by the fact that the animated melodic material on wind instruments is heard
‘through’—or at least ‘against’—this wall of sound and the attempted
detachment from the typical apprehension of the orchestral landscape is
heightened by the use of a pre-recorded loom-shuttle which passes back and
forth across the acoustic space. This type of aural metaphor relates more
strongly to our notion of a natural morphology of musical objects (which is
to be discussed below) than to the more conventional idea of association. In
fact, Stockhausen attempts to distance the audience from the normal
perception of the orchestral landscape through the orchestral staging of the
piece. The metaphor of the curtain is paralleled by the use of an actual
curtain on stage and through the physical disposition of the string players
who are seated in a line across the stage, hiding the wind instruments, who
thus play ‘through’ the curtain, and they are instructed to bow exactly in
parallel with one another so that the individuality of the players is
subordinated to the theatrical (and sonic) image (see Figure 7.31).
Despite, then, its general playing-down in the discussion of musical
structure, the figurative has played a part in the traditional Western
instrumental and vocal repertoire. Furthermore, the relationship between the
figurative and the so-called abstract is not a simple one. A glance at the
development of painting would lead us to a similar conclusion. Up until this
century painting was generally thought of as essentially figurative. It is
interesting to note that the impact of technology on music and painting has
been in opposite directions: the tape recorder has introduced the
representational more easily into music, while the camera has tended to
replace the figurative role of painting and allow painting to pursue the non-
figurative domain.2

Virtual acoustic space: redefinition of landscape

With the arrival of sound recording the question of source-identification of

sounds became of great importance. On the one hand, it allowed the electro-
acoustic musician to isolate a sound physically from its producing medium
by recording it and hence enabled Schaeffer's conceptualisation of the
acousmatic. At the same time the reproduction of sound on loudspeakers
caused the question “what is the source of the sound?” to become
problematic.
Whereas previously we have defined the landscape of a sound to be
the perceived physical source of the sound, what are we to make of a
recording of Beethoven's Pastorale symphony played on loudspeakers? In
this case the physical source of the sounds is the vibration of the cones of
the loudspeakers. But, of course, the loudspeaker is able to re-produce
sounds from any other source so that noting that a sound originates from
loudspeakers tells us almost nothing about that sound, except that it has
been recorded. We must therefore seek a redefinition of the term landscape.
If the term is to have any significance in electro-acoustic music, we must
define it as the source from which we imagine the sounds to come.
 Figure 7.3 Layout of the orchestra in Stockhausen's Trans.

Thus, when we are sitting in the concert hall listening to the

Beethoven symphony the landscape of the sounds is the orchestra. When
we are sitting in our living room listening to a recording of the Beethoven
symphony the landscape of the sounds remains the orchestra. The
loudspeaker has in effect allowed us to set up a virtual acoustic space into
which we may project an image of any real existing acoustic space such as
that of the concert hall or, for example, in the case of a wildlife recording,
that of a wood at night.
The existence of this virtual acoustic space, however, presents us with
new creative possibilities. The acoustic space which we represent need not
be real and we may in fact play with the listener's perception of landscape.
This aspect of sonic architecture was not an aspect of the traditional craft of
the musician because, before the invention of sound recording, it was not
open to the composer or performer to control. It is therefore easy to dismiss
it by linking it with the somewhat cruder and culturally circumscribed
procedures of associationism (programme music) and mimicry which exist
as a somewhat marginal aspect of the central vocal and instrumental
tradition of Western art music. This, however, would be foolish. Not only
does the control and composition of landscape open up large new areas of
artistic exploration and expression, in the sphere of electro-acoustic music it
will enter into the listener's perception of a work regardless of the
composer's indifference to it.
One of the first composers to seize upon this new compositional
possibility was Luc Ferrari, whose comments were quoted at the opening of
this chapter. He described his approach as ‘anecdotal’. Presque Rien Nr.1,
from the vantage point of the musician the most radical work in this genre,
takes a recording of several hours of activity on a beach and compresses (by
selective choice) this material into twenty minutes, without in any way
attempting to negate or transform (other than enhance by concentration) the
perceived landscape of the beach. Alternatively, in the work Music
Promenade four independent linear montages of recognisable material
(including military music and marches, crowd sounds, conversations etc.)
are presented simultaneously and articulated by the occasional heightening
of sound-events through the use of echo or focus (for example, isolating a
single breath or a single phrase).
Parmegiani, on the other hand, developed a kind of surrealistic
approach to recognisable sources, lying more easily within the frame of
reference of musique concrète. As Simon Emmerson (1982: 125–126) has
noted, Parmegiani has used such devices as duality (recognisable/abstract),
perspective (near/far), and the juxta- or superposition of unrelated images to
achieve an effect similar to that of certain surrealist paintings (those of Max
Ernst, for example)—the intimation of real landscape without its actually
being present (Example 7.5 is the transition from Matières induites to
Ondes croisées from De Natura Sonorum). Duality refers to the use of
recognisable sounds which have then been reordered in such a way as to
appear almost abstract, for example as sound-objects with scintillating
overtones etc.. Changes in aural perspective on an object can be obtained by
recording it at a normal listening distance and then close-miking it. These
produce quite different acoustic results and when they are juxtaposed in the
aural landscape our sense of aural perspective is transformed. Finally,
unrelated images (e.g. water/fire or electronic/concrete) may be brought
into close proximity and by the use of similar musical gestures or the
creation of similar overtone structures made to have a very strong aural
interrelationship.
Even earlier, however, a similar concern with aural landscape
developed in the radiophonic departments of radio stations broadcasting
drama. Whereas Parmegiani was concerned to hint at possible landscapes in
an essentially acousmatic approach to sound-materials, radiophonics was, at
least initially, concerned exclusively with the suggestion of real landscapes
in the virtual acoustic space of the radio receiver which were not physically
present in the recording studio. In most cases these landscapes were
associated with spoken text or dialogue which itself might suggest a
landscape context. Often simple devices such as distance from the
microphone, the use of echo, or the sound of a door opening or closing
might be sufficient to ‘set the scene’ but as time went on more elaborate
scenarios were developed through seventeenth century battles to the purely
imaginary landscape of alien spacecraft.
These so-called ‘sound-effects’ fall largely into the category of what I
would call contextualising cues which permit us to identify a particular
landscape. Technically they may be divided into four categories: actuality,
staged, studio and mixed. Actuality involves, as its name suggests, an actual
recording of the event which is to be represented, such as a football match,
a crowded street, a clock striking and so on. Similarly, staged involves the
real staging of an event which could not be otherwise recorded for practical
reasons (a rare event, an event involving obsolete vocabulary or technology
such as a seventeenth century battle and so on). Studio sound-effects on the
other hand were simulations of actual sounds using materials far removed
from the original sound-sources. Thus in staging a medieval dual we would
involve two actors and actual swords of the period; this would be staged. To
generate the sound of horses’ hooves, however, it was typical to use a pair
of coconut shells in a tray of sand. For most listeners, and with the added
cue of human speech (possibly modified to suggest that the speakers were
being jogged up and down by horses), this was an acceptable acoustic
analogue of the sound of horses’ hooves. In some cases indeed studio-
produced sound-effects proved to be more acceptable than the real thing! In
the case of the horses’ hooves it may be largely a matter of getting a clear
recording of the type of horse movement that we want (how do we get away
from the sounds of modern traffic, how do we get the horse to behave in
exactly the way we want and so on). In the case of, for example, fire we
have a different problem. Even with an excellent digital recording, the
sounds of fire seem somehow less impressive when we hear them from a
recording than when we hear them in real life. This is something to do with
the fact that a fire is a multimedia experience. The impact of the visual
movement of flame and the heat from the fire are much more intense than
the impact of the fire sounds and the whole adds up to a very powerful
sensory experience. The recreation of the effect ‘fire’ by purely auditory
means, can simply fail to evoke the power of the multi-media image of fire.
In this particular case, where we are restricted to the medium of sound, the
use of studio fabrication (such as the recording of crinkled cellophane and
its subsequent speed-changing, filtering and mixing with other sources)
provides an aural image which is more acceptable than the real thing. This
apparent paradox is explained by the fact that in this case the aural image
has to partly replace our visual and tactile experience. It also points out that
the concept of aural landscape is far from straightforward!
Generally speaking, sound-effects have been used in this fairly
straightforward manner but on some occasions their use has been extended.
In 1978 the BBC presented a play whose dramatic and narrative content
was carried exclusively by sound-effects as there was no text present.3
Although the play demonstrated the inherent power of sound-effects
techniques, it appeared, however, as a poor substitute for a real text-based
narrative rather than the development of an entirely new art-form based on
the subtle composition of aural landscape. The only really creative use of
sound-effects seems to have been in the sphere of radio comedy where
absurd structures and events (such as endless staircases, maps as large as
cathedrals, and so on) were suggested by the simple transformation of aural
landscapes (Example 7.6). The appearance of science fiction drama
permitted the development of less obvious landscape devices but
radiophonics largely confined itself to simply exploiting the unusualness or
novelty of a sound-object (which very rapidly became clichéd and
predictable). Before the very particular quality of aural landscape
manipulation could be explored and developed (see below), the
concreteness of the television image took over and radiophonics declined to
an even more marginal position. The use of landscape concepts in
radiophonics is, however, very important in our attempt to conceptualise the
area of aural landscape for the purposes of sonic composition.
Finally, we should also mention that the landscape technique of
mimicry has much wider application than its use in music would imply.
Mimicry is often used in everyday conversation and in formal presentations,
for example by comedians or sound-poets, as a means to indicate an
alternative source for the information than the actual speaker or mimic.
Thus the imitations of accents, age and sex characteristics of vocalisation,
or even the imitation of birds (by whistling) or dogs may be catalogued.
Mimicry is not, as one might suspect, an entirely human occupation. It has
been observed in the behaviour of birds, both in imitating humans (human
whistling) and in the imitation of other species of birds (usually predators,
which suggests some kind of denotative function).

Defining characteristics of landscape

To briefly recapitulate, we have defined the landscape of a sound-image as

the imagined source of the perceived sounds. The landscape of the sounds
heard at an orchestral concert is ‘musicians-playing-instruments’. The
landscape of the same concert heard over loudspeakers is also ‘musicians
playing instruments’. In some cases it is difficult to identify the source of
the sounds. This fact is particularly pertinent to music designed for
projection over loudspeakers. When listeners who are accustomed to
listening to music at concerts where instrumentalists perform attend
concerts of electro-acoustic music projected on loudspeakers they often
express a sense of disorientation. This is usually attributed to the lack of
any visual focus at the concert. However, it seems clear that this reaction is
prompted by an inability to define an imaginable source, in the sense of a
landscape, for the sounds perceived. This sense of disorientation produced
in some listeners by the impact of electronic sounds was the basis of the
early use of electronic sound-materials for science fiction productions. The
inability of the listener to locate the landscape of the sounds provided the
disorientation and sense of strangeness which the producer wished to instil
in the listener. The development of the concept of the acousmatic and the
general tendency in mainstream musique concrète to destroy clues as to the
source or origins of the sounds can be seen as a specific reaction to the
problem of landscape in electro-acoustic music.

Figure 7.4 Defining characteristics of landscape.

What aspects of our perception of an aural image enter into our

definition of landscape for the sounds we hear? We may effectively break
down our perception of landscape into three components (see Figure 7.4)
which are not, however entirely independent from one another. These are:

(1) the nature of the perceived acoustic space;

(2) the disposition of sound-objects within the space;
(3) the recognition of individual sound-objects.

We will discuss each of these in turn.

In practice the nature of the perceived acoustic space cannot be
separated from our perception of the sound-objects within it. We obtain our
information about, for example, the reverberant properties of the space, by
hearing out the temporal evolution of the sound-objects within the space
and, for example, the different reverberation times of different objects
within the space may give us further clues as to the overall acoustic quality
of the implied sound-environment. Furthermore, the acoustic of the space
may change and in some works such changes may be organised in such a
way as to prevent us building up any overall acoustic image. Although this
‘destruction’ of landscape has been an aim of certain kinds of electro-
acoustic music composition, we will here put it on one side, as a very
special case, and deal with landscape as normally experienced.

Landscape: the nature of the acoustic space

Usually, any sort of live recording will carry with it information about the
overall acoustic properties of the environment in which it is recorded. These
might include the particular resonances or reverberation time of a
specifically designed auditorium or the differences between moorland (lack
of echo or reverberation, sense of great distance indicated by sounds of very
low amplitude with loss of high frequency components etc.), valleys
(similar to moorlands, but lacking distance cues and possibly including
some specific image echoes) and forests (typified by increasing
reverberation with distance of the source from the listener). Such real, or
apparently real, acoustic spaces may be recreated in the studio. For
example, using the stereo projection of sounds on two loudspeakers we may
separate sound-sources along a left-right axis, creating a sense of spatial
width. We may also create a sense of spatial depth by simultaneously using
signals of smaller amplitude, with their high frequencies rolled off.
Depending on which type of acoustic space we wish to recreate, we might
also add increasing amounts of reverberation to these sources, the more
distant they appear to be. In this way, depth is added to the image and we
create an effective illusion of two-dimensional space. This illusion is
enhanced if sound-objects are made to move through the virtual space (see
Figure 7.5). (A detailed discussion of the control of spatial motion will be
found Chapter 10.)
The digital technique known as convolution allows us to impose in a
very precise manner the acoustic characteristics of any preanalysed sound-
environment upon a given sound-object. Ideally, the sound-object itself
would be recorded in an anechoic environment. To implement the process
of convolution we begin by measuring the impulse response of the acoustic
environment we wish to recreate. This involves playing a single very brief
impulse (Figure 7.6a1) with a very broad and flat spectrum (see Chapter 3),
e.g. a gunshot, into the natural acoustic environment and recording the
result; this is represented in Figure 7.6a2 as a series of ‘instantaneous’
digital samples - the impulse response of the environment. Provided all
audible frequencies are equally represented in the initial impulse, the
resulting recorded signal should indicate the overall resonance
characteristics of the environment. (If the impulse is specifically pitched we
will be measuring only the resonance characteristics of the environment at
some particular frequency.) Let us now assume that we have a digital
recording of our sound-object. Because digital recording involves a
sampling process, the sound-object may be regarded as a collection of
instantaneous impulses which taken together define the overall waveform of
the sound-object. If we now replace each individual sample in the digital
representation by the graph of its impulse-response (which will be
magnified or reduced according to the amplitude of each impulse) and then
sum at each sampling instant the resultant values (see Figure 7.6b4) we
obtain a waveform corresponding to the sound perceived as if the sound-
object had been recorded in the sound-environment which we analysed.
 Figure 7.5 Representation of depth in a stereo field.

Sound recording and the presentation of recorded material has,

however, brought with it a number of other acoustic spaces which are
conventions of a mode of presentation. We shall refer to these as formalised
acoustic spaces to distinguish them from the real acoustic spaces we have
previously been discussing. There is, of course, not a clear dividing line
between these two categories because, as broadcast sound becomes an ever
more present part of our real sound-environment, it becomes possible to
question, for example, whether we hear orchestral music, the sounds of a
radio (playing orchestral music) or the sounds of a person walking in the
street (carrying a radio (playing orchestral music ))! These questions are not
trivial when we come to discuss electro-acoustic music, such as
Stockhausen's Hymnen, which attempts to integrate existing recorded music
into its sonic architecture. And in fact some sounds which have been
modified by the technological means of reproduction might almost be
accepted into the category of real acoustic space, for example, the sound of
a voice heard over a telephone or the sound of a voice heard over a distant
distorting channel as with the sounds of voices transmitted from space.

Figure 7.6a Generating the impulse response of an environment.

 Figure 7.6b The process of convolution.

The formalisation of acoustic space is found in all kinds of

contemporary music production. The negation of any consistent landscape
in some kinds of electro-acoustic music has been mentioned previously. In
the studio recording of rock music albums, various formal manipulations of
the acoustic space are taken for granted. The most obvious is the
rebalancing of instruments by means of differential amplification. A soft-
spoken singing voice or a flute may become much louder than a whole
brass section. A piano, initially heard in close perspective, may ‘duck’ to a
lower dynamic level at the instant at which a singing voice enters. Popular
music is also very much concerned with the sense of intimacy or distancing
involved in the recording of the voice. Electrical amplification permitted
‘crooners’ to sing at a very low level, and hence adopt the vocal
characteristics of intimacy only available at low vocal amplitudes, yet still
be heard against a large orchestra. At the other extreme, singers commonly
use effects of reverberation, echo (often in a self-consciously clearly non-
natural way) and phasing (an effect that is confined in the natural world
almost exclusively to our perception of aeroplanes passing overhead!) to
distance themselves from the listener or appear larger than life. Such
techniques have parallels in certain pre-literate cultures. However, as these
techniques are usually applied on an all-or-nothing basis in a particular song
and are used extremely commonly, they begin to cease to have any
landscape ramifications and become conventions of a formalised musical
landscape. We begin to perceive these things as new timbres in a formalised
acoustic space; they become no more unusual than the rebalancing of
instruments. At this stage in the development of these devices, they can
only become useful tools for the elucidation and elaboration of landscape
properties of our perception of sonic art if we abandon the all-or-nothing
usages and investigate the effects of transformations between different
representations of the voice, having at the same time a sensitivity to the
concept of aural space.
This kind of formalisation of acoustic space does not apply merely to
music. It is an aspect of the formatting of productions for radio (or TV)
broadcasts. A typical formalised landscape might be that of the disc jockey
presentation of rock music. We are not meant here to recreate an image of a
person sitting in a studio speaking into a microphone and putting records on
turntables (in fact in many Californian rock stations this has now been
entirely replaced by automated synchronised tape recordings). We have
merely a formalised conventional structure. The record ends or draws to a
close, the voice of the disc jockey enters and speaks. This is often mixed
with special snippets of music or sound-effects; a kind of logo in sound, or
pure linkage material—‘keep the music playing’—until the next song
begins. This controlling voice floating in a sea of musical snippets has now
become a broadcasting convention and has no special landscape
implications. It is interesting, however, to speculate what one might do by
using this conventionalised format and beginning to break down the
underlying conventions. This would be simple to do in the direction of the
abstract (in the sense of a space which is neither universally accepted as a
convention nor real) but might be more interesting if moved in the direction
of the real or the surreal (see below).
 Finally, we might consider the case of electro-acoustic compositions
such as Stockhausen's Telemusik or Hymnen which use elements of
recognisable recorded music within a sonic architecture using a wider
sound-palette. In both pieces the absorption of finite portions of
recognisable recorded musical forms within an ongoing stream of
electronically generated sound-materials tends to suggest a particular sort of
relationship to the preexisting music i.e. their absorption in a larger process
and hence a view from outside or above the cultural substratum from which
the musics come. We are here in a sort of ‘cosmic’ media space which,
generally speaking, has no real-world reference. This ‘distancing’ or
detachment from the real (no recognisable acoustic space, no recognisable
real-world referents, although some kinds of natural processes are
suggested in Hymnen) predisposes us to perceive the pre-recorded musics
(themselves often heavily distorted) in a distanced way. As, however, they
are the only elements which refer directly to our experience of real acoustic
spaces, we are viewing the real world as if at a distance. Such generalised
use of landscape phenomena (i.e. the sense of detachment underlined by the
lack of real-world referents) tends to become accepted as convention as
time goes by and this aspect of Stockhausen's aural metaphor may not
survive (it may be perceived merely in formal terms) as our familiarity with
electronic sound-sources increases. Already in our discussion of Hymnen
we are touching on the areas of sound-source recognition and the vantage
point of the listener which we will deal with more fully below.

Landscape: the disposition of sound-objects in space

Given that we have established a coherent aural image of a real acoustic

space, we may then begin to position sound-objects within the space.
Imagine for a moment that we have established the acoustic space of a
forest (width represented by the spread across a pair of stereo speakers,
depth represented by decreasing amplitude and high-frequency components
and increasing reverberation) then position the sounds of various birds and
animals within this space. These sound-sources may be static, individual
sound-sources may move laterally or in and out of ‘depth’ or the entire
group of sound-sources may move through the acoustic space. All of these
are at least capable of perception as real landscapes. If we now choose a
group of animals and birds which do not, or cannot, coexist in close
proximity to one another, and use these in the environment, the landscape
would be, ecologically speaking, unrealistic but for most listeners it would
remain a ‘real’ landscape.
Let us now begin to replace the animal and bird songs by arbitrary
sonic objects. We might accomplish this by a gradual process of substitution
or even by a gradual transformation of each component of the landscape. At
some stage in this process we begin to perceive a different kind of
landscape. The disposition of the objects remains realistic (in the sense that
we retain the image of the acoustic space of a ‘forest’) yet the sound-
sources are not real in any sense of the word. Here we have the first
example of an imaginary landscape of the type unreal-objects/real-space.
If we now take the original sound-objects (the animal and bird
sounds) and arbitrarily assign different amplitudes and degrees of
reverberation or filtering to each occurrence, we achieve a second but quite
different kind of imaginary landscape of the type real-objects/unreal-space.
If we now imagine a more extreme example of the ecologically
unacceptable environment (!) described earlier, we arrive at a third type of
landscape. For example, imagine that, by appropriate editing and mixing
procedures, we are able to animate a duet between a howler monkey and a
budgerigar or a whale and a wolf, we have a landscape in which the sound-
sources are real and the perceived space is real, yet the relationship of the
sound-images is impossible. This bringing together of normally unrelated
objects in the virtual space created by loudspeakers is closely parallel to the
technique of bringing together unrelated visual objects in the space defined
by a painting, a technique known as surrealism and I therefore propose to
call this type of imaginary landscape (real-objects/real- space) surrealist.5
Example 7.7 from the television soundtrack of my Automusic
illustrates one of these landscape genres. Earlier in the same piece, we have
heard the sound of traffic which is characterised both by the typical spectra
type of motorised traffic and the typical rise and fall in amplitude with the
intervening Doppler pitch shift which is a feature of the movement of the
sources through space.6 Here at the end of the piece the amplitude envelope
and Doppler shift attributes of the motion of traffic are retained but the
sounds themselves are not clearly recognisable (they are in fact aggregates
of vocally-derived sounds). This, then, is an example of unreal sounds/real
space landscape. (Note that the ‘reality’ of the aural space has been
established by its earlier presentation in an entirely real context.)
 Motion in space may also be used to alter the perceived characteristic
of a landscape. In Chapter 3 we discussed some aspects of aural imaging
and described a sound which began life as an oboe situated in the centre of
the stereo space but then apparently diverged into a ‘cello and a voice-like
sonority which moved off towards the two separate speakers. Such imaging
phenomena as this do not occur in our everyday aural experience, but can
be generated by precise computer control of sound-materials. Even the
rapid contrary motion of several voices in a quadraphonic (sound-surround)
or three-dimensional space (without the accompanying sound of running
feet!) defies our everyday intuition. As has been mentioned previously such
devices, if used indiscriminately and in an all-or-nothing fashion, become
accepted as mere conventions of electro-acoustic art. If, however, we can
learn to control with subtlety the transition from one type of acoustic space
and one type of spatial disposition of sound-objects to another, control of
landscape and spatial motion can become an important structural and
expressive ingredient in sonic art. Once more than one sound is set into
motion in the same virtual acoustic space, we may begin to consider the
interaction of different kinds of spatial motion. The topic will be dealt with
more fully in Chapter 10.
The change in the apparent disposition of sonic objects in the acoustic
space may alter the perspective of the listener. For example, where various
sound-objects move in different directions, or a single sound spins around
the listener's position (in a quadraphonic space), the listener may reasonably
assume himself to be at a point of rest. If, however, a number of sound-
objects are spaced at various points of the compass around a quadraphonic
space, and the entire frame of reference made to rotate, we might suggest to
the listener that he himself is spinning in the opposite direction whilst the
frame of reference remains still. At a simpler level, differences in amplitude
and also in timbral qualities caused by the closeness or distance of the
microphone to the recorded object alter not only the listener's perceived
physical distance from the source but also the psychological or social
distance. With vocal sounds, depending on the type of material used,
closeness may imply intimacy or threatening dominance, distance a sense of
‘eavesdropping’ or of detachment and at various stages in between a sense
of interpersonal communication or more formalised social communication.
A similar kind of psychological distancing which parallels the social
distancing may be experienced even in the case of inanimate sound-sources.
To hear sounds in our normal acoustic experience in the same perspective
that close-miking provides we would usually need to be purposefully
directing our aural attention to the sounds (by, for example, bringing our ear
very close to the sounding object). Listening to sounds recorded in this way
can thus produce the effect of perception through an aural ‘magnifying
glass’ and is quite different from our experience of normal acoustic
perspective. Changes in aural perspective in the course of an event or
extended composition are of significance to the listener and need not be
approached in a purely formalist manner.
The sense of perspective or distancing we are discussing here is quite
different from that discussed earlier in relation to Stockhausen's Hymnen.
Here we are referring to perceptible acoustic properties of our aural
experience which relate to the distance of the sounding object, not merely to
a general sense of mental detachment caused by a landscape with no
perceptible referents. Let us speculate on another landscape device whereby
Stockhausen's sense of distancing from the real might be achieved which
uses true aural perspective. Imagine that we create a two-dimensional stereo
image (using stereo width with amplitude and equalisation for depth) of a
group of people which we project as a narrow image in the centre of a
stereo pair of loudspeakers in front of the listener (see Figure 7.7). We now
increase the loudness and high frequency response of the four signals and
move them in such a way that at their maximal point they are distributed
over the four loudspeakers of the (sound-surround) quadraphonic space.
The transformation continues until the image (now reversed in depth) is
situated in a narrow band in the middle of the rear stereo pair. Here we have
created the illusion of flying through space, either of the aural image of the
group of people (front to rear) or of the observer (flying forwards) and
graphically underlined his detachment from the group of people who
remain fixed in their position in relation to each other.
 Figure 7.7 The illusion of ‘flying through space’.

Landscape: the recognition of sources

The final element in our definition of landscape is our ability to recognise

(or not recognise) the sound-objects themselves. With sound-objects
specifically generated by electronic synthesis or musique concrète
techniques, the concept of recognition may be extremely problematic, but
with sound-sources recorded directly from the acoustic environment,
recognition would seem to be a fairly straightforward matter. Even with
superb digital recording, however, this is not necessarily the case.
In our normal working lives our experience of the environment is a
multi-media one. In particular we rely very heavily on the visual medium to
assist in our recognition of objects and events. This recognition may be
direct in the sense that we see the object or event which emits the sound, or
indirect in the sense that a physical location (e.g. a railway station or a
particular type of terrain) or a social occasion (e.g. a concert performance)
may enable us to identify a, perhaps indistinctly heard, sound-source. Once
we remove the visual and other clues to sound recognition, we must rely
entirely on our aural perspicacity. In the virtual acoustic space of an electro-
acoustic work projected on loudspeakers, even where sounds directly
recorded from the real world are presented, they may be completely devoid
of aural context. Furthermore, where sounds which would normally be
individually recognisable are placed in an imaginary space or in an
imaginary relationship to one another, the problem is compounded.
Note here that what we might call the surrealism effect is usually
quite different in the sphere of sonic art to the same phenomenon in the
visual arts. Normally speaking in a surrealist painting, the juxtaposition of
everyday objects in an ‘unnatural’ way does not interfere with our
recognition of the individual objects. In fact, the impact of such images
arises partly from the very recognisability of the objects themselves. In
sonic art, however, this kind of juxtaposition may make it difficult for us to
identify the source of the sounds, to recognise the individual objects. This
particular feature of sonic art will be discussed in more detail later. For the
sake of the following discussion we will make somewhat arbitrary
distinctions between intrinsic recognition and contextual recognition. It
may also be generally assumed that throughout this section we are
discussing sounds directly emitted by natural events, mechanical vibrations
or utterances in the real world and not specially-contrived sonic objects
made in the electroacoustic studio.
Certain sounds retain their intrinsic recognisability under the most
extreme forms of distortion. The most important sound of this type is the
human voice, and particularly human language as the formant structure of
the human voice itself has a high intrinsic recognisability for human beings.
This is partly due to the obvious immediate significance of the human voice
to the human listener, but also the unique complexity of articulation of the
source. The ability to produce a rapid stream of timbrally disjunct entities is
uncharacteristic of any other source (except perhaps bird mimics of human
beings such as parrots). We may mistake electronic or mechanical
vibrations of certain types for the sounds of crickets (though with increasing
aural sensitivity, this confusion can be reduced considerably) because the
number of features entering into the articulation of the particular sound-
objects produced by crickets is quite small. Without any context, we may
recognise a human voice, even where its spectral characteristics have been
utterly changed and it is projected through a noisy or independently
articulated channel. However, given a stridulation type of sound in
isolation, it might be difficult for us to identify it as cricket, electronic
oscillator or mechanical vibration. If the latter three sounds were heard
simultaneously or in quick succession, we would most probably be able to
differentiate between the sources, but in this case it is context which is
enabling recognition. By setting up a context of other sounds (for example,
the sounds of frogs or other creatures living in the same habitat as crickets)
within certain limitations it would be possible to pass off the electronic or
mechanical sound as the sound of a cricket.
Furthermore, recognition cannot be entirely separated from
disposition in space. In the making of Red Bird I wished to include the
sound of a fly. Although it might seem a relatively simple matter to
approximate sufficiently closely the sound through an electronic or even a
filtered vocal source, this proved very difficult. The problem was
aggravated by the fact that I wished to achieve a transformation from a
recognisable vocal sound into the sound of the fly buzzing. Although in an
isolated context the transformed vocal sound might have been acceptable as
the sound-image ‘fly’ to the listener, in a context in which the voiced
sibilant ‘zzzz’ occurred, the vocal source of the ‘fly’ sound became
transparent. After much frustrating work I discovered a biology researcher
working with bluebottles (a large blue-coloured fly). As all a bluebottle's
sense receptors are in its feet, it was possible to attach their backs to rods.
When a bluebottle was thus lifted off a surface, it began to ‘fly’ but being
attached to the rod, of course, did not move. (Incidentally, after such
experiments the bluebottles could be removed from the rods and flew away
quite happily!) It was therefore possible to make a recording of a bluebottle
in flight with the microphone at a fixed distance from the creature.
Replaying the sound in the recording studio, however, it bore little
resemblance to our typical percept of a ‘fly’ sound. Only at sufficiently low
amplitude and with treble roll-off (a distance cue) and in particular, spatial
motion, both laterally across the stereo image and in depth through changes
in amplitude, could the aural image ‘fly’ be recreated. The sound thus
redistributed in space resisted fusion and confusion with its vocal imitation
and was even recognisable in unlikely aural landscape contexts. In certain
other cases our recognition of a group of sources may be strongly correlated
with a mutual disposition of motion through the acoustic space.
In a second example from Red Bird I wished to achieve the
transformation of a sound-object from a recognisable book being slammed
on a surface to a recognisable door being slammed. In this particular
instance the door-slamming sounds could be made recognisable by subtle
cues related to the mechanical noises of door-handles. Door sounds of this
type could thus be differentiated from arbitrary ‘slam’ sounds. An arbitrary
‘slam’ sound might be taken to be a door if presented in the context of
doors-with-handles. A book-slam is, however, a fairly unspecific aural
image. In order to ensure recognition of this sound (particularly in a context
where it would be juxtaposed with slams from other sources) it was
necessary to provide a contextualising cue, in this case the sound of the
pages of a book being turned which has a much higher intrinsic
recognisability.
Contextual cues may not only change our recognition of an aural
image, but also our interpretation of the events we hear. As a simple
example imagine a recording of a vocal performance accompanied by
piano. Imagine that the vocal performer uses many types of vocal utterance
not normally associated with the Western musical repertoire, for example,
screaming, glossolalia, erotic articulation of the breath etc.. The presence of
the piano in this context will lead us to interpret these events as part of a
musical performance, perhaps the realisation of a pre-composed score. The
utterance will lie within the formalised sphere of musical presentation. If,
however, we now were to hear a similar recording in which the piano were
not present and no other clues were given about the social context in which
the vocalisations occurred, we might not be able to decide whether we were
listening to a ‘performance’ in the above sense, or ‘overhearing’ (for
example) a direct utterance of religious ecstasy or the ravings of an insane
person! The latter brings us onto the subject of utterance and the way our
perception of the intention of utterance influences our response to the aural
events. This will be discussed in Chapter 11.
Contextualisation has its inverse in the phenomenon of masking. Put
simply, one sound may ‘get in the way of our perception of another. A very
loud sound may prevent us from hearing quieter sounds. Loud sounds in
particular frequency bands may prevent us from hearing those frequencies
as constituents of another sound. This kind of masking is of particular
importance in the virtual acoustic space projected on loudspeakers as we
have no visual cues from which to get alternative information. The building
of a representational aural landscape is not therefore merely a matter of
assembling the appropriate recognisable sound-objects. Spatial and
temporal disposition (and relative placement within the frequency
spectrum) will influence our ability to recognise individual objects and
must therefore enter as compositional criteria in sonic art projected on
loudspeakers. In its simplest form, this fact is appreciated in the rock
recording studio, where spectral emphasis (through the use of filters) is
used to enhance the ‘visibility’ of each timbral layer in a typical multi-track
production. Masking is, however, not a totally passive phenomenon. As
experiments in both visual and aural perception have demonstrated, our
brain has the ability to reconstruct certain kinds of images and messages,
even where these are heavily masked (see Figure 7.8). This
reconstructability applies particularly to objects with a clear and well-
known structure, for example pictures of objects or icons with which we are
very familiar or the structured sound-stream of language or music where we
are familiar with the ‘text’ or at least with its style. It can be shown that the
brain is capable of reconstructing a message from partial information, even
where the remainder of the message is not present. As is illustrated in
Figure 7.9, the brain may assume that an image or soundstream continues
behind a masking object or sound (and hence reconstruct the image or
soundstream) even if this is not the case. Thus, if elements in a linguistic
stream are edited out and replaced by bursts of loud noise, the listener will
perceive the original text as if white noise had merely been superimposed
on an uninterrupted text stream. The same results are discovered with
musical material where the listener can reasonably predict the sonic
progression of events, for example, within a well-known musical style or in
a simply articulated musical process such as a continuous transformation
from one state to another.

Figure 7.8 Visual masking.

 In another situation I wished to create the aural image of a machine.
A typical mechanical machine which repeats the same process over and
over has certain acoustic characteristics superficially resembling the
linguistic stream, thus sound-objects of quite different timbral
characteristics may follow each other in rapid succession, just as in the
linguistic stream. The machine differs, however, not only in the regularity
of its repetition (which can be closely approximated in vocal performance)
but in its lack of a characteristic defining formant structure: at most, the
sound might take on the overall resonance structure of the room in which it
is situated. In this particular instance it was necessary to construct a
convincing aural image of a machine, but using as the sound components
the syllabic constituents of a certain phrase. A direct aural imitation of a
machine sound is, in fact, just that. We have the aural image of vocal
mimicry, not of the machine. The mere repetitive formalisation of vocal
material is therefore not adequate. To distance the aural image from the
vocal source material, each syllabic component was therefore articulated
separately with quite different spectral and other articulatory characteristics
and the resulting elements differently filtered, spliced and mixed on tape.
The latter procedures destroyed the intrinsic continuity of the vocal stream,
which persists despite the linguist's emphasis on the discrete nature of
phonemic units (see later chapters, especially 13). Given all of these factors,
however, it was necessary to use further contextualising cues before the
aural percept ‘machine’ was received by the listener. In this case the typical
reverberant acoustic of a factory was added and finally, at very low level,
various other mechanical sounds, recorded directly in a factory situation,
were added to the aural landscape acting directly as contextual clues.

Images
Figure 7.9 Mental reconstruction of an image from masked data.

The control of recognition in this latter example is of a different order

to that discussed in relation to Parmegiani's ‘surrealism’; in that case we
were concerned with a subtle interplay between the recognisability and
lack-of-recognisability of various sources. Here we are concerned with
transforming our recognition of an object. We take an object which is
recognised as one thing and transform it in such a way that it is recognised
as another (rather than just not recognised). This opens up the concept of
the composition of landscape to include the notion of landscape
transformation, not merely in the simple dimension of recognition or lack of
recognition, but in the transformation of one aural image into another. We
can thus contemplate the construction of a mode of composition concerned
with the interplay of recognisable aural images which is quite different from
the use of degrees of recognisability in an otherwise formalised musical
landscape.

Transformation: the intrinsic ambiguity of aural space

With an understanding of various properties of the aural landscape, we can

begin to build compositional techniques based upon transformations of the
landscape. Digital technology offers us tremendous power to manipulate the
inner substance of sounds so that transformations between different
recognisable archetypes can be effected with great subtlety. For example,
the computer language Chant permits us to generate a transition from a
bell-like sonority to a male voice by simultaneously manipulating the
relative amplitude and width of the formants and the rate of occurrence of
impulsions, which when sufficiently fast, correspond to the perceived pitch
of the voice (Example 7.8).
The next two examples (Examples 7.9 and 7.10) are taken from Red
Bird and employ pre-digital techniques. In the first example we hear the
sound of a book being slammed on a surface (in an attempt to swat a fly)
gradually changing into the sound of a door slamming. This transformation
depends heavily on the use of contextualising cues. Thus, as mentioned
previously, the sound of the book slam, if heard in isolation, is not easily,
uniquely assignable to the source ‘book’, no matter how realistic the
recording may be. The aural image interpretation ‘book’ is thus imposed on
the sound when first heard by the prior contextualising cue ‘sounds of book
pages being turned’. The sounds of door-slams, however, may or may not
be easily assignable to the source ‘door’ as discussed above. The
introduction of handle noises makes this assignment easier. We can
therefore effect a transition from a door-slam with no extra cues to a door-
slam with very clear handle noises, gradually introducing the sound of the
handle and thus making identification as ‘door’ increasingly definite. In the
acoustic image intervening between these two extremes the assignment of
the sound ‘slam’ becomes difficult, especially where we use speed changes
and the interleaving of the two types of slam sound (see Figure 7.10). The
ambiguity of the source can be resolved only by context clues which allow
us to move from the interpretation ‘book’ to the interpretation ‘door’.
In a second example, the vocal syllable ‘lisss’ transforms into the
sound of birdsong. This transformation is achieved by a mixture of vocal
mimicry and textural melange (a kind of self-masking) as indicated in the
diagram (Figure 7.11). In this case the two poles of the transformation are
clearly recognised without contextualising clues and the transformation
passes through an artificial acoustic space generated by mixing it into a
mono stream.
It is interesting to compare these aural transformations with similar
transformations which might be effected in visual space (for example in
animated cartoons). First of all we should note that there is not necessarily
any parallel in visual space of a transformation in aural space. Although the
change ‘book’ to ‘door’ may be quite easily conceived in visual space, the
processes involved in the transformation of spatial form are quite different
from those involved in the transformation of the acoustic object. The fact
that book shape and door shape are similar, and at the same time book
sound and door sound are similar is a happy coincidence in this case. There
is no intrinsic correlation between geometrical shape (by itself) and sonic
properties. The distinction is even more evident when we consider the
second example. What could be the visual equivalent of this
transformation? Is it a mouth which changes into a bird or some
representation of breath which becomes a bird. It soon becomes clear that
seeking such a parallel is rather fruitless. Visual transformation and aural
transformation take place in different dimensions. Not only this, they have
quite different qualities. We normally have little difficulty in recognising a
visual object, even at a glance. Rapid transformations between clearly
recognisable objects are therefore quite simple to achieve in visual
animation (speaking aesthetically rather than technically!). The whole
process has a very ‘concrete’ feel, a certain definiteness. Aural images,
however, almost always remain a little ambiguous. We can never be sure
that we have made a correct recognition, especially where the
transformation of sound-objects denies us the normal contextual cues.
Transformation in aural space therefore tends to have a ‘dreamlike’ quality
removed from the concrete, distanced and often humorous world of visual
animation. At the same time, although transformations of abstract forms or
between recognisable and abstract forms may be achieved in the visual
sphere, these tend not to have the same affective qualities as time processes
taking place in the domain of sound. Landscape composition, therefore, has
a quite different ‘emotional’ feel to the sphere of visual animation.

Images
Figure 7.10 Schematic representation of ‘book-slam’ transforming into ‘door-slam’ (from Red Bird).

Images
Figure 7.11 Transformation of ‘Liss-’ (from ‘listen’) into birdsong (from Red Bird).

The intrinsic ambiguity of aural space also means that certain kinds
of transformations may be effected in aural space which it is very difficult
to relate in any way to a visual analogue. In the piece I am sitting in a room
by Alvin Lucier (Example 7.11) the initial sound-image is that of a voice
speaking in a room with a given acoustic (at this stage our attention is not
drawn to the room acoustics). The voice is then recorded and the recording
played back into the room. This process is repeated over and over again. As
this process proceeds the recording becomes increasingly coloured by the
room acoustic until finally at the end of the piece we hear essentially the
room resonance vaguely articulated by the amplitude fluctuations of the
voice. In this case our perception of what is the sound-object and what is
the acoustic space in which it is projected have been conflated. At the
beginning of the piece we would unreservedly state that the sound-object is
the voice. At the end of the piece the sound-object is clearly a more
‘abstract’ entity whose characteristics derive from the room acoustic.
Somewhere in between these extremes our perception passes over from one
interpretation to the other. Not only, therefore, can we control the
dimensions of, on the one hand, simple recognition/non-recognition and on
the other hand recognition-as-A/recognition-as-B, but also the dimension
acoustic-space/‘sound-object within an acoustic space’.
From what has been said so far about the intrinsic ambiguity of aural
space, it might seem unlikely to be able to generate an aural image which is
specifically ambiguous, i.e. which has two very specific interpretations.
This, however, can be quite simply achieved in certain cases, particularly
where one of the sound-sources is the human voice. Thus we may use the
vocoder to impose the articulatory structure of unvoiced speech onto, for
example, the sound of the sea. The two recognisable aural images remain
simultaneously perceptible in the resulting aural stream. Similarly the
digital technique of cross-synthesis allows us to transfer certain
characteristics (e.g. the changing formant structure) of one recognisable
sound-source onto another recognisable source, creating sound-images
which demand two simultaneous landscape interpretations.
One final comment on ambiguity and recognisability: we might ask
the question, what enables us to recognise a sound as ‘like’ another? What
is the aural basis of mimicry? If we are capable of perceiving that certain
sounds, even so-called abstract sounds, are similar to other, possibly
concrete, sounds, does this not then affect our perception of all sonic
structures? Is there any kind of relationship between our perception of the
morphological properties of the natural sonic landscape and our
appreciation of composed sound-events and sound-structures? This point
will be discussed more fully in the following chapter.

Landscape in music: some sketches

Landscape concepts have now been used extensively by composers of

electro-acoustic music. In Luc Ferrari's Presque Rien Nr.1 (Example 7.12)
an existing landscape is simply recorded on tape and concentrated in
intensity by compressing the events of a morning into the space of twenty
minutes. The landscape itself remains entirely realistic, except that our
attention to it is concentrated by the condensation of materials already
mentioned and also by its being brought into the concert hall or living room
as an independent sonic landscape. In Larry Wendt's From Frogs we hear a
landscape consisting of the songs of many frogs but in this case these songs
are slowed down and hence the rich content of the spectrum becomes more
apparent to our ears (Example 7.13). The effect is similar to that of close
perspective (recording sounds with the microphone placed extremely close
to the source). We have a sense of entering into the inner details of the
sound-world. The landscape itself remains ‘natural’, except for the effect of
magnification resulting from the slowing and lowering of the frog sounds.
We appear to be in a landscape of giant frogs! A similar effect can be
achieved by slowing down birdsong. Roughly speaking the pitch range of a
bird is related to the length of its windpipe which is similarly related to its
overall size. There seems also to be a relationship between size and the
speed of articulation. Hence, slowing down the song of a small bird, we
hear what appears to be the song of a large bird. This process can be
extended a long way. I have in fact used birdsong at a sixteenth of its
original speed, yet without destroying the source-image “bird”. Human
beings, on the other hand, do not come in such a vast assortment of sizes
and even at double or half speed we are no longer prepared to accept vocal
sounds as naturalistic.
In Alvin Lucier's I am sitting in a room (discussed above), on the
other hand, the acoustic characteristics of the physical space in which the
sound-object is first presented are brought to the centre of our attention by
the process of natural resonant feedback. Lucier's approach to acoustic
space is literal and objective. He is concerned with drawing our attention to,
and utilising, the acoustic characteristics of existing spaces, rather than
establishing a virtual space through the medium of the loudspeaker. His
work also tends to objectively demonstrate or elucidate these characteristics
rather than attempting to utilise them for some other expressive purpose.
In Michael McNabb's Dreamsong (Example 7.14), however, we are
concerned with the transformations of virtual acoustic space. The
transformations are neither simply relatable to existing acoustic spaces nor
do they relate to any conceivable or visualisable events in the real world. As
its title suggests, therefore, we find ourselves travelling in a dream
landscape which has its own logic. The further elucidation of this
‘dreamlike’ virtual landscape is the basis of the piece Red Bird which we
have mentioned previously and will be discussed at length in the next
chapter.
The conventions or idiosyncrasies of media landscapes may become
the basis of compositional structures. Two light-hearted examples of this are
suggested (Examples 7.15 and 7.16). In the first the defects in an old
gramophone record cause the needle to jump about randomly on the disc
surface. This effectively arbitrary editing or looping of short segments
becomes the basis of the organisation of the material in Musical Box (from
my Menagerie). In the second example the music-plays in a typical disc
jockey sequence are replaced by quite different sounds (part of Still Life
from Menagerie).
In Stockhausen's Hymnen (discussed above) we encountered the use
of a formalised media landscape to produce a general impression of
detachment from real world musical objects (the national anthems used in
the piece). In the earlier piece Gesang der Jünglinge (Example 7.17),
landscape considerations enter into our perceptions of the piece (though it is
difficult to know to what extent Stockhausen was aware of, or intended this,
when he originally composed the piece). Thus, the appearance of the boy's
voice in different acoustic perspectives (with different amplitude levels and
with different degrees of reverberation) disembodies the singer. The voice is
floating in a strange and unreal space. At the same time Stockhausen
attempts to set up a kind of serial mediation between the sound of the boy's
singing voice and pure electronic tones with various degrees of vocal
recognisability or electronic abstractness in between. Here is perhaps the
first use of the metaphorical in landscape composition (this will be
discussed in the following chapter). There is a sense in which the boy's pure
voice appears as the utterance of the individual singer, whilst the pure
electronic tones have no such sense of utterance; they are altogether more
distanced in our perception. The mediation between these two sound-types
suggests some kind of mediation between individual human expression and
something much more abstract and distant from human expression, a sonic
metaphor for Stockhausen's continuing religious preoccupation. In case this
may seem far-fetched to a reader schooled in musical formalism, consider
what effect it would have upon our perception of the piece if the
transformation between the vocal and the pure electronic were replaced by,
for example, a transformation between violin sound and trumpet sound
(such as those so effectively executed at Stanford using the new digital
technology). The piece might have been composed with effectively the
same structure but our landscape interpretation of it would have been
entirely different. It is the sense of mediation between the personal
(represented by the recognisable human voice) and the more general and
abstract which carries a metaphorical content beyond the mere use of serial
interpolation between sound-objects.
Finally, in the extract from Berio's Visage (Example 7.18), we hear an
aural landscape in which a clearly-recognisable source, the human voice,
interacts with a more ‘abstract’ set of (electronically-generated) sound-
objects. Just as in typical radiophonic drama work, the human voice
provides a contextualising cue for the other sounds. However, instead of
words being used to outline the action, we are provided with paralinguistic
utterances of fear, pain or suffering. However, it is not merely the vocal
context which makes us perceive the more ‘abstract’ sounds as ‘violent’ or
doing violence to the actress. We could not replace these sounds with
arbitrarily different sounds, even if they were equally loud and produced the
same impact of violent action. Something in the internal characteristics of
the sounds themselves points to a violent origin or gesture. Here, then, we
have a quite different use of the sound-object. In the case of metaphor,
recognisable sounds or their relationships point outside themselves to ideas
and relationships which do not reside essentially in the aural landscape. In
the present case, however, the sonic objects point inside themselves towards
some kind of characteristic of their morphology which we read as indicative
of violence or violent action. Is there a natural morphology of sound-
objects? This question will be taken up in the next chapter.

1 There is a further pair of loudspeakers behind the audience.

2 The complicated relationships between representation, association, visual metaphor, visual gesture
and ‘abstraction’ can be intimated by comparing four paintings of Paul Klee: Polyphony (1932), Gate
in the Garden (1926), The Future Man (1933), Around the Fish (1926) (Hall 1992: (plate 33 (:97);
plate 21 (:73); plate 35 (:101); plate 23 (:77)).
3 ‘The Revenge’ by Andrew Sachs, first broadcast on BBC Radio 3 on the 1st of June 1978.
4 In addition, we need to apply a so-called ‘normalisation’ procedure to ensure the final signal does
not go ‘out of range’ of the digital system.
5 In the original text the author does not discuss the fourth possible combination of his ‘real/unreal’
categories. Presumably unreal-objects/unreal-space is formed when we “begin to replace the animal
and bird songs by arbitrary sonic objects” (unreal objects) and “arbitrarily assign different amplitudes
and degrees of reverberation or filtering to each occurrence” [of them] (Ed.).
6 Raised pitch as a source approaches the listener, lowered as it moves away.
 Chapter 8
SOUND-IMAGE AS METAPHOR: MUSIC AND
MYTH

The true answer is to be found, I think, in the characteristic that myth and music share of
both being languages which, in their different ways, transcend articulate expression, while at
the same time—like articulate speech, but unlike painting—requiring a temporal dimension
in which to unfold.
(Lévi-Strauss 1970: 15)

Music and myth

Having established that landscape considerations enter into our perception

of sonic art and that representational sound-images are potential working
material, what special implications does this have for the sonic artist? In
particular, what forms may we develop based on our sensitivity to sound-
images. The pieces by McNabb and Stockhausen, which we have already
discussed, point towards some new and interesting possibilities. In
Dreamsong the transmutation of aural landscape is suggestive of the
scenarios of dreams, whilst the mediation between the human voice and the
‘pure’, ‘abstract’ electronic sounds in Gesang der Jünglinge points towards
a metaphorical interpretation. Gesang takes this single metaphorical
opposition and embeds it in a complex musical structure. Although it
continues to contribute to our perception of the work, it is not further
elaborated as metaphor. What would happen if we were to establish a whole
system of relationships between sound-images, each having strong
metaphorical implications. By articulating the relationships between the
sound-images we could develop not only sonic structures (as in the
McNabb) but a whole area of metaphorical discourse.
In 1973, having worked for some time on electro-acoustic
compositions which utilised sound-images within an otherwise musical
mainstream conception of sonic structure, I decided to attempt to set up a
sonic architecture based on the relationship between the sound-images
themselves which would however remain compatible with my feelings
about musical structure. What I discovered through working on the piece
Red Bird was that the twin concepts of transformation and gesture,
discussed earlier in relation to non-representational sound-objects, may also
be applied to the sound-image. On the one hand, sound-images of the voice,
or animal and bird cries, have an intrinsic gestural content. More distanced
sound-materials, for example, textures developed out of vocal syllables,
may be gesturally articulated by appropriate studio techniques.
Transformation now becomes the gradual changing of one sound-image
into another with its associated metaphorical implications, and a landscape
can be seen as a particular kind of timbre-field applying to the space of
sound-images (see Figure 8.1). These parallels are not, of course, precise,
but they do form the basis of a meeting ground between musical thinking
and a discourse using sound-images as concrete metaphor.
In his book The Raw and the Cooked, Lévi-Strauss (1970) draws
certain interesting parallels between the structure of music and the structure
of myth. At one level, in adopting a structural approach to the analysis of
myth, Lévi-Strauss calls upon structural categories from the tradition of
Western music (such as ‘sonata form’ and ‘double inverted counterpoint’).
However, he also implies a deeper relationship between music and myth.
Using people, objects and wild animals having a particular significance to
the group, the myth illuminates more abstract relationships and categories
of thought. At the same time the myth gains its power from its unfolding in
time. The way the myth is told is of great importance. The parallels with
conventional musical structures are obvious and in fact Lévi-Strauss points
to Wagner as the first person to attempt a structural analysis of myth. The
fact that he used music as a medium for this is, to Lévi-Strauss, no
coincidence.
 Figure 8.1 Comparison of sound composition with sound-image composition.

Wagner's methodology establishes a relationship between delimited

musical structures (leitmotifs) and people, objects or ideas, primarily
through association—at some stage musical object and its referent are
juxtaposed. By then developing these musical objects and interrelating
them, he is able to carry on a discourse which is not subject to the spatial
and temporal limitations of the opera stage. This discourse is partly to do
with unspoken ‘emotions’ and partly with metaphor. Using sound-images in
the virtual space of the loudspeakers, we can create a world somewhere in
between the concreteness of the opera staging and the world of musical
relationships. We do not need to associate a musical object with, for
example, a bird and thence with a metaphorical meaning, we may use the
sound of a bird directly. And the concreteness of theatrical staging is
replaced by a dreamlike landscape hovering between musical articulation
and ‘real-world’ events (see Figure 8.2). Having drawn these parallels,
however, from this point on there will be little in common between our
perception and conception of sound-image composition and that of
Wagnerian opera!

Sound-image as metaphor
 Figure 8.2 Comparison of music-drama with electroacoustic sound-image composition.

In looking at Stockhausen's Gesang we have already noted that sound-

images may be used metaphorically, although in this particular case this use
may not strike us immediately as other features of the sonic architecture are
more strongly articulated and apparent to our perceptions. It is interesting to
note, however, that even in this case the metaphorical interpretation
depends on the existence of a transformation. It is the mediation between
the sound of a voice and the electronic sound which gives rise to a
metaphorical interpretation which would not arise if no mediation were
established between the two. In a similar way, we might consider using the
aural image ‘bird’ (as in Red Bird) as a metaphor of flight (and hence,
perhaps freedom or imagination). In itself, however, the sound of a bird
need conjure up no such metaphorical association. If, however, we now
make the sonic transformation “lisss”->birdsong (see the previous
chapter1), the voice ‘takes flight’ so to speak; the metaphorical link with the
concept ‘imagination’ is suggested. If this transformation is set within a
whole matrix of related and transforming images the metaphorical
implications become increasingly refined and ramified. Similarly, the sound
of a mechanically repetitive machine has no implicit metaphorical
implications, but a word-machine made out of syllables of the phrase ‘listen
to reason’ and the relationship between the normally spoken sentence and
the word-machine begins to establish a metaphorical dimension in our
perception of the sound-image. We may gradually establish a network of
such relationships, defining a rich metaphorical field of discourse and just
as a change of contextual cues may alter our interpretation of a sound-
image (see the previous chapter), it may also have metaphorical import. The
sound-image ‘bellows/water-pump’ (Example 8.1) may be interpreted as
the functioning of the machine or the functioning of a human body and
when our perception of it changes from one to the other, a metaphor is
implied. (These relatively crude attempts to describe the aural image are not
intended to be taken absolutely literally, as the aural landscape is always
much more ambiguous than the visual). The listener may of course deny or
blank out the metaphorical implications but this is possible with all other art
forms which use metaphor.
In putting together a sonic architecture which uses sound-images as
metaphors, we are faced with a dual problem. We must use sound
transformations and formal structures with both sonic impact and
metaphorical import. We must be both sonically and metaphorically
articulate. Using concrete metaphors (rather than text) we are not ‘telling a
story’ in the usual sense, but unfolding structures and relationships in time
—ideally we should not think of the two aspects of the sound-landscape
(the sonic and the metaphorical) as different things but as complementary
aspects of the unfolding structure. This fusion of conception took place
slowly and unnoticed in my mind during the composition of Red Bird. The
new approach has immediate practical implications. Achieving a
convincing transformation between two sounds is a practical problem of
sonic art. Computer technology, allowing us to examine the internal
structure of sounds, is beginning to make this task easier. The
transformations between the sounds of different instruments playing the
same note (e.g. oboe-flute—listen to examples in IRCAM (1983)) are very
convincing as sonic transformations but unfortunately totally uninteresting
as metaphors. The transformation voice->bird-sounds is metaphorically
quite interesting but much more difficult to generate.2
A second, related problem presents itself. In conventional musical
practice, we may conceive of a sonic architecture which we notate in a
score, partly because the sound-objects we use are universally known and
fixed in the culture. With musique concrète, on the other hand, we work
from a specific sound-object and can thus continually aurally monitor the
sound-structures we develop. If, however, we conceive of an exciting
metaphorical structure involving the transformation of sound-images, we
still have to find or generate specific sound-objects which meet the
necessary recognition criteria and are capable of undergoing the
transformations. The computer, together with a detailed personal knowledge
of the internal architecture of sound, may make it easier for the composer
actually to construct any sound which he can conceive of. At the moment,
however, we are some way from this ideal and in fact Red Bird was made
without the intervention of computers at all. The only available procedure
was to amass a very large number of samples (of birdsong, articulations of
particular syllables etc.), catalogue these according to perceptible (rather
than a priori) criteria and then institute a search through the material when
a specific transformation was to be realised.

Deep structure—surface structure

Lévi-Strauss, in his analysis of myths moves from a description of the

surface structure to an elucidation of the deep structure of the myths
concerned. In this case the deep structures are seen to be various
oppositions and transformations which constitute basic elements of human
thinking. In consciously constructing a myth-structure (an activity which I
imagine Lévi-Strauss would profoundly disagree with), we must somehow
move in the opposite direction. As Lévi-Strauss has pointed out himself,
however, myths are not ultimately about what can easily be said, otherwise
we would just say it. However, in order to explain what I think is going on
in Red Bird it is necessary to give some indication of what I perceive to be
the underlying ‘model’. (Incidentally I do not wish that the listener hear Red
Bird within any framework suggested verbally by myself. Sonic structures
either work or do not work in themselves. If you have not yet heard the
piece in its entirety, either listen to it now or skip the next few paragraphs!).
Some things are expressed much better through metaphor and the dynamics
of sound architecture. Verbal explications may sound unwieldy or
pretentious. With these qualifications, let us continue.
At the deepest level, Red Bird is about the opposition between open
and closed conceptions of the world. This opposition can be defined in
relation to specific and different areas of thought. In relation to our
conception of knowledge (epistemology), a closed view of the world would
regard science as an institution enabling us to proceed from relative
ignorance towards an absolute knowledge of the nature of the world, thus
giving us greater and greater, and finally complete, control over our
environment. The aim is a world in which we may rationally control every
aspect of our existence. An open view would regard science as a powerful
heuristic tool which allows us to gain some control over our environment. It
does not, however, tend towards completeness and its preconceptions need
to be regularly overhauled. Furthermore, there is no conception that there is
a final end to scientific endeavour. The focus is upon the search, the
breaking of new ground and not upon consolidation and security.
In social and political terms a closed view sees human society as a
functional totality in which each person has, or should have, an assigned
role. Society can be entirely rationally organised and ultimately the control
of human affairs is to be left to experts. Those who cannot accept this state
of affairs are either insane, heretical or criminal. An open view regards
human society as rationally orderable only up to a point. There are,
however, always ideas and ways of doing things which no-one has yet
thought of. These do not arise out of rational discourse because this
normally takes place within established frameworks. It is therefore
necessary to allow different and often conflicting ideas and modes of
organisation to co-exist within the same environment because even those
which appear most ridiculous under present circumstances may turn out to
be extremely efficacious. We simply cannot know.
A closed view of language (linguistics) would regard meaning as
residing entirely within the sphere of semantics as defined by the
permutation of various grammatical units. Questions which could not be
clearly formulated within this framework would therefore have no meaning
(the view of linguistic philosophy of a certain type). An open view of
language would regard semantics as merely a vehicle for the approximation
of a person's meaning (see Chapter 2). Linguistic communication may
involve paralanguage or the poetic use of assonance, alliteration or even
song and none of these is arbitrary. Meaning can in fact be conveyed where
the words cannot be found to express it. A closed view of natural
philosophy would see the world as ultimately reducible to a set of ‘natural
laws’. Such concepts as free will and imagination are mere chimera. An
open view would regard the world as well-ordered but not deterministic.
Free will and imagination are real, not merely in the sense that we
experience them to be real but in the absolute sense that their consequences
are not predetermined.
I am not suggesting that Red Bird is about all these things (!), but
about the difficult-to-verbalise opposition ‘open/closed’. People's reactions
to it, however, are usually expressed in a political, linguistic or other frame
of reference. Furthermore, in the long run, Red Bird, like any other myth or
work of art, means what people take it to mean, regardless of what I have to
say on the matter!
Establishing a sound-image structure

In order to build up a complex metaphoric network we have to begin

somewhere. We need to establish a set of metaphoric primitives which the
listener might reasonably be expected to recognise and relate to. Just as in
the structure of a myth, we need to use symbols which are reasonably
unambiguous to a large number of people. The metaphoric primitives
chosen for Red Bird may seem crassly obvious. However, these are only the
basis upon which the metaphoric structure is to be built. The use of
intrinsically esoteric referents (in the manner of T. S. Eliot) would not be
appropriate in this context, apart from any other objections we might raise
to it. The four basic sound-types used in Red Bird are (see Figure 8.3)
Words (especially ‘Listen to reason’), Birds, Animal/Body and Machines.
Although in certain cases these categories are quite clearly distinguishable,
ambiguities may arise (and are used intentionally). For example, non-
linguistic vocal utterances (from a human or animal voice) may approach
linguistic forms and vice versa. A repeated high-frequency glissando may
be taken to be of ornithological or mechanical origin. Articulated mid-range
pitched material may be both bird-like and animal-like. Each symbolic type
is chosen because it either has a conventional symbolic interpretation
(birds: flight, freedom, imagination; machine: factory, industrial society,
mechanism) or such an interpretation can be easily established (the phrase
‘listen to reason’ points to itself).
The situation is already, however, more complicated than this might
suggest. For example, the phrase ‘listen to reason’ is open to fragmentation
and permutation and many different kinds of gestural articulation (see
Figure 8.4) which means that its semantic surface may be utterly
transformed or even negated (listen to Example 8.2). More importantly,
there are two particular kinds of landscape in which the sound-images of
the piece may be placed. These may be described as the ‘garden’ landscape
and the ‘reason’ landscape. Apart from the sound-images involved, the
former is characterised by a sense of coexistence of the sound-images in
acoustic space and time (listen to Example 8.3), the latter by a sense of
spatial and temporal regimentation, the individual images being
subordinated to a rigid, rhythmic structure (listen to Example 8.4). In fact,
the machines which inhabit this landscape are made up either from
phonemes or bodylike visceral sounds, whilst the squeaks and squeals of
the machinery's operation are vocal, animal or bird noises.

Figure 8.3 Basic sound-image classification in Red Bird.

Figure 8.4 Space of possible transformations of the phrase “listen to reason”.

In both cases the landscape as a whole may be transformed. In a

previous example we heard the re-interpretation of the body/machine
sound-image through a change in contextualising cues. In a similar way, the
garden structure can be broken down through the gestural interaction of its
constituents. In Example 8.5 animal-like sound-images emerge from the
background (8.5a), take on a phonemic articulation (8.5b) and gesturally
interact with the bird sounds (8.5c). The garden structure breaks down.
In fact, once we begin to explore the possible metaphorical
ramifications of this network of sound-images, the possibilities become
almost daunting. A methodology is required to allow us to make some kind
of systematic search of the field of possibilities. This is certainly not a case
where we can expect to set up some straightforward permutation procedure
which will allow us to mark off all of the reasonably distinct possibilities,
the field is in fact open. The procedure I shall describe (actually used during
the composition of Red Bird) is meant therefore to be merely a heuristic
tool, an enabling device to force the imagination to consider possibilities
which might not otherwise have occurred to it. It in no sense defines or
delimits the possibilities. Nor does it attempt to separate every conceivable
parameter involved. In fact, definite parameters have been involved in
setting up a search procedure. So long as the procedure is not meant to
define our final use of the material, however, this is of little importance.
The scheme used in Red Bird is illustrated in Figure 8.5. Sound-
images in the piece were broken down into twelve broad categories (no
connection with serialism!). This included sub-divisions of the four basic
categories (e.g. BIRD = birdsong consisting of the repetition of a simple
element; BLACKBIRD = bird-song of more elaborate structure) and a
number of subsidiary sound-images (BUZZ = the fly; BOOK, DOOR,
SHARP = each a category of loud staccato sounds, in most cases made
from phonemes or the slams of books or doors). Every combination of one,
two, three, etc. elements was then considered in terms of the organisational
categories illustrated in the diagram. Then the imagination needed to be
used! Some possible interpretations of these permutations are also
illustrated in the diagram. A brief glance at these interpretations will reveal
that the permutations were not regarded as definitive of the sound
organisations generated. One interesting feature of this approach was that
when I began the piece there was a clear distinction in my mind between
organising something as music and organising something as landscape. In
the actual studio construction of the piece, however, this distinction ceased
to have any significance.
 Figure 8.5 Heuristic scheme for searching the field of possible sound-images for Red Bird.

Aspects of landscape composition

The sound-images used in a landscape may be organised to suggest

different interpretations of the landscape, or even different interpretations of
the sound-images themselves. The garden landscape discussed previously,
organises the sound-images in a ‘naturalistic’ way. Although the
juxtaposition of species is ecologically impossible, our perception is of a
‘natural’, if somewhat dreamlike, environment. Conversely, in the bird-
cadenza (listen to Example 8.6), the material is organised according to
formal criteria derived from musical experience, but because of the sound-
imagery used (various kinds of birdsong), the percept of a natural
environment has not been destroyed.
In the case of the word-machine (Example 8.7) which we have
discussed previously, our attention is not focused upon the phonemic
constituents of the sound-landscape and our interpretation is of a
mechanical, not a verbal, sound-image. We hear phonemes as a machine. In
the next example phonemes are also subjected to a special kind of
organisation. Here, however, the dense and diverging texture is more nearly
suggestive of the flocking of birds. We hear phonemes as if they were
flocking birds (Example 8.8).
In the next example, the phoneme ‘rea’ of ‘reason’ gradually
becomes the sound of an aggressively barking dog, adding a further
dimension to the three already mentioned for our presentation and
interpretation of verbal elements (Example 8.9). In yet another example, the
word ‘reasonable’ immediately becomes the sound of bubbling water. In
this case the final phoneme ‘-ble’ apparently bursts to reveal the bubbling
water (Example 8.10). In another situation (Example 8.11) constituents of
the landscape may be revealed to be other than we had imagined. In this
particular case context has suggested that the sounds we hear are the
squeals and whines of machinery. When the foreground constituents of
machinery are halted, however, we can momentarily glimpse a landscape of
animals, insects and frogs—or is it? Such transformations may be
developed in extended sequences. In the next illustration, the screamed
syllable ‘rea’ is first developed as a texture (Example 8.12a, see Figure 8.6
(I)), then, by a gradual shaping of the envelope and the addition of
reverberation the character of the gestalt is completely altered (Example
8.12b, see Figure 8.6 (II)). Finally, it passes over imperceptibly into the
ticking of a clock (Example 8.12c, see Figure 8.6 (III)).
In the penultimate section of Red Bird we have the most complex
interaction of the many features of landscape composition. The overriding
landscape image is that of the ‘factory’ or ‘torture-chamber’. The two
machines, as we have mentioned, are composed respectively of phonemes
and body-sounds. The image of interminable machinery is articulated
through:
(1) changes in speed of either one or both machines, i.e. changes in the
rate at which the machines cycle (not changes in their pitch) which
may be achieved by shortening individual elements or omitting certain
elements in the machine cycle;

Figure 8.6 The three stages of evolution of screamed “rea-” (from reason) into a clock tick.

(2) interruption of either one machine or other or both by interpolated

material. Much of this is condensed recapitulation of sound-images
and events which have occurred earlier in the piece, for example, bird
sounds which previously transformed into a wind-like texture now
descend even further in a rushing downward glissando, voices which
previously emerged into birdsong now emerge into screams, the book
which attempted to swat the fly and changed into the sound of doors
now pursues the human being.

At the same time animal, bird and human vocal sounds are integrated into
the mechanical as the squeaks and squeals of the machinery. These may
then:

(1) be revealed, as in the case of the animal sounds discussed previously;

or
(2) emerge, as in the case of the human cries which become stronger, less
reverberant and finally detach themselves from the machinery to
complete a full vocal articulation, before being reabsorbed in the
mechanical cycle.

Here we have a classic case of condensed recapitulations of materials for

which we would find the model in conventional musical thought. The
landscape reorganisations and transformations, however, are clearly
metaphorically potent. Musical and metaphorical thinking converge in the
articulation of sound-images in time. The concepts of transformation and
gesture remain fundamental.
One final question remains; can any sound metaphor be said to be
‘universal’ in its interpretation? In his book ‘The Tuning of the World’
Murray Schafer has traced the almost universal symbolism of certain
natural sounds:
The sea has always been one of man's primary symbols in literature, myth and art. It is
symbolic of eternity: its ceaseless presence. It is symbolic of change: the tides; the ebb and
flow of the waves. [...] It is symbolic of reincarnation: water never dies. [...] When angry it
symbolises, in the words of W. H. Auden, ‘that state of barbaric vagueness and disorder out of
which civilisation has emerged and into which, unless saved by the effort of gods and men, it
is always liable to relapse’ [...] When the sea is worked into anger it possesses equal energy
across the entire audible spectrum; it is full-frequencied, white noise. Yet the spectrum always
seems to be changing; for a moment deep vibrations predominate, then high whistling effects,
though neither is ever really absent, and all that changes is their relative intensity. The
impression is one of immense and oppressive power expressed as a continuous flow of
acoustic energy.
(Schafer 1977: 170)

By comparison with the barbaric challenge of the sea, the wind is devious and equivocal.
Without its tactile pressure on the face or body we cannot even tell from what direction it
blows. The wind is therefore not to be trusted. [...] Jung speaks of the wind as the breath of
the spirit. ‘Man's descent to the water is needed in order to evoke the miracle of its coming to
life. But the breath of the spirit rushing over the dark water is uncanny, like everything whose
cause we do not know—since it is not ourselves. It hints at an unseen presence, a numen, to
which neither human expectations nor the machinations of the will have given life. It lives of
itself, and a shudder runs through the man who thought that “spirit” was merely what he
believes, what he makes himself, what is said in books, or what people talk about’. (Schafer
1977: 171)

Schafer goes on, however, to point out that modern man, living in cities,
sheltered from the elements in air-conditioned buildings and travelling
between continents in aeroplanes, tends not to perpetuate this primeval
symbolism. The sea, for example, becomes a romantic image associated
with holidays. How, then, can we use any metaphor with the certainty that it
will be understood? The answer, I think, lies in the embedding of the
metaphor in a structure of interrelationships and transformations, as in Red
Bird, such that various oppositions and distinctions are established. This is
very much the way that musical objects (e.g. motifs) operate. The
significance of the symbolisation is clarified through its relation to other
symbols. Through suitable structures we could establish either the primeval
or the romantic symbolism of the sea or in fact both, and generate subtle
resonances and transformations between the two interpretations.

1 The syllable ‘lisss’ is understood to be from the phrase ‘listen to reason’; in Red Bird this is
established by context.
2 In fact, the programme Chant now makes this transformation quite straightforward and we may
anticipate that problems of sound-transformation will become increasingly transigent as our
experience with computer synthesis and analysis increases.
 Chapter 9
IS THERE A NATURAL MORPHOLOGY OF
SOUNDS?

Modes of continuation: a physical interpretation

While investigating sound-objects from a landscape point of view, it is

interesting to reconsider certain categories from the acousmatic description
of sound-objects given by the Groupe de Recherches Musicales. In
particular the category of continuation refers, as its name suggests, to the
way in which a sound-object may be continued in time. Three basic
categories emerge: the discrete, the iterative and the continuous. Discrete
continuation describes such sounds as a single (unresonant) drum-stroke or
a dry pizzicato on a stringed instrument. Iterative continuation applies to a
single-note ‘trill’ on a xylophone (i.e. the sustainment of a sound which will
be otherwise discrete by rapidly re-attacking it), a drum-roll or the impulse-
stream of ‘vocal fry’ or a bowed note on a double bass string which has
been considerably slackened off. Continuous continuation applies to a
sustained note on a flute, a synthesiser or a bell.
As we are here talking in terms of landscape, we shall relate our
observations largely to sound-events occurring in the natural physical world
(or, of course, to recordings of these heard in the virtual acoustic space of
loudspeakers). We will then apply these observations to our apprehension of
new sound-types generated in the studio through manipulations or
synthesis. In the physical world, therefore, these three types of continuation
imply something about the energy input to the sounding material and also
about the nature of the sounding system itself. We may, in fact, say about
any sound-event that it has an intrinsic and an imposed morphology (see
Figure 9.1). Most sound-objects which we encounter in conventional music
have a stable intrinsic morphology. Once the sound is initiated it settles
extremely rapidly on a fixed pitch, a fixed noise-band or more generally on
a fixed mass in the case of, for example, bell-like or drum-like sounds with
inharmonic partials. Furthermore most physical systems will require a
continual (either continuous or iterative) energy input to continue to
produce the sound. Others, however, (such as bells or metal rods) have
internal resonating properties which cause the sound energy to be emitted
slowly with ever-decreasing amplitude after an initial brief energy input.
Hence, from a landscape point of view we will split apart the category of
continuous continuation to give on the one hand, sounds where there is a
continuous input of energy, the continuation is due to the imposed
morphology (e.g. flute sounds, violin sounds etc.), and, on the other hand,
sounds where the continuation is due to the physical properties of the
sounding medium (continuation through intrinsic morphology).

Figure 9.1 Nine types of imposed morphology.

The reason for making this distinction is simply that the imposed
morphology tells us something about the energy input to the system and
ultimately relates to what we have called the gestural structure of sounds.
Clearly we can gain more information about this energy input where it is
continuous and least where it is in the form of an initiating impulse. Where
energy (mechanical, hydraulic, aerodynamic or electrical) is continuously
applied to the system, we can follow its ongoing subtle fluctuations. The
sounding system is gesturally responsive. Where a sound-event is initiated
by an impulse (e.g. a drum or bell-stroke), however, very little gestural
information can be conveyed—effectively, only a difference in loudness
relating to the force of the impulse. Iterative continuation is ambiguous in
this respect. Iteration may be entirely an aspect of the applied force (as in
the case of the xylophone ‘trill’), purely an aspect of the physical nature of
the medium (vocal fry or slack double bass strings), or an interacting
mixture of the two (a drum-roll).
Clearly, on a synthesiser we can generate events of any kind without
actually supplying any immediate energy input from our own bodies. Two
things need to be said about this. First of all, the mode of continuation (and
attack-structure, articulation etc.) of a sound will tend to be read in terms of
the physical categories I have described. The distinction between, for
example, continuous and impulse-based excitation is not a mere technical
distinction but relates to our entire acoustic experience and ‘tells us
something’ about the sound-object even though it may have been generated
by an electrical procedure set up in an entirely cerebral manner. We can, of
course, transcend these categories of the physical experience of sound-
events, but I would suggest that we do so in the knowledge that this
background exists. In a similar way, for example, we may generate
glossolalia through the statistical analysis of letter frequencies in texts.
Hayes (1983) generated the following examples from analyses of Latin
(Virgil), Italian (Dante) and French (Flaubert) respectively:
AD CON LUM VIN INUS EDIRA INUNUBICIRCUM OMPRO VERIAE TE
IUNTINTEMENEIS MENSAE ALTORUM PRONS FATQUE ANUM ROPET PARED LA
TUSAQUE CEA ERDITEREM [...]

QUALTA'L VOL POETA FU’ OFFERA MAL ME ALE E'L QUELE ME’ E PESTI FOCONT
E'L M'AN STI LA L'ILI PIOI PAURA MOSE ANGO SPER FINCIO D'EL CHI SE CHE CHE
DE’ PARDI MAGION[...]

PONT JOURE DIGNIENC DESTION MIS TROID PUYAIT LAILLE DOUS FEMPRIS ETIN
COMBRUIT MAIT LE SERRES AVAI AULE VOIR ILLA PARD OUR SOUSES LES
NIRAPPENT [...] (Hayes 1983: 19)
But the reader will always hear or read the results against the background of
his knowledge of one or several languages. The forms of sound-objects are
not arbitrary and cannot be arbitrarily interrelated.
Composers who have weighted their activities towards live
electronics rather than studio-based synthesis seem to me to have been
strongly affected by the fact that a morphology imposed upon electronic
sound-objects through the monitoring of performance gesture can be much
more refined and subtle than that resulting from intellectual decisions made
in the studio. The directness of physiological-intellectual gestural behaviour
carries with it ‘unspoken’ knowledge of morphological subtlety which a
more distanced intellectual approach may not be aware of. This is not to say
that theorising cannot lead to interesting results, but that it can lead to a loss
of contact with the realities of the acoustic landscape.
Even where the imposed morphology is a mere impulse, the loudness
of the sound carries information about the force of that impulse. The
association of loudness with power is not a mere cultural convention
although loud sounds have often been used to project political potency
(massed orchestras and choruses, the musicians of the Turkish army etc.).
As far as we know, continuous changing in overall dynamic level
(crescendos, diminuendos) were an invention of the Mannheim school of
symphonic composition in the eighteenth century (though of course it was
possible to differentiate different dynamic levels on instruments such as the
organ in previous ages). The formalistic assignment of a series of different
dynamic levels to musical objects, which was experimented with in the total
serial aesthetic leaves a sense of arbitrariness or agitation (neither of which
is usually intended) because it ignores the landscape basis of our perception
of loudness.
Sounds undergoing continuous excitation can carry a great deal of
information about the exciting source. This is why sounds generated by
continuous physiological human action (such as bowing or blowing) are
more ‘lively’ than sounds emanating, unmediated, from electrical circuits in
synthesisers. The two natural environmental sounds, not of human origin,
which indicate continuous excitation—the sound of the sea and that of the
wind—tend to have an interesting ongoing morphology which may relate to
the symbolic associations of these sounds (see the previous chapter). In the
case of the sea, the excitation (the pull of the moon's gravity) may be
regular but the form of the object (the varying depth of the sea) results in a
somewhat unstably evolving (intrinsic) morphology. The sound ‘of the
wind’ is usually in fact the sound of something else animated by the motion
of the wind. In this case it is the exciting force (the wind itself) which varies
unpredictably in energy giving the sound its interestingly evolving
(imposed) morphology. Murray Schafer has pointed out in his book The
Tuning of the World that it is only in our present technological society that
continuous sounds which are completely stable (the hum of electrical
generators etc.) have come to be commonplace (Schafer 1977, Chapters 5
and 6). The ability of the synthesiser to generate a continual stream of
sounds says something about our society's ability to control energy sources;
but if we take this continuous streaming for granted, like the humming of
machinery, it tends to lose any impact it might have had on the listener. The
machine has no intentions and therefore it inputs no gestures to its sound.
The synthesiser can sound the same way!

Intrinsic morphology of complex sound-objects

If we accept that continuous, iterative and discrete modes of continuation

can be used to describe the imposed morphology of the sound-object we
can suggest a number of other possibilities. In particular, let us propose the
category unsteady continuation. All of these categories, apart from the
discrete, imply some kind of ongoing energy input and we can imagine
transformations between these types of excitation. The various extensions
of the modes of continuation are illustrated in Figure 9.1.
One important critique of the acousmatic analysis of sound-objects is
that it reduces the two dimensions of imposed (gestural) morphology and
intrinsic morphology to a single dimension. Even though the distinction
between these two is not totally clear-cut and in the virtual acoustic space of
loudspeaker the problem of sound-origins can be problematic; I would
argue that the two dimensions continue to enter into our perception of
sound-objects. Different kinds of intrinsic morphology affect us differently
and this is something to do with the assumed physicality of the source
(which is not the same thing as source-recognition). Imposed morphology
we react to more directly, having an immediate relation to the workings of
our own physiological-intellectual processes.
Most musical instruments have a stable intrinsic morphology. When
energy is input in a steady stream or as an impulse, they produce a sound-
object of the attack/resonance type. There is an initial excitation which
generates a distinct spectrum (either pitched, inharmonic, or noise-based)
which then dies away in amplitude either rapidly or with varying degrees of
sustainment. Not all physical objects, however, behave in a similar fashion.
If a steady stream of air is input to a siren it will take time to glissando up to
a steady pitch and even more time to fall from that pitch once the air stream
is stopped, to its lowest possible pitch. Similarly, let us compare the
application of a stream of air whose pressure gradually increases to two
systems. In the case of the flute we find that we set up a given fixed pitch
which becomes louder until we reach a (‘catastrophe’) point at which the
note changes rapidly to the next highest harmonic and so on. With the siren,
however, we find that the pitch slides gradually upwards as the pressure is
increased. Incidentally, as we will discuss in Chapter 12, the human voice is
an interesting case in this respect because, although in music it is usually
used to model the ‘musical instruments’ we have, in moments of extreme
stress it tends to react more like the siren (the scream).
Thus, the flute and the siren have a different intrinsic morphology
revealed by their response to a similar energy input. In this particular case
the sound morphology of the siren is related in a direct (though delayed)
way to the energy input. For this reason the siren is a fairly straightforward
example. If, however, we take sound-objects whose intrinsic morphology is
very complex or unstable how can we relate to these? Are they merely
formless or random? I would like to propose that there are a number of
archetypes which allow us to perceptually classify these complex sounds.
The following discussion is speculative and suggests what a few of these
might be. It relates mainly to sounds all of whose characteristics evolve in a
complex way and does not necessarily apply to typical instrumental tones.
Turbulence: when liquid is caused to flow down a pipe, it normally
does so in a regular and well-ordered manner. When, however, its speed and
pressure reach a certain point, it may begin to exhibit turbulence. The
characteristics of the flow become difficult to predict. This applies to all
fluid flows (water flowing in open channels, flows of air, see Figure 4.8,
Chapter 4). One can imagine a continuous sound-object exhibiting
turbulence. Extreme turbulence would undoubtedly be perceived as noise
but there are probably many other states in between.
Wave-break: a sound of complex morphology which exhibits increase
in internal activity during a crescendo followed by a change in the nature of
the spectrum—particularly a broadening—near the point of maximum
loudness, followed by a diminuendo in which the morphology of the sound
drifts slowly, will be perceived as a wave breaking. The crucial aspect of
this seems to be the spectral broadening near the point of maximal
amplitude, but this is not in itself sufficient to generate this archetype. This
is one kind of natural anacrusial tension and resolution.
Open-Close: if we take any continuous sound whose mass or
statistically-average mass is reasonably constant, impose a low-pass filter
on the sound and then gradually open up the filter so that the high
frequencies are revealed, we produce the aural image of opening or
revealing. With the reverse process, where the filter is closed down again,
we have the aural image of closing, muffling or disguising.
Siren/Wind: in any continuous sound-object where the pitch or mass
rises and falls in parallel with the dynamic level this particular archetype is
suggested. It can in fact be broken down into a large range of sub-categories
and relates to a large number of aspects of utterance in animals and human
beings.
Creak/Crack: when increasing pressure is applied to certain physical
systems (e.g. slats of wood) they emit stress sounds (‘creak’) which are
often intermittent or unsteady and tend to rise in pitch and high frequency
partial content as the pressure increases. The object finally gives way with a
low instantaneous wide-spectrum sound (‘crack’). Sounds having the
characteristic of creaks or creak/crack will relate very strongly to these
concepts of physical tension.
Unstable-settling: if an object which will resonate when struck (such
as a long thin plate of spring steel) is put into a state of complex artificial
tension, then struck and simultaneously released from the state of tension,
the morphology of the sounds it emits will reflect its attempt to reach a state
of equilibrium. In the case of the spring steel there will be rapid spectral and
pitch portamenti in various directions which will finally settle onto the
natural resonance of the unflexed state.
Shatter: some objects, when vibrated by an impulse above a certain
amplitude respond by fracturing. The sound of fracturing is an aggregate of
smaller sounds emerging out of one initial louder and wider spectrum
sound. Typical characteristics of this kind of sound are suggested in Figure
9.2. The particular spectral components of the sound elements will vary
from material to material but the morphological characteristics of shattering
remain approximately the same.
Explosion: a sudden wide-spectrum attack, followed by a wide, low
spectrum ‘aftermath’ with an unstable morphology (rumbling). This is
found in natural and man-made explosions and the sound of thunder. The
evolution of the sound is quite different from the smooth linear decay of the
cymbal or the slowly evolving sonority of the tam-tam.
Bubble: a brief sound whose attack is accompanied by the rapid
opening of a filter and decay by a rapid closing may fit into this archetype.
This description is not complete. It may be necessary for the pitch or mass
of the sound to move even slightly and the nature of the attack is quite
important. Given these limitations, however, the morphology ‘bubble’ can
be imposed on any sound of mass (rather than noise) type. The morphology
of the sound-object presumably relates to the nature of the physical process
of a bubble breaking the surface of a fluid (see Figure 9.3).
This thumbnail sketch is meant to be merely suggestive of what I
have in mind. From working with sounds of dynamic morphology in the
studio, one begins to sense that there is some kind of mental categorisation
of the sound-objects related to the nature of physical processes of change or
instability. I am suggesting, therefore, that a particular sound has a certain
impact, not because it merely reminds one of, for example, breaking glass
or bubbling soup, but that there is a deeper-level morphology of natural
processes which is revealed through the sounds which they make. I sense
that the development of catastrophe theory may throw some light upon this
matter, but at the moment it remains in the sphere of aural intuition.

Figure 9.2 Suggested morphology of the ‘shatter’.

 Figure 9.3 Speculative physical origin of the bubble's upward glissandoing sound.

The sort of approach we are discussing may be extended to verbal

sounds. Although the linguistic sign is essentially arbitrary (i.e. there need
be no intrinsic relationship between the sounds of vocal speech and the
objects referred to) some words in any language do seem to bear some
morphological relationship to the sound-events (or even visual objects)
which they stand for. This effect may be consciously aimed at, as in a poet's
use of onomatopoeia. It may also be that sounds originally mimicking
natural events were absorbed into the language as apparently purely
conventional signs (like most words). The English poet, Gerard Manley
Hopkins, in his diaries, would collect sets of words linked by a consonantal
root:
Grind, gride, gird, grit, groat, grate, greet, [...] Crack, creak, croak, crake, graculus, crackle.
These must be onomatopoeic.
[from Early Diaries (1863) in Hopkins 1959: 5].

The consonantal structure which begins these words suggests a particular

kind of natural morphological sound structure which we find in cracking or
grating type natural sounds. Jean-Paul Curtay, the French lettriste poet, has
suggested that there is a physical analogy between the process of sound
production taking place in the mouth and the process of sound production in
the events represented by words of certain types. These he has termed
mouth symbols.
An image of two surfaces getting in contact with each another becomes obvious when one
gathers: clap, flap, map, slap, snap, strap, tap. [...] what is interesting [...] is that this image
is again consistent with what happens in the mouth when one articulates ‘-ap’, coming from
the larger aperture vowel to a non-voiced bilabial occlusion [...] the same me chanisms can
explain -ump, large rounded volume (bump, clump, dump, hump, lump, mumps, rump, stump)
or ‘-ang’, vibrant resonance in an extended space (bang, clang, bing, ping, ring, gong, song)
symbolically ‘mimed’ with a double voicing in expansion from throat to nose.
The indeniable [sic] existence of these phenomena in English, in French and many
more languages, does not wipe out Saussure's rule of the arbitrariness of the words; it just
modulates it, showing that brain integrative abilities have played a role when they could do
it.
(Curtay 1983)
Natural morphology in group phenomena

It may even be possible to extend this kind of analysis to phenomena where

many individual sound-sources are amassed. Again, I give a brief
speculative sketch of some possibilities.
Alarum: when a colony of animals or birds is disturbed (for example,
by a predator), the resulting mass of individual sounds has a very
characteristic morphology. It will begin with an individual loud cry, which
rises very rapidly in pitch and/or amplitude to a maximum where it remains.
As it does so, many other cries are triggered until a whole mass of
individual cries is created. This then gradually disperses, becoming less
dense and possibly (as in the case of birds) more distant and falling off in
pitch (in this case indicating a fall-off in gestural intensity).
The Dunlin-effect: really a more generalised version of the alarum.
This refers to the particular characteristics of motion of a mass of objects
which I first observed when watching Dunlin flying across mud flats in
Morecambe Bay in England. The Dunlin is a gregarious wading bird and
when the group is disturbed a mass of birds takes to the air and the swarm
can be seen to move along, rise and fall in characteristic swirling patterns
(see Figure 9.4). This type of motion seems to be characterised by three
features. First of all there are birds at the leading edge of the group who set
a course and change direction only in a curving pattern (i.e. there are no
very abrupt changes of direction, although the swirl might be quite tight).
Secondly, the other birds follow this pattern, but not in a mechanical way;
they fly generally in the same direction and alter direction in generally the
same way as the leading birds, yet the trajectory of each bird is quite
unique. Thirdly, the overall pattern of variation of any bird's motion from
the lead birds’ motion is quite random. The flock moves therefore as a
cohesive unit whose overall shape and size may, however, constantly
fluctuate in a manner which is never abrupt but never entirely predictable. It
seems to be the ‘third order’ randomness, combined with the smoothness of
the motion which gives the flight of the flock its mellifluous quality. It is
fairly straightforward to see how this quality might be mapped onto the
patterning of a mass of sound-objects in acoustic space (either pitch,
dynamic and gestural criteria in a mono-space or involving real spatial
criteria in a stereo, quad, or other space).
 Figure 9.4 Flight of Dunlin.

Streaming effects: certain changes occurring in continuous streams of

sounds may perhaps be related to models developed in catastrophe theory.
Some of these changes in regime-type are illustrated in Figure 9.5. One can,
of course, imagine several others, especially where several dimensions of
musical articulation are taken into account.

Some conclusions

If we now return to the two-dimensional model suggested at the beginning

of this chapter (see Figure 9.1), we may modify this to accommodate what
we have learned about the nature of physically existing sound-sources
(apart from the electro-acoustic medium). Generally speaking, all
physically existent sound-sources will be found beneath an arc to the
bottom left of Figure 9.6. Within this arc, sounds initiated by a single
impulse (and not modified by the aid of the electro-acoustic music studio)
will be found in the bottom right wedge. The evolution of these sounds is
given almost exclusively by their intrinsic properties and after the initial
impact the performer can have no effect on the evolution of the sound-
object. At the other extreme, the human voice is to be found in the upper
left-hand wedge of the figure. The larynx and vocal tract are so intrinsically
tied up with human physiology that it is almost impossible to talk about any
kind of intrinsic morphology for vocal sound. There are, of course, sounds
which the voice cannot produce—for example steady-state, inharmonic,
bell-like sounds—but compared with most other sound-sources (including
most other living creatures), the human voice is an almost infinitely flexible
source of sound-objects. Except in extreme cases (extreme distress
generating screaming, for example) the particular formation of the larynx
and vocal tract and hence the morphology of the resultant sound are the
result of the intentional disposition and motion of the vocal physiology. The
morphology of the sound-object is almost exclusively an imposed
morphology.

Figure 9.5 Change of regime in continuous streams.

Typical conventional musical instruments lie somewhere to the

middle of this figure; blown and bowed instruments a little more to the
upper left than plucked instruments. Here there is a balance between
intrinsic morphology and imposed morphology which, in fact, is channelled
in a particular way in conventional musical practice (intrinsic morphology
is streamed—see the discussion of instrumental streaming—so that the
distinction between imposed and intrinsic morphology may more easily be
grasped by the listener). The development of electro-acoustic music and
particularly the development of digital synthesis potentially expands the
field of possibilities (as shown in the figure). We may take a source like (or
akin to) a struck bell and then articulate its ongoing morphology in several
dimensions at once, if we so wish, using studio techniques. Similarly, we
may take the human voice and impose upon it a physically impossible
intrinsic morphology while at the same time retaining its typical
articulation. In a sense, the distinction between imposed and intrinsic
morphology breaks down and to the composer it becomes a matter of
aesthetic or ideological bias whether one regards creating sounds in the
studio as a matter of carefully modelling gesture or building analogues of
physical objects. The age-old debate about music as human expression and
music as objective ‘harmony’ develops a new and intriguing twist.

Figure 9.6 Real sounds in relation to imposed or intrinsic morphology.

The attempts of the Pythagorean school to see human musical

practice in terms of objective laws relating to numerical ratios on the one
hand and the imputation of intent to sound-sources (such as the wind)
having a morphology akin to that imposed by human vocal articulations,
meet in our attempt to generate entirely new sound morphologies and
structures in the studio. In one sense we can say that all the sounds we make
are intended and their morphology imposed by us but, of course, we may
impose a morphology which is modelled on that of a naturally-occurring
sound-object of strong intrinsic morphology! At this level the debate is
meaningless. However, I would argue that at the level of perception this is
very important because archetypes of natural morphology and the
interpretation of human gestural actions will enter into the listener's
interpretation of sound morphology whatever intellectual attitudes the
composer may adopt.
One final and strangely bizarre possibility presents itself. We might
imagine a music whose logic was based entirely upon the logic of the
evolution of natural events as evidenced by the natural morphologies of the
sound-objects used. The sound-objects themselves, however, might be
entirely artificial in origin (and not merely attempting to mimic natural
sound-sources). Is this ‘anecdotal’ music?
 Chapter 10
SPATIAL MOTION

Perceptions of space in Sonic Art

In listening to Western instrumental music either live or over loudspeakers

various conceptions of space enter into our perception of the events.
Musicians have always implicitly accepted some kind of metaphorical or
analogical link between perceived pitch and height. We speak about ‘high’
pitches and ‘low’ pitches and about melodic lines or portamenti going
‘upwards’ or ‘downwards’. As we have seen, it is fairly easy to hold to this
metaphor when one is dealing with a set of sounds of fixed timbre. There
are, however, two components to its perception: one to do with what
Shepard has called the chroma of a note (i.e. whether we perceive it as C, C
sharp, D, E flat etc.) and the other to do with where the energy is
concentrated in the spectrum. As can be demonstrated with the Shepard
tone, it is possible to make the chroma and the spectral ‘height’ move in
opposite directions, yielding the aural paradox of a sound which moves up
and moves down simultaneously. Clearly, then, the association of high pitch
with physical height is not unequivocal. We might ask why things are
perceived this way round rather than the other, i.e. why do we not consider
a glissando moving from 1600 Hz towards 16 Hz (and whose spectral
energy does not change in a contradictory way) as moving ‘upwards’
instead of ‘downwards’? I am not certain if anyone has proposed a solution
to this quandary, but I would suggest that there would be an environmental
metaphor involved. Any creature which wishes to take to the air, with one
or two exceptions, needs to have a small body weight and therefore tends to
have a small sound-producing organ and produce high frequency
vocalisations. Conversely, any large and heavy creature is essentially
confined to the surface of the earth and at the same time will possess a
correspondingly larger sound-producing organ and a consequently deeper
voice. Crudely speaking, airborne creatures have high voices and earth-
bound creatures low voices. Whatever the explanation of this musical
phenomenon, composers of instrumental music have often exploited the
spatial metaphor or analogy. For example, the open textures of some of
Sibelius’ orchestral writing where a low bass line moves against a high
melody with little or nothing in the intervening ‘space’ generates a sense of
a vast and empty ‘landscape’ (Example 10.1).
Stereo reproduction on loudspeakers offers us, as we have discussed
previously, a virtual acoustic space with both width and depth. There are,
however, certain limitations to our perception of this (or any other) space.
We can locate the direction of origin of high-frequency sounds fairly easily
(our brain detects phase differences between the signals arriving at the two
ears and other factors) but low frequency sounds with little energy in higher
partials are very difficult to localise. In some multi-loudspeaker
reproduction systems such as the Gmebaphone, developed by the Groupe
de Musique Expérimentale de Bourges, the lower bass frequencies are
therefore reproduced on a single large bass speaker placed directly in front
of the listener, whilst higher frequency signals are distributed on
loudspeakers placed in the conventional way, symmetrically to the right and
left of the listener. The sense of spatial depth which, as discussed
previously, may be generated in the stereo field by correlations between
falling off in amplitude and high-frequency roll-off (and in certain cases
reverberation) can be further extended in a multi-loudspeaker system by
using planes of loudspeakers at different distances from the audience. The
sound-image may thus be moved literally backwards and forwards by cross-
fading from one set of loudspeakers to another (see Figure 10.1).
In certain circumstances the spatial metaphor associated with
frequency may interact with real spatial motion in stereo space. I remember
particularly a performance of Denis Smalley's Orouboros in which noise-
based sounds which rose and fell in frequency band height in an undulating
manner moved forward and outwards through the stereo space, creating the
impression that the sound was tumbling towards the listener (see Figure
10.2). In our discussion of spatial motion in the current chapter we shall
assume that the apparent location of a sound-object is unequivocal and that
its motion can be adequately described. The ideas developed however
cannot be applied uncritically (i.e. without listening) to arbitrary sound-
material as the internal evolution of the spectrum of the sound-material may
well affect the perceived motion.
A virtual acoustic space may be created all around the listener using
four loudspeakers, a quadraphonic format.1 The generation of an illusion of
depth as in the stereo case allows us to expand the perceived virtual space
out beyond the rectangle defined by the loudspeakers. Another perceptual
problem arises in quadraphonic space. As in the real world it is not always
easy to distinguish whether a sound emanates from a direction forward of
the head or to the rear of the head. The doubt can usually only be resolved
by moving the head itself.
 Figure 10.1 The Gmebaphone (schematic).

Figure 10.2 Tumbling effect produced by spatial and spectral changes.

Although a satisfactory quadraphonic illusion can be created through

an expansion of the techniques we have described for stereo, some more
refined approaches have been suggested. Through a precise analysis of the
exact relationship between the fall-off in amplitude of a sound, its loss of
high-frequency content and changes in its reverberance properties, John
Chowning at Stanford developed a computer program to control more
precisely the illusion of motion in the virtual acoustic space defined by four
loudspeakers (see Chowning 1971).2 The program also models the Doppler
effect: when a sound-source moves towards us its apparent wavelength is
shortened; conversely, when it moves away from us its apparent wavelength
is lengthened. Thus any sound passing the hearer appears to fall in pitch
(from higher as it approaches to lower as it retreats). Ambisonics on the
other hand relies on phase differences generated at the different
loudspeakers to convey information about the spatial location of a sound
object. In some ways it may be regarded as a technical refinement of
quadraphony. Ambisonic technology, however, allows the spatial image to
be rotated, expanded, contracted or (if used in three dimensions) tumbled,
in a technologically quite straightforward way. Recently more refined
computer models of spatial encoding in loudspeaker projection have been
developed.
Finally, virtual acoustic space may be expanded into three dimensions
with as few as four loudspeakers (in a tetrahedral arrangement) or, more
typically, eight forming a quadraphonic rectangle at ground level and a
second above the audience. More ambitious three-dimensional systems
have been presented. The spherical concert hall of the German pavilion at
the Osaka World's Fair in 1970 had a whole network of loudspeakers
distributed around the spherical surface of the auditorium walls, including
beneath the platform on which the performers were suspended. The
auditorium was used for performances of Stockhausen's Spiral3 in which
the sounds produced by the performers were picked up by microphones and
made to move across the surface of the sphere. Such a multi-loudspeaker
system (see Figure 10.3) conveys much more accurate information about
the direction of a sound-object than any technology developed on fewer
loudspeakers can. In recent years the cinema has moved from mono to
stereo to surround reproduction of sound and we can expect the technology
of sound location in space to develop rapidly in the coming years,
particularly with the possibilities offered by computer modelling and
control.

Aesthetic functions of spatial motion

In the previous chapters we have outlined one primary use of spatial
localisation and spatial motion of sound-objects in the definition and
transformation of musical landscape. Certain sounds (e.g. the fly) even need
a spatial motion component in order to be recognisable. Spatial motion may
also be used to underline contrapuntal developments and interactions
between different streams of sound. At the opening of the piece Vox I a
single stream of multiplexed vocal sounds (generated by four voices)
emerges from the tape background and begins to transform and
differentiate. Gradually there is a separation into a higher register stream
(carried by two female voices) and a low register stream (carried by two
male voices) with somewhat different timbral materials. The sense that the
sound-stream has differentiated into two distinct entities (the model of
growth and division of cells was consciously used) is underlined by the
spatial division of the sound-stream as it moves from front centre stage
through the listener and back to two separate locations on the two front
loudspeakers (see Figure 10.4). One sound-object generates two sound-
objects. In a similar way spatial convergence might be used to underline
timbral convergence of two musical streams.
More generally, we may look on spatial movements as (musical)
gestures, consider the typology and implications of different types of spatial
gesture and how the spatial motion of one sound-object might relate to the
spatial motion of others, and thus build up a concept of counterpoint of
spatial gestures. In fact the concepts of transformation and gesture
developed in earlier chapters might be extended so that gestural articulation
in space might be added to our repertoire of possibilities for developing an
articulate contrapuntal music. We could consider using spatial gesture
independently of other musical parameters or in a way which reinforced,
contradicted or complemented other gestural features of the sound-objects.
The gestures of spatial motion (as opposed to the articulation of different
spatial locations) occurs in Stockhausen's works Gruppen and Carré where
sounds are passed from one instrumental or vocal group to another,
sometimes circling around the audience. In the piece Spiral, projected in the
spherical auditorium at Osaka, spatial gestures played an essential part in
the projection of the music. In this case the gestures were entirely
improvised. Here we are going to attempt to analyse spatial gestures in
more detail with a view to understanding the ‘vocabulary’ of motion but not
as yet attempting to define any language.
Figure 10.3 The spherical auditorium at the Osaka World's Fair.
 Figure 10.4 Spatial motion and sound evolution at start of Vox-I

To simplify our discussion we will confine ourselves to an analysis of

the two-dimensional horizontal plane. Motions in the up-down (vertical)
dimension will be referred to in passing. Furthermore we will confine
ourselves to looking at connected paths, i.e. if a sound is to move from
point A to point B it must pass through all locations on some line
connecting these points in sequential order. That line may be as complicated
or convoluted as we wish (Figure 10.5). A sound, however, which appears
at A and reappears at B has not followed any connected path (this is more
like a switching operation than a spatial motion). Clearly we might consider
the latter kind of discontinuous motion as a special category and even
consider an intermediate type of motion which traces out a path from one
point to another but by a series of discrete leaps. There is of course nothing
inherently impossible or unmusical about any of these kinds of motion but
for the moment we shall leave them out of our discussion (Figure 10.6).

Characteristics of horizontal space

Figure 10.5 A variety of connected paths.

We are going to analyse the situation in which the listener is situated at the
centre of a virtual acoustic space so that sound-objects may appear in front,
to the left, to the right and behind the listener (see Figure 10.7). The
listener, in fact, forms a frame of reference for this space which allows us to
talk about ‘in front’, ‘behind’, ‘left’, ‘right’. From a purely geometric point
of view the space is entirely symmetrical and there are no preferred
directions. For the listener, however, certain directions have different
psychological implications to others so that the frame of reference we are
imposing on the space is not just a convention related to the ear-geometry
of the head but a psychological/aesthetic aspect of our perception.
 Figure 10.6 Non-connected and connected paths.

Figure 10.7 Oriented acoustic space around the listener.

The principal distinction to be made is between ‘in front’ and

‘behind’. In purely perceptual terms it can often be difficult to decide
whether a stationary sound is located in front or behind the head. Motion of
the sound (or the head), however, usually allows this distinction to be made.
There is a slight difference in quality between the same sound heard from in
front of the head and from behind the head. The orientation of the pinna and
masking effects of the head itself tends to mean that most sounds are heard
most clearly (in greater detail) when we turn our face towards them. More
important, however, in its natural environment, on hearing a sound—
particularly an unusual or frightening sound—an animal or bird will orient
its face towards the direction from which the sound comes in order to be
able to see the source of the sound. In the case of a sound coming from in
front the creature will have probably seen the source of the sound before the
sound is heard, but this is not the case with sounds coming from behind.
Such sounds, therefore, tend to be more stressful, mysterious or frightening.
This separation of ‘in front’ and ‘behind’ also has a social dimension
for most higher animals. We almost always turn to face the person with
whom we are conversing. Sounds heard from behind may be ‘overheard’ or
‘commands’ but not usually part of a mutual discourse. At a concert or
poetry reading we sit facing the performers. Metaphorically we ‘face up to
things’ or ‘face the music’ and suspect things that happen ‘behind our
back’. The distinction, therefore, between sounds heard from in front and
sounds heard from behind is not merely a function of the geometrical
asymmetry of the human body in the front-back direction (as opposed to the
left-right direction) but the psychological/aesthetic dimension of perception.
The distinction between left and right on the other hand is not so critical.
Not only is the body symmetrical in the left-right direction but (apart from
sounds with no high-frequency components) it is normally very
straightforward for us to differentiate between left and right in locating the
source of a sound and there is no essential qualitative difference between
the same sound heard from a similar angle in front, to the left or to the right.
In certain artificial test situations using specially synchronised tones
played on headphones it can be shown that right-handed individuals have a
tendency to orient their perception such that, for example, high-frequency
sounds are assigned to the right side of the head and low frequency sounds
to the left side of the head, even when such an assignment is contradicted
by the physical placement of the sources. In our normal acoustic
experience, however, we may assume that such effects are marginal,
particularly if, like myself, you are left-handed! From an aesthetic
viewpoint, therefore, a sound heard on the left, or moving from left to right
does not have different implications to a sound heard from the right or
moving from right to left. The distinction will only be of significance where
we have different sounds placed or moving in the same space. Then, for
example, a movement from left to right may be counterpointed with a
movement from right to left. Except in very special circumstances,
however, playing an entire composition with the two loudspeakers switched
around will not alter our aesthetic perception of the piece.
The distinction between ‘above’ and ‘level’ (or ‘below’), on the other
hand, has quite different psychological implications, at least for earthbound
human beings (it might well become different for astronauts). Because we
live on the surface of the planet to which we are bound by gravity, energy is
required for any object to move upwards, whereas objects above us, unless
constrained, will naturally fall down. Sounds moving upwards therefore
will be linked metaphorically to flight or at least with the requirement of
energy input. A sound which moves upwards, slows down and then
descends has in some sense a ‘natural’ motion. A sound which appears from
above and descends may suggest ‘supernatural’ or at least ‘extraterrestrial’
origins as it enters the horizontal plane of our normal acoustic perceptions
from a plane (above) which is normally outside those perceptions. Although
I have overstressed these distinctions in an exaggerated poetic way, these
gravity-related orientation distinctions between ‘above’, ‘level’ and ‘below’
will enter into our aesthetic experience of sounds using the up-down
dimension, even where the symbolic elaboration of these distinctions plays
no part.

Figure 10.8 Grid of nine distinguishable spatial positions.

In order to analyse the qualitatively different types of motion in the

two dimensional plane, I am going to assume a grid of nine distinguishable
positions (see Figure 10.8). It is assumed that the listener is at the position
marked ‘centre’, looking towards the position marked ‘front’. Assuming a
quadraphonic array of loudspeakers placed in the front and rear corners of
the room, it should not necessarily be assumed that the positions ‘front
right’ and ‘front left’ (for example) in our diagram correspond to the
positions of two such loudspeakers. We are discussing positions in the
virtual space created by the loudspeakers and this virtual space may be
much larger than the actual rectangle formed by the loudspeakers
themselves (for example, by using amplitude, high-frequency roll-off and
reverberation to create the illusion of depth beyond the frame of the
loudspeakers). In many cases, we will be able to distinguish many more
directions and distances from the centre position than these nine. This grid
has been chosen for two reasons. First of all in order to make qualitative
distinctions between types of motion we must be able to define the start and
end points of a motion. Two motions will only be distinct (in purely spatial
terms) if they start or end at positions which are qualitatively different to
the listener. This grid gives us the simplest qualitative division of the
acoustic space into perceptually distinguishable and qualitatively different
positions. At the same time our aural discrimination of spatial position is
not so refined as, for example, our discrimination of pitch. Particularly
where the virtual acoustic space is projected on a limited number of
loudspeakers (e.g. stereo or quadraphonic projection). Advances in
computer simulation of spatial position or the general development of
multi-loudspeaker concert halls may soon improve this position. However,
even with just nine positions, we will find that the analysis is quite
complicated enough! For the moment motion between points intermediate
to the grid points may be regarded as segments of larger motions from one
grid point to another (see Figure 10.9).
Considering both the perceptual limitations of the ear and the
mathematics of curves it will be possible to describe all distinguishable
types of motion in terms of straight line or circular motion, or some
combination of these. We will also introduce the idea of random
fluctuations in a motion. Finally, we may distinguish between the motion of
an object and the motion of the frame (of reference). Clearly from a purely
mathematical point of view, the motion of the object and the motion of the
frame are but two representations of the same motion (see Figure 10.10(a)).
Perceptually, however, we can make a distinction between, for example, the
rotation of a single object and the co-ordinated rotation of all the sound-
objects in an acoustic space. The former we will refer to merely as a
rotation and the latter as a frame rotation (see Figure 10.10(b)). Clearly
there will be borderline cases where it is difficult to say whether we
perceive the motion of objects or the motion of the frame, and certain kinds
of complex motions among a group of objects may have the characteristics
of both modes of perception. We will adopt what seems to be the most
perceptually relevant description, indicating areas of ambiguity where these
might arise.
 Figure 10.9 Nine-point spatial grid corresponds to nine spatial areas.

Figure 10.10 Object and frame rotations: distinction of the mathematical and the perceptual.

Finally, two points should be stressed. Our analysis is a qualitative,

not a mathematical, analysis. Mathematically speaking any motion in the
two-dimensional plane can be expressed in terms of two separate motions
along straight lines (and in many other ways). Furthermore the analysis
aims at a qualitative understanding of our perception of motion in acoustic
space and not as a formalisation of compositional procedures.

Direct motions

Given the left-right symmetric, front-back asymmetric nature of the

acoustic space (see above) we can define just three straight-line (non-
diagonal) spatial paths which pass through the listener (see Figure
10.11(a)). In the diagrams dotted line arrows indicate paths which are
aesthetically equivalent to accompanying solid line arrows. However, there
are four such paths along the edge of the space as motion across the front
can be distinguished from motion across the back (by front-back
asymmetry) and forward motion can be distinguished from backward
motion (again by front-back asymmetry) (Figure 10.11(b) and 10.11(c)).

Figure 10.11a Straight line motion: centre-crossing.

Figure 10.11b Straight line motion: edge-hugging.

Figure 10.11c Straight line motion: edge-hugging partial motions.

In the case of paths which pass through the listener's head, it is

perceptually quite clear whether the path is a straight line or not as we can
use front-back and left-right cues. Paths which pass along the edge of the
space, however, are more difficult to judge in this respect as we must rely
purely on distance criteria (which are not so clear-cut). We shall, therefore,
for the moment assume that straight lines and arcs which do not pass
through the listener's head are at least similar in their aesthetic impact.
However, arcs which do pass through the listener's head will be clearly
distinguishable from straight paths and we may distinguish four paths of
this type (see Figure 10.12).
A second set of paths moves simultaneously along the left-right and
front-back axes. We will call these diagonal paths. Backward moving and
forward moving diagonals are clearly distinguishable (front-back
asymmetry) and we may in fact distinguish seven types of diagonal motion
(see Figure 10.13). For these paths which do not pass through the listener's
head the same comments about lines and arcs apply. This means, however,
that we must distinguish a further set of diagonal paths which arc through
the listener's head (see Figure 10.14). In a sense, these ‘centre-hugging’
diagonals may be regarded as spatial articulations of the direct diagonals, a
spatial gestural-articulation imposed on a spatial motion type.

Figure 10.12 Centre-crossing arc motions.

Figure 10.13 Diagonal paths.

Figure 10.14 Centre-hugging diagonal paths.

A further class of movements is concerned with motion to and from

the centre of the space (centring and decentring respectively). These are
illustrated in Figure 10.15. Again, as these motions move to and from the
listener's head, arc-like motions can be distinguished from straight lines and
so we must also consider the set of spatial gestural-articulations of these ten
types where the straight lines are replaced by arcs of various depths. Note
that when the motion is to or from the front centre or rear centre positions
(see Figure 10.16) we need make no aesthetic distinction between arcs
which move out to the left or to the right; but in all other cases the initial (or
final) direction of the arc is of great importance because forward and
backward motion are both quite different from lateral motion (Figure
10.16).

Figure 10.15 Centring and decentring motions.

Figure 10.16 Equivalence and non-equivalence of centring arcs.

Cyclical and oscillatory motions

Next we must consider circular motion. This is the first example of a

motion type which is (potentially) cyclic. The motions we have discussed
up to this point we shall call direct as they trace out a path from one distinct
point to a different point and therefore must take a finite time to execute.
Cyclic motions, however, continually retrace the same path and therefore
may continue ad infinitum. Cyclic motions have different possibilities to
direct motions. In particular, they may be combined with each other or with
direct motions to produce qualitatively different classes of motion (see
below). In this respect they are similar to randomly wandering motions
which, however, differ in that they are not cyclic. In a superficial sense
diagonal motion may be regarded as qualitatively distinct from, yet
derivable from, front-back and lateral motion. Although this is true, I would
not regard diagonal motion as belonging to an altogether different class of
motion from front-back and lateral motions. I cannot in the end give any
hard and fast criteria for these distinctions; they are matters of aesthetic
judgement. We might also argue that diagonal motion is nothing more than
centring followed by decentring (see Figure 10.17). This is a slightly
different matter, however. Whether we observe a motion as diagonal or a
centring followed by a decentring in the same direction depends upon how
we perceptually divide up the spatial motion into distinct spatial-gestural
events; this has partly to do with the sound-material involved but also with
the temporal evolution of the motions involved. This will be discussed
further below.
Because left-right/front-back cues are much more reliable than
distance cues, circular motion is most easily recognised when it passes right
around the head of the listener (central circular motion). Circular motion
which does not do this (peripheral circular motion) is much more difficult to
establish in the listener's perception. The motion may however go right
around the listener's head without being centred upon it (eccentric circular
motion) so that in any event we can define a number of perceptually distinct
circular motions (Figure 10.18). For the moment we will also not make a
distinction between motions along particular closed polygons (see Figure
10.19) and related circular motions. The straight-line paths will either be
distinct because of their time (and sonic) articulation, in which case we can
regard the motion as a set of distinct direct motions, or these clear
articulations will not be made, in which case distinguishing the polygon
from motion along an arc will be quite difficult.

Figure 10.17 Possible derivations of diagonal motion.

 Figure 10.18 Circular motion types.

Figure 10.19 Equivalence of closed polygon and circle.

It is of course possible to take any direct motion and retrace the path
in the opposite direction, thereafter repeating this cycle. I would, however,
prefer to call this motion an oscillation. The motion is partly defined by its
two end points and essentially oscillates between these two positions. There
is no such sense of oscillation in circular motion. As all points along the
circle are equivalent, there is no ‘turning point’. Again, this is not a
mathematical or a semantic distinction but a question of the aesthetic import
of such motions. We might liken circular motion to the motion of a Shepard
tone which, though apparently continually rising, never in fact moves out of
its initial tessitura. An oscillation, on the other hand, is much more like a
trill or vibrato. If we imagine a circular motion in which the diameter of the
circle successively decreases and increases in a cyclic fashion, then the
circular motion would take on the character of an oscillation. These
distinctions begin to blur when we consider eccentric circular motion or
narrow eccentric ellipses (see Figure 10.20).
A related movement type is spiral motion (see Figure 10.21). An
inward spiral which approaches the centre slowly may be perceived as a
circular motion in which the frame is contracting towards the centre (see
Figure 10.22). More commonly, however, spiral motion will be perceived as
direct, as it has a definite start and end point: it is not cyclic. This is
particularly evident in the case of a very shallow spiral (Figure 10.23)
which is more like a mellifluous spatial ornamentation of linear motion. In
between the two extremes, the spiral displays some characteristics of both
circular and direct motion. Like circular motion it tends to negate the
orientation of the space, making all directions equivalent. In its place,
however, and unlike circular motion, it establishes inwards and outwards
motion as significant. Motions which spiral inward and then outward or
vice versa (see Figure 10.24) should also be distinguished. Where this
motion is extended into an oscillation (inwards to outwards to inwards to
outwards etc.) we have the oscillating circular motion discussed previously.
It seems to me unlikely that in two-dimensional acoustic space, spiral
motion which is not centred on the listener's head can effectively convey
the vortex feeling of spiralling.

Figure 10.20 The relation of cyclic to oscillatory motions.

Figure 10.21 Spiral motion.

Figure 10.22 Relation of slow spiral to circular motion.

 Figure 10.23 Shallow spirals.

Finally, let us consider motion along a figure-of-eight. There are two

symmetric figure-of-eight pathways (see Figure 10.25). Motion along these
paths illustrates very well the oriented asymmetry of acoustic space. Non-
cyclic paths may be divided into self-crossing and S-shaped (Figure 10.26).
The important difference between these is that self-crossing paths pass
through the centre of the space twice and S-shaped paths only once. Motion
along an S-shaped path will be differently perceived in one direction than
the other because the sound will pass either front-to-back or back-to-front
through the listener, depending upon the direction of motion (see Figure
10.27). With self-crossing paths, however, there is a difference between
motions which move along the edge (lateral) and motions which move
along the front or back. Lateral paths pass through the centre twice in the
same direction, which will be either twice forwards or twice backwards,
depending upon the direction of motion along the path (Figure 10.27). With
backward self-crossing motion, however, the path crosses the listener's head
in both directions—first front-to-back and then back-to-front—and, thus
reversing the direction of motion, does not alter the aesthetic impact of the
path (only the left-right symmetry). It does, however, make a difference
whether we begin at the back of the space or the front of the space.

Figure 10.24 Inward and outwards spirals.

 Figure 10.25 Two types of figure-of-eight motion.

Figure 10.26 Types of pathway around figure-of-eight.

Figure 10.27 Symmetric and asymmetric motions around figure-of-eight paths.

We can therefore classify the non-cyclic figure-of-eight motions as

shown in Figures 10.28 and 10.29. Cyclic figure-of-eight motions are
affected by the same asymmetry considerations so that motions which move
across the front and rear of the space pass through the centre in both
directions, whereas motions which pass along the sides of the space pass
through the centre either always forwards or always backwards (Figure
10.30). Motions along the diagonal figure-of-eight are more simply
asymmetric (Figure 10.31). Note also that S-shaped paths can be shaded
over into arc-articulated linear paths while self-crossing paths remain quite
distinct (Figure 10.32).
 Figure 10.28 Typology of S-shape figure-of-eight motions.

Figure 10.29 Typology of self-crossing figure-of-eight motions.

Figure 10.30 Symmetries of cyclic figure-of-eight motions.

Figure 10.31 Motions along the diagonal figure-of-eight.

 More extended self-crossing pathways need not be symmetrical.
Figure-of-eight self-crossing motions are characterised by the reversal of
curvature of the path after each crossing of the centre. This results in a
particular patterning of the directions in which the path crosses the central
position. In the case of the backward (or forward) cross, this crossing is
always in the same direction. In the case of the lateral cross, backward and
forward crossings alternate regularly (see Figure 10.33). A quite different
self-crossing motion is illustrated by the clover-leaf, in which the half-loops
of the motion themselves cycle regularly around the space (Figure 10.33).
This motion then carries two senses of ‘circling’ but lacks the ‘twist’ of
figure-of-eight motion. This type of double motion will be discussed more
fully below. Note, however, that there is a regular pattern of centre-
crossings (forward, forward, backward, forward, forward etc.). We may
now introduce irregular reversals and curvature into the motion to produce
the irregular self-crossing motion illustrated in Figure 10.33. Here the
pattern of centre-crossings is irregular (backward, forward, forward,
backward etc.) and although the acoustic space is clearly articulated into
centre-crossing diagonal motions and edge-hugging motions, the overall
effect remains irregular or unpredictable.

Figure 10.32 Transition of S-shape to arc-articulated linear path.

Double motion

Figure 10.33 Regular and irregular paths around figure-of-eight.

We may derive further classes of motion by combining the aesthetic
qualities of motions we have already discussed. The word ‘addition’ used in
this section does not necessarily imply that the motions can be derived by
mathematical addition of the two or more motions from which they
aesthetically derive. As a first example, we may combine a linear oscillation
with the motion perpendicular to the plane of oscillation. This gives us
various types of zig-zag motion illustrated in Figure 10.34. Note that in the
case of the lateral zig-zag, provided there is enough zig-zagging motion, we
need not differentiate between a motion which begins at the rear and one
which begins at the front as the path will spend as much time in front of us
as behind us. Doubled motions, however, depend very much on the relative
time-bases of the two contributory motions (more about this later) and we
will find that there are limiting cases where the quality of the motion is
quite different. Thus in the case of a motion which traverses a single zig-zag
we must differentiate between a motion which begins at the rear and one
which begins at the front (see Figure 10.35). We can also differentiate other
classes of zig-zagging motion (for example Figure 10.36). These are more
easily discussed in terms of the transformation of a one-dimensional frame
(see below).

Figure 10.34 Zig-zag motions.

Figure 10.35 Zig-zag front/back symmetry/asymmetry.

 Figure 10.36 Other possible zig-zag type.

If we now combine a back-front and a lateral zig-zag motion, where

the time-bases of each contributory zig-zag are quite different, we will
produce merely an oscillation between backward and forward zig-zagging
(or leftward and rightward zig-zagging). Where the time-bases are almost
but not exactly equal we will set up an oscillating pattern in the space (see
Figure 10.37). Different patterns of fluctuation between the two senses of
‘diagonality’ may be set up by relative fluctuations in the time-bases of the
two zig-zag motions4.
Combining circular and direct motion we produce looping motion
which proceeds in a particular direction in space whilst continually looping
back on itself (see Figure 10.38). If we now allow the linear motion to
oscillate back and forth across the space the looping motion will do
likewise. In this situation the limiting case is equivalent to the self-crossing
motions discussed earlier. Applying linear motion to an inward-outward
spiral we obtain a pulsating looping motion (see Figure 10.39).

Figure 10.37 Oscillation from two zig-zag motions.

 Figure 10.38 Looping motions.

If we now expand the linear oscillation into a narrow ellipse we may

differentiate between two distinct senses of the ‘addition’ of the qualities of
motion. If we apply the circular motion to every point along the path of the
ellipse we obtain the path illustrated in Figure 10.40(a). If, however, we
apply this circular motion to a particular defining parameter of the ellipse—
in this case the position of one focus of the ellipse—we obtain the
precessing ellipse illustrated in Figure 10.40(b). Both of these motions have
the quality of circling and of elliptic motion but they are combined in a
qualitatively quite different sense. The first I will refer to as internal
addition (of the circle to the ellipse) and the latter as external addition. Note
that we can look on this addition the other way round and regard the first
motion as imposing an elliptic motion on a defining parameter of the circle
(the position of its centre point) and the latter motion as the internal
addition of elliptic motion to all points on the path of the circle.
Similarly we may add zig-zag motion internally or externally to
circular motion (see Figure 10.41). The process of addition of motion in this
sense may be elaborated even further (see, for example, Figure 10.42) but,
at least for the moment, there are distinct limitations on our ability to
perceive the characteristics of such complex types of motion. We may also
add cyclical motion to itself, producing the circling loop motion illustrated
in Figure 10.43. This motion has several special cases, from the rotating
single loop (Figure 10.44) to circular and elliptic four-cloverleaf formations
(see Figure 10.45). We may also describe three-cloverleaf formations and,
even within this category, we can define two distinct types, the normal and
the maximally-swung (see Figure 10.46). The latter is produced by varying
the time-base of the motions appropriately. By a similar process we can
produce two different kinds of four-cloverleaf pathways (see Figure 10.47)
and even motions which are asymmetric with respect to the acoustic space
(for example, the ‘butterfly’ motions illustrated in Figure 10.48). Finally we
may imagine motions which loop closely around the centre, throwing out
larger loops of either regular or irregular sizes at irregular intervals. These
may be regarded as irregularly oscillating circular motions (Figure 10.49)
and lead us into a consideration of randomness in spatial motion.

Images
 Figure 10.39 Pulsating loop motion.

Images
Figure 10.40 Internal and external additions of circle and ellipse motions.

Images
Figure 10.41 Internal and external additions of circle and zig-zag motions.

Images
Figure 10.42 An elaboration of additions of circle and zig-zag motions.

Images
Figure 10.43 Addition of circular motions.

Images
Figure 10.44 A rotating single loop.

Images
Figure 10.45 Four-cloverleaf formations.

Images
Figure 10.46 Three-cloverleaf formations.

Irregular motion

We may also consider motion-types which involve irregular paths through

the acoustic space. Such a path may be completely unlocalised or localised
in a particular area of the space (see Figure 10.50). Alternatively, an entirely
unlocalised (or partly localised) motion may be weighted so that the sound-
object spends more of its time in particular areas of the space than in others
(Figure 10.51). We will leave consideration of such time-averaged
properties for a later section. Clearly irregularity is a matter of degree and
we can imagine a whole array of paths between completely unlocalised
irregular motion and small fluctuations around a direct motion. Movements
of the latter type are best considered as double motions, combinations of
motion types we have already discussed and irregular motion. Clearly,
irregular fluctuations in space can be applied to any direct motion. If,
however, we apply them to double motions, we are able to do this internally
or externally as before. Consider for example zig-zag motion. We may
apply irregular motion internally to the zig-zag paths themselves or
externally to the defining end-points or centres of the zig-zag motion
(Figure 10.52). Clearly the perceived qualities of the three motions are quite
different. In the first case we have an unsteadiness within the zig-zagging
motion itself. In the two other cases, however, the zig-zagging motion is
quite definite but its orientation is unpredictable.

Images
Figure 10.47 Two types of four-cloverleaf formations.

Images
Figure 10.48 ‘Butterfly’ pathways.

Images
Figure 10.49 Irregular oscillating circular motion.

Images
Figure 10.50 Localised and unlocalised irregular paths.

Images
Figure 10.51 Time weighted irregular paths.

Images
Figure 10.52 Addition of irregular and zig-zag motions.

Images
 Figure 10.53 Addition of irregular and circular motions.

Images
Figure 10.54 Addition of irregular and pulsating looped motions.

We may similarly combine irregular motion with circular motion, or

looped circular motion (see Figure 10.53) or to pulsating looped motion
(Figure 10.54). In the case of these circular motions, the external addition
causes the circle centre to wander about in a random manner. With the
pulsating loop motion, however, there is another external parameter, the rate
of oscillation of the pulsation. This, too, may be made to fluctuate
randomly. These variables combine to produce a motion with one degree of
order (looping) and two degrees of disorder (random circle centre and
random pulsation rate (time-base randomness) (Figure 10.55)).
Clearly, when a motion becomes too complex we cease to perceive
any pattern in it at all. Thus internal addition of irregularity will be
perceived as a ‘jitteriness’ of the motion only where it is a small component
of the motion. Beyond a certain point the motion will appear quite random.
External addition of irregularity is in fact more interesting because the zig-
zagginess, or loopiness of the motion is preserved. These qualities should
differentiate one kind of randomly wandering motion from another.

Time

A motion is characterised not only by its path in space but also by its
behaviour in time. We may distinguish the first order time properties
(different speeds of motion) and second order properties (the way in which
the speed changes through time, the acceleration or deceleration of the
motion). We might even consider in some cases third order properties of the
motion (the way in which the acceleration or deceleration changes through
time) but for the moment we will assume that this degree of precision is not
generally audible.

Images
Figure 10.55 Pulsating looped motion with two degrees of randomness.
 The absolute speed of the motion will determine its perceived
aesthetic character. A very slow motion will be experienced as a mere
relocation of position or even as ‘drift’ rather than a movement with some
definiteness or ‘intention’. As the speed of the motion increases the
apparent energy associated with that motion is increased. Motion at
intermediate speeds has a feeling of definiteness or ‘purposefulness’, an
intention to get from one location to another. Fast motions carry a feeling of
urgency or energy. Where fast motion is introduced suddenly into a
relatively static frame, there is a sense of sudden surprise. The similar
introduction of a very slow motion into a static frame may induce a sense of
gentle disorientation. Very fast motion in a circle may even induce a sense
of head-spinning dizziness.
Considering now different categories of speed change we may
broadly differentiate six classes of motion (see Figure 10.56). Accelerating
motions, with their sense of rushing towards a final position, thus increase
in spatial ‘definiteness’ or ‘intention’ and point to the significance of their
target point. They are a kind of spatial ‘anacrusis’. Decelerating motions, on
the other hand, have exactly the opposite effect, a definiteness in leaving
their point of origin and a sense of coming to rest at their target, a calming
or spatial ‘resolution’.
Accelerating-and-decelerating or decelerating-and-accelerating
motions allow us to define some new types of linear motion. Figure 10.57
defines a whole class of there-and-back linear (or narrow elliptical)
motions. Where these have a decelerating-accelerating time-contour they
are perceived as ‘thrown’ elastic motions. It is as if the sound-object is
thrown out from its point of origin on an elastic thread whose tension slows
down its motion and then causes it to accelerate back towards the source.
Simple constant speed motion along any of these paths would usually break
down in our perception into two separate motions, one in the outward and
the other in the inward direction. The time-contour, however, gives the
whole motion a special kind of unity. Conversely, the accelerating-
decelerating time-contour gives the feeling of ‘bounced’ elastic motion, the
motion gathering energy and then being forcibly repelled by the edge of the
space it defines, losing energy as it returns. Again, the overall there-and-
back motion is unified by the time-contour.
Where a motion cyclically accelerates and decelerates, our aesthetic
interpretation may depend on our position in relation to the motion.
Consider the maximally-swung elliptic four-cloverleaf motion of Figure
10.47. We may apply a synchronised pattern of accelerating and
decelerating motion to this path in one of two ways. In the first case the
movement on the elliptic outer loops will accelerate whilst the motion close
to the observer will be slow. As the sound-object will therefore spend most
of its time circling slowly around the observer's head, the motion will
appear rooted in the centre but making dramatic swings out into the distant
space. The motion will thence appear ‘bounced’ elastic. In the opposite
case, however, the motion along the outer ellipses will be slow, accelerating
towards the centre and moving very quickly around the observer's head.
Here the sound-object will spend most of its time on the outer edges of the
space, making sudden (and perhaps disturbing) close loops around the
listener. The motion will then appear ‘bounced’ elastic but in the opposite
direction (inwards) to the first case (see Figure 10.58). We can imagine a
third situation in which the motion around the listener's head is at a medium
pace, suddenly accelerating before it moves off along the outer ellipses
where it decelerates. In this particular case the motion at the centre has a
stable phase (where it is moving at a medium rate) and the listener may thus
feel that the sound is rooted in the centre of the space but ejected to the
edges by ‘thrown’ elastic motion. This example illustrates the way in which
subtle interactions between motion contour and spatial path may influence
the aesthetic impact of a particular spatial motion. As another example,
consider the inward spiral (see Figure 10.21). Where this motion accelerates
towards the centre we have a sense of the sound-object rushing towards, or
being sucked into, the centre of the vortex. Conversely where the motion
decelerates there is a feeling of the sound-object coming to rest at the centre
of motion.

Figure 10.56 Time contours (classes of motion).

 Figure 10.57 Elastic or bounced motions.

Figure 10.58 Bounced elastic four-cloverleaf motions.

In the case of cyclical and oscillatory motions, changes in the motion

contour may synchronise with the rate of oscillation. In this way, a cyclical
or oscillatory motion may be given an entirely new character. For example,
a circular motion may start slowly at the front and accelerate as it moves
towards the back of the space, decelerating as it moves back to the front.
The circle thus no longer defines all directions as equivalent. It becomes
oriented as with most other motions in acoustic space. In the case of double
motions, the relationship between the time-cycles of the contributory
motions will determine the type of spatial pattern traced out by the path of
the sound-object. Aesthetically, however, it is more profitable to analyse the
paths of the resulting patterns and this we have done in a number of cases in
the previous section.
There is, however, another sense in which temporal considerations
enter into our perception of spatial motion. Returning to Figure 10.51 we
may remember that random motion may be weighted in the sense that the
sound-object may spend more of its time in particular areas of the space
than others. The otherwise random motion does have certain time-averaged
characteristics which allow us to distinguish it from other random motions.
There is an interesting parallel here with our perception and analysis of
noise-based signals. Taking a time-average of a white noise signal would
show it to contain all possible frequencies with equal probability. We may,
however, filter the noise such that the occurrence of certain frequencies
becomes more probable, and we then obtain a different sound, a noise with
a distinctive ‘colour’. The weighting of an unlocalised random motion
corresponds exactly to this concept of filtering.
We may go even further. Even more patterned motions such as
circling-looping motion may be time-weighted so that the sound-object
spends more of its time in a particular part of the space (see Figure 10.59).
This is equivalent to stressing a particular ‘formant’ (or quadrant) of the
motion; as the analogy with filtering time-averaged sounds still stands but
we are dealing with a regular pattern. This analogy with filtering procedures
gives us a powerful tool for analysing our perception of complex types of
motion. We might, for example, apply it to discontinuous motions (where a
sound appears and disappears in various locations in the space without
passing through the intervening positions).

Frame motions
 Figure 10.59 Time-weighted circling-looping motion.

In certain situations a group of sound-objects, or a single oscillating sound-

object, may define a line (which need not be straight) in the space. This
may be regarded as a one-dimensional frame and we may investigate
motions of the total frame (as opposed to motions of the individual objects).
In a sense, a frame motion could be seen as merely a set of simultaneous
motions of independent objects. If, however, certain types of symmetry are
preserved between the objects (or in the nature of the oscillation) we will
perceive the group of objects to move as a whole. As well as the fairly
straight-forward motions we will consider here, a frame may be considered
to ‘writhe’ in all sorts of strange ways if the objects defining it move in
elaborate relative motion. There is no clear dividing line between a complex
frame motion and a sense of independent motion of the sound-objects.
We may consider frame translation, swing, twist, flip and rotate.
These are illustrated in Figure 10.60, together with their application to a
one-dimensional linear oscillation. Note that a frame twist is only effective
with an asymmetric one-dimensional sound-image. Otherwise it will be
read merely as a frame translation which contracts towards the centre and
then expands outwards again to beyond the centre (a three-dimensional
twist would be different). The frame may also be contracted or expanded
and these operations may be combined with the previous class of
movements (see Figure 10.61).
We may also consider motions of a two-dimensional frame. The
frame defined by the entire acoustic space may rotate, contract or expand
(see Figure 10.62). By combining these types of motion the frame itself
may spiral inwards or outwards (Figure 10.63). We might also consider
translations of an entire two-dimensional frame (Figure 10.64(a)). The
change in the listener's perspective implied by this motion, however, would
be better suggested by a corresponding expansion and contraction (Figure
10.64(b)); the observer passes through the acoustic landscape (or vice
versa). We might also consider various distortions of the two-dimensional
frame as illustrated in Figure 10.65. We cannot go too far along this road,
however, without the sense of ‘frame’ being lost and the sound-objects
appearing to move independently of one another. There is a sense in which
any mutual movement of the sound-objects in the space which preserves
certain symmetries can be considered a frame motion or distortion, but how
we actually perceive this will depend upon the particular circumstances.

Figure 10.60 One-dimensional frame motions.

 Figure 10.61 Further one-dimensional frame motions.

Figure 10.62 Two-dimensional frame motions.

Figure 10.63 Further two-dimensional frame motions (spiral forms).

Images
Figure 10.64 Further two-dimensional frame motions (translations).

Two types of frame motion are of particular interest: the first (see
Figure 10.66) involves the expansion of sound-objects from the centre into
the surrounding space. If this is accompanied by an accelerated motion it
should give a sense of growth, whereas if accompanied by a decelerating
motion which is initially quite fast, a sense of exploding will be conveyed.
Conversely (see Figure 10.67) all the elements in a space may collapse into
the central position and, if this is achieved with an accelerating motion, a
sense of imploding will be created. In more complex situations we may
imagine most of the objects in the acoustic space undergoing a symmetrical
rotation whilst a single object pursues an independent course. How we
perceptually group the objects in these situations will depend partly on the
various relative motions of the objects and partly on various landscape
aspects of our perception (for example, recognition or sonic relatedness of
the sound-objects).

Images
Figure 10.65 Two-dimensional frame distortions.

Images
Figure 10.66 Frame motion (expansion).

Images
Figure 10.67 Frame motion (contraction).

Some principles

We may draw the following set of conclusions to this part of our discussion:

(1) acoustic space is an oriented space: in particular, front and back are to
be clearly distinguished from one another;
(2) individual motions in the space may be direct or cyclical/oscillatory;
(3) motions may have more than one characteristic;
(4) a degree of irregularity may be imposed internally or externally on any
basic pattern of movement;
(5) the temporal characteristics of a motion will significantly affect its
character: with direct motion (or cyclical motions which have directed
characteristics such as the cloverleaf) the motion contour will
determine the ‘gestural’ feel of the motion, while with cyclical double
and random motions the motion contour will contribute to the spatial
structure of the path;
(6) in certain cases we may consider a one-dimensional or the entire two-
dimensional frame of reference to move.

The counterpoint of spatial motions

Having now established an enormous potential vocabulary of spatial

motion, we may consider how the motion of distinct sound-objects in the
acoustic space may be counterpointed with one another. Direct motions (or
directed aspects of motions, either spatial or temporal) may in this respect
be distinguished from cyclical or oscillatory motions. This distinction, we
will discover, is akin to that between sounds of dynamic morphology and
sounds of stable mass or tessitura in that the former motions may be
organised gesturally whilst the latter may be organised in a sense
‘harmonically’. Any directed aspect of a motion may be considered as a
spatial gesture, These gestures may then be made to move independently, to
interact, or to trigger one another just as with sonic gestures.
For example, we may have three sound-objects in the acoustic space:
sound A wanders slowly around the edges of the space, sounds B and C,
however, dart about rapidly in the space always avoiding each other. In this
situation we would tend to hear the movement of A as a separate and
independent spatial layer, not interacting with B and C. B and C, however,
would form an interactive contrapuntal system because their spatial motions
clearly interact with one another. In Figure 10.68, various motions of the
sound-objects B and C are plotted. These motions might be independent,
interactive, or triggering. In the latter case, for example, the arrival of sound
B at a particular location will suddenly cause sound C to move off (the two
locations in question need not be the same). Just as with sonic gestures,
gestural interaction relies on the relative temporal coordination of the
gestures in time and their intrinsic qualities. We may for example make
gestures which have similar temporal structures but different spatial
qualities, for example, motions which are accelerating in a synchronised
way but which move differently in direction and spatial contour (see Figure
10.69). Similarly, the spatial interrelatedness (or lack of it) between two
gestures may be established in many ways, particularly with reference to
the symmetry of the space (see Figure 10.70). In this way we may establish
a subtle interplay between the relative timing and the spatial characteristics
of various spatial gestures which is akin to the counterpoint of gesture and
transformation. This was discussed in Chapter 6.
Gesture and transformation in space may underline or counterpoint
other (possibly gestural) properties of the sound-objects themselves. Clearly
a motion accelerating towards a point through a crescendo has a quite
different feel from a similar motion accelerating through a diminuendo. In
cases such as the merging and divergence of sonic streams, the coordination
of spatial motion and timbral transformation is obviously of the utmost
importance. Conversely, it is not possible to conceive of a collision between
two sound-objects having identical sonic structures. We would perceive at
worst a mono image as our ears located the sound-object between the two
(hypothetically) moving sources or we would achieve a merge; a sense of
collision could only be created between two objects of quite distinct sonic
properties.

Images
Figure 10.68 Gestural interactions of spatial motions.

Images
Figure 10.69 Synchronised motions with different spatial contour.

Images
Figure 10.70 Multiple motions symmetry considerations.

The symmetries (or lack of them) established between the relative

spatial positions of the sound-objects simultaneously help to define the total
space itself. Once all objects in the space are in motion—particularly if
these motions are asymmetric—a sense of disorientation can be created as
there is nothing left to define the limits or orientation of the space (this is
where closing the eyes becomes important as we can always establish a
visual reference grid). Conversely, if such motions are set against a
background of distant but static sound-objects, a sense of energetic activity
within the frame may be achieved. A cyclic motion on the other hand may
be regarded as a kind of spatial ‘resonance’, mapping out as it does a
particular way of dividing up, and in many cases orienting, the space (see
Figure 10.71). Various motions may then be spatially or temporally
coordinated in order to create various degrees of ‘consonance’ or
‘dissonance’.
The harmonic analogy is not so far fetched as it may at first sound.
Consider for example the motions in Figure 10.72(a). Here two motions
follow the same circular pattern, the same direction. Clearly when the two
objects start from the same point at the same time we hear the rotation of a
single image. Alternatively, if the objects are placed exactly at opposite
sides of the space (Figure 10.72(b)) they are symmetrically oriented with
respect to the head and we hear what amounts to a one-dimensional frame
rotation. These may be regarded as harmonically related states of the
rotation of the system. If we now make the two objects rotate at slightly
different speeds (Figure 10.72(c)) they will continually pass through the
states of parallel rotation and anti-parallel rotation (Figures 10.72(a) and
(b)). We are producing a kind of portamento between two harmonic states
of the system.
We can discover not only spatial ‘harmonics’ but temporal
‘harmonics’ in a rotating system. Let us imagine two objects rotating along
the same circular path but in opposite directions. If they rotate at the same
speed, they will always cross at the same points on the circle (at 0° and
180°) (Figure 10.73(a)). Similarly if one rotates twice as fast as the other
they will always cross at the same four points of the circle (0°, 90°, 180°,
270°) (Figure 10.73(b)). All of these may be regarded as temporal
harmonics of the system. Again, we may make one of the sound-objects
move slightly faster than the other and we may observe the motion pass
through these various harmonic states (Figure 10.73(c)). A temporal
portamento of motion has been created.

Images
Figure 10.71 Spatial division and orientation from cyclic motion.

Images
Figure 10.72 Spatial ‘harmonics’.

Images
 Figure 10.73 Temporal ‘harmonics’.

Images
Figure 10.74 Spatial coordination of two motions.

In Figure 10.74 the two objects move on different paths. The two
paths, however circulate around the space in the same direction (always
anticlockwise). They are therefore in some kind of spatial ‘harmony’ with
one another. If at the same time the cycle times are coordinated so that, for
example, they are both at the centre rear of the acoustic space at the same
time a further temporal ‘harmony’ is achieved between the two motions. In
Figure 10.75 a group of sound-objects rotates around the centre of the
space. If they all preserve the same angular velocity we hear merely a
rotation of a one-dimensional frame around the centre. If, however, they all
have the same linear velocity the outer objects will gradually lag behind the
inner objects. The motions of the various objects are however spatially
‘harmonised’ with each other or at least they set up a particular feeling or
structuring to the space which is more vaguely akin to an inharmonic
resonance. Examples of this type may be multiplied ad libitum.
Furthermore, gestural motions may be superimposed on these situations,
either through the movement of other objects through gestural articulation
of the cyclic motions, or through the consecutive use of gestural and
‘harmonic’ modes of organisation. The counterpoint of spatial motions is
thus in itself an extremely rich field for the sonic artist to explore.

Images
Figure 10.75 Coordination of rotations.

Conclusion

There is clearly even more we can say about spatial motion. We have not
yet considered the up-down dimension; we have not considered oscillating
motions which are so fast as to produce amplitude modulation of the signal
(the timbral effects of spatial motion); we have not considered analogies
with the sphere of dance.
 As the technology is further developed which permits us to analyse
and control the various parameters which enable us to accurately locate
sounds in space and, as reproduction facilities and acoustics are improved,
we can expect this analysis to be expanded and refined; certainly at this
stage it cannot claim to be complete. The organisation of spatial motion is
undoubtedly a growth area in sonic art.

1 Or even just three loudspeakers, in principle.

2 Chowning (1971) deals, strictly speaking, with an illusory space outside of the square defined by
the loudspeakers. Other theories (ambisonics, for example) have tried to overcome some of the
ambiguities which occur within the square (three distinct ways of localising a sound at the centre
point of the square, for example). This does not invalidate the principles of the author's spatial
morphologies, although many of the examples could not strictly be synthesised using Chowning's
algorithms (Ed.).
3 In fact Stockhausen's group performed many programmes of his works throughout their residency
at the Osaka World's Fair. For a description of Stockhausen's spatial ‘mill’ see Cott, 1974: 45–6
(Ed.).
4 Note that the precessing ellipse motion discussed below may be mathematically described in terms
of a front-back and a lateral oscillation with appropriate time bases. It is, however, discussed in a
different context here because it is perceptually related to circular and elliptic motion and not to
linear oscillations.
 Part 3

Utterance
 Chapter 11
UTTERANCE

Man’s languages have objective status as internally organised systems that are independent
of the people who speak them, whereas animal communication is precisely the social
interaction of the animals. Notice that this difference is not just a matter of the cultural rather
than genetical transmission of human languages, for many animal signaling actions,
references, or significances may be culturally acquired either separately or together. [...]
Because languages are, as such, not behaviours, their properties cannot be compared with
the properties of animal communication.
(Bastian 1968: 589)

If [...] verbal language were in any sense an evolutionary replacement of communication by

kinesics and paralanguage, we would expect the old, preponderantly iconic systems to have
undergone conspicuous decay. Clearly they have not. Rather, the kinesics of men have
become richer and more complex, and paralanguage has blossomed side by side with the
evolution of verbal language. Both kinesics and paralanguage have been elaborated into
complex forms of art, music, ballet, poetry and the like, and, even in everyday life, the
intricacies of human kinesic communication, facial expression, vocal intonation far exceed
anything that any other animal is known to produce. The logician’s dream that man should
communicate only by unambiguous digital signals has not come true and is not likely to.
(Bateson 1968: 614–5)

Imagine that we switch on the radio and tune into the sound of an orchestra
playing a familiar piece. The music proceeds for a short while but then we
begin to hear the sound of falling masonry. Performance of the music
becomes disorganised and stops. We hear the sound of chairs being knocked
over and running footsteps and then people screaming. How do we interpret
this sequence of events? The most likely interpretation is as follows: when
we begin to listen we hear the soundstream as music. For many listeners
(though not for all) even the landscape of ‘people playing instruments’ will
be ‘bracketed-out’, their attention will be focused on the syntax of pitch
relations. Once the masonry starts to fall, however, people's attention will
switch very rapidly to the landscape. It will be apparent that some event is
taking place in a location. The landscape of the musical performance can no
longer be bracketed-out. Finally when the screams are heard we perceive an
utterance, that is we assume the sounds that the people are making have
some fairly immediate intent and are not just some new development in
avant-garde musical technique. It may, of course, be objected that the
original musical performance is a type of utterance but it is clear even at
this stage that there is a marked distinction between the two types of sound-
event. In the first (orchestral) case, the utterance is highly formalised, the
sound is patterned according to conventional syntactic rules and we do not
even ascribe the patterning to the performers themselves. They are (up to a
point) merely the agents involved in producing the structure. In the latter
case (screaming) we assume, however, that the individuals involved ‘mean’
their utterances in some immediate sense, i.e. fear, danger. Although we can
make no absolute distinction, for the moment at least we will not describe
the conventionalised case as an utterance. As we proceed with our argument
these distinctions and their ramifications will become clearer.
Now imagine that all of a sudden the pandemonium ceases and an
announcer comes on the air to inform us that we have just been listening to
a new electroacoustic work by a certain young composer. In a sense, with
hindsight, we can now say that the entire experience was a formalised
utterance, a piece of clever tape montage created by a composer, but clearly
it is not as simple as this. Even where we know that we are dealing with a
conventionalised situation, we cannot normally completely obliterate from
our minds the interpretation in terms of (direct) utterance. If we compare
two extreme cases, for example: I tread on my dog's foot and it yelps;
someone says something which he does not mean which is then quoted in a
play by an actor, a recording of which is then used as the basis for a tape
composition which is overheard on the radio by someone who speaks a
different language. Here we feel we can make a clear distinction between
what is a direct utterance and what is a completely different kind of sonic
‘communication’.
 In most normal cases, however, where human beings are heard to
produce sounds, then we will tend to impute intention to the sonic event.
We will hear it at some level as an utterance. In particular, whenever the
human voice is used as a source of sound in whatever context, the concept
of utterance will enter at some level into our apprehension of the event.
This becomes particularly important in the sphere of electro-acoustic music
projected in the virtual space of loudspeakers where we can no longer rely
on the physical and social cues of the concert hall to conventionalise and
sanitise the vocal events. In general, sounds produced by individual
creatures may be taken to indicate or express something about internal state,
reactions to environmental events, responses to utterances by other
creatures and so on, becoming more involved, convoluted and to some
extent detached as we move up the cerebral hierarchy, finally reaching the
etiolated heights of artistic manifestation. At whatever level, the sense of
utterance, whether as indicator, signal, symbol, sign or syntactic or
semantic-syntactic stream, enters into our perception of the events.
The study of utterance will have a bearing on sonic art in two related
ways. First of all, wherever voices enter into sonic structures, we will have
to deal with the special characteristics which pertain to the sonic
architecture of utterance, for example, universal utterance-gestures, para-
language, phonemic objects, language-stream articulation. At the same
time, aspects of utterance may be observed in, or structured into, the
morphology of other sound-objects and events. Just as we can imagine a
landscape containing an utterance, so we can imagine an utterance
containing a landscape (a crude example would be vocoded sea sounds).
Either of these may be aspects of an essentially musical composition.

Animal communication; intrinsic and imposed morphology

In order to consider the utterances of various creatures, we are going to

return to our analysis in terms of intrinsic and imposed morphology of
sound-objects. This will be a useful starting point for the analysis but, in
fact, in the case of utterance, we will find that these distinctions become
difficult to maintain. When dealing with instruments it is intuitively clear
what properties of the sound-object are determined by the intrinsic
properties of the vibrating medium and what properties are determined by
the way in which that is articulated by the human performer. Superficially it
will also seem that we can make a clear distinction between these two on
the grounds that the former is, in a sense, inevitable, while the latter is a
result of conscious choice or intention. Unfortunately, however, matters are
not so simple.
Let us consider first the human voice: the physical structures which
make up the ‘vibrating medium’ are complex and constitute an integrated
system from lungs to larynx to mouth and nasal cavities, tongue, lips etc..
These various components can be set in so many different ways that talk of
an intrinsic morphology of the sound-source is, to say the least, extremely
problematic. Furthermore within the physiology of the organism there is in
no sense a dividing line between the vibrating medium and the musculature
used to articulate this medium. They are one and the same and are
furthermore intrinsically connected to the nervous system and thence to the
brain. In this case, then, to talk of intrinsic morphology becomes
meaningless.
Considering, on the other hand, a much simpler organism, such as the
cricket, the idea of an intrinsic morphology makes more sense.
For many of the lower species, the signal's morphology is a close physical expression of the
mechanical structure of the emission apparatus. These signals have thus a sort of obligatory
physical form, rigidly determined by the elementary movement of the organs.
(Busnel 1968:137)

We may thus, at this level, make an analysis in terms of intrinsic and

imposed morphology. However, now the concept of intention comes into
question. To what extent can we say that the cricket intends the sounds that
it makes. Here we get into interesting philosophical hot water. We can deal
with this problem initially in a somewhat clinical manner, as René-Guy
Busnel writes:
Other signals, on the contrary, have a flexible physical structure due to the possibility of
varied uses of the same organ (such as the bird syrinx, higher vertebrate vocal cords, and the
delphinid larynx) and to a directing brain capable of making choices.
(Busnel 1968: 137)

The higher we go in the animal kingdom, the more diffuse and heterogeneous become the
motor zones, introducing a notion of the degrees of freedom. The production of complex
signals depends upon numerous centers which interfere with each other, and thus no longer
permit the ‘all or none’ responses of invertebrates or lower vertebrates. In mammals, zones
corresponding to a specific signal are not found. Instead, generalised phonation zones can be
 described which are diversely activated by other centers concerned with different emotional
behavior patterns.
(Busnel 1968: 135)

In order therefore to understand the sound emission of the cricket in terms

of intrinsic and imposed morphology we must for the moment sever the
connection between imposition and intention. As we move up the ‘mental
hierarchy’ of the animal kingdom, we will need to consider the different
degrees of voluntariness or involuntariness of utterance, and this will at the
same time deepen our conception of the nature of human utterance.

The biological instrumentarium

The range of sound emission organs found in the animal kingdom is quite varied; they are
usually bilateral in invertebrates and very often unpaired in vertebrates. They may be
restricted to one sex, or they may present a considerable sexual dimorphism. They are found
on all different parts of the body. For example, the following may be found functioning as
sound emission organs in invertebrates: chitinous toothed files which, by friction, stimulate a
vibrating body—wing, elytra, antenna, thorax, leg, abdomen (Orthoptera, Crustaceans);
friction or vibration of nonspecialised organs such as the wings (mosquitoes and some
moths); semi-rigid plates on a resonant cavity stimulated by neuromuscular contractions
(Tymbal method—cicada); reed-like organs which function by aspiration and expiration of
air (death's-head hawk moth, Sphinx atropos). [...] two species have been found [...] which
can automate legs. These species have no special stridulatory organ; however, when the legs
are separated from the body, they emit sounds. When they are intact, they are silent. The
hypothesis is that the noise emitted by the leg attracts predators, leaving the animal free to
flee [...].
In lower vertebrates, [...] nonspecialised organs may produce friction, as do
vomerine teeth in certain fish; osseous, rattle-type apparatuses may be found, made up of
moving, oscillating parts which knock each other when agitated, as in rattlesnakes; whistling
or vibrating apparatuses which function by air expulsion through a more or less
differentiated tube (larynx) ending in an aperture (glottis) with more or less functional lips.
The expelled air is supplied by the lungs themselves or by being in contact with an air pocket
reserve [...] (vocal sac of some amphibians); and finally, membranes may be stretched over
resonating pockets (as is the swim bladder of fish). These apparatuses are activated either by
external percussion (fin beating) or by contraction of muscles disposed in different ways
around the cavity.
Sounds produced by nonspecialised organs are also found in higher vertebrates.
These include breast-beating in the gorilla, organ-clapping, such as wing-beating of the
wood pigeon, drum-rolling in the hazel grouse and gold-collared manakins and trembling of
remiges (primary) and rectrices (tail feathers) in the woodcock and snipe. Owls and storks
use their beaks, and some bats [...] and some insectivores [...] use their tongues. In many
higher vertebrates specialised organs are found, usually working by propelled or aspirated
air in a more or less differentiated tube equipped with modulating membranes or slit systems.
These organs are vocal cords, muscular glottal lips, the larynx [...] and the bird larynx and
syrinx. These apparatuses often have additional organs which form air reservoirs or
resonators (clavicular and cervical air sacs) as found in the bustard, ostrich, crane and
 morse. In some monkeys these features are found in the thyroid cavity, as is the gibbon's vocal
sac or the hyoid bone resonating chamber of the New World howler monkey. Curious
peripheral sound organs are also found, such as the fifteen-spined sound apparatus in
tenrecs, [...] and the tail bell of the Bornean rattle porcupine.
(Busnel 1968:131–132)

Figure 11.1 illustrates two examples of the many forms of stridulatory

apparatus found in insects: on the left the pointed rostrum scratches the
striations on the prosternum. On the right the apparatus is situated on the
antennae; a plectrum (Pl) is rubbed against a row of small tubules, the pars
stridens (PS) situated on the other antenna. Figure 11.2 shows a typical bird
syrinx: on the left it is illustrated from behind (A), from in front (C) and
laterally (B and D). The trachea may be greatly extended as in the Trumpet
Bird (right). The syrinx is surrounded by an air-sack and this, in
combination with the tympanic membranes, makes up the sound-producing
system of the bird. Figure 11.3 shows the human vocal tract (above) and
changes in the vocal cords during speech (below). It should be emphasised
of course that sound is not the only means of communication for animals.
Two important areas which may often be associated with the use of sounds
are the use of gestures of physical movement of the body, face etc. (kinesis)
and, in humans, the use of written language. The temporal patterning of
kinesics may be analysed in a similar fashion to sonic utterances (see Figure
11.4). For the purposes of this book, however, we will not concern
ourselves with either the relationship between written text and the resulting
sound or with the kinesic pathway of communications either in animals or
in Art performance, with one exception: in particular circumstances,
physiological postures and motions will directly influence the morphology
of sounds, thus the characteristic turning-down of the corners of the mouth
which can be observed both in humans and in many primates, visually
indicating a state of ‘sadness’, also influences the formant structure of
vowels and the articulation structure of various consonants. Similarly the
‘deep ululation’ of sobbing (or laughing) influences not only kinesic signals
but also the nature of the sounds emitted. It is the sounds, however, with
which we are concerned.
Figure 11.1 Two examples of stridulatory apparatus in insects.

Figure 11.2 The syrinx of birds.

 Figure 11.3 The human vocal tract and vocal cord changes during speech.

Figure 11.4 Head-bobbing patterns in lizards.

Indicators

Cicadas, crickets and grasshoppers have a very simple sound-producing

mechanism. This may be a specialised stridulatory apparatus (crickets),
specialised convex portions of the abdominal body wall (tymbals) which
are crinkled at rapid rates, or specialised hind legs which are rubbed against
the wings or abdomen. In most of these cases the individual sound-event is
some kind of impulse whose sonic characteristics are determined by the
nature of the biological materials which come into contact. Generally,
however, these impulses appear in short or long groups at a (usually very
rapid) speed. They may also vary in loudness according to their specific
contour (Figure 11.5).
The individual impulse, then, is not open to articulation by the
creature itself; the intrinsic morphology is quite clearly defined. On the
other hand, these impulses may be grouped differently or changed in
intensity (or intensity variation) or speed. This gives a limited number of
imposed morphological characteristics to the signals produced. In our
original schematisation (see Figure 11.6) the sounds of these creatures may
be thought of as occupying a position down to the far right where intrinsic
morphology has a more strongly determining role on the sound-object than
imposed morphology.
However, we run into a new difficulty here. To what extent is the
‘imposed’ morphology imposed? In the case of the cicada, the tymbals are
crinkled by a large pair of muscles to which they are attached. These in turn
are activated by nerve impulses sent out by the cicada's brain. It can be
demonstrated that the sound of a particular species of cicada will vary
according to particular situations, for example, external threat, sexual
arousal, etc.. In this sense the different signals can be taken to indicate the
cicada's state. But does the cicada intend to indicate? Even in human beings
certain acoustic utterances may be taken to indicate the state of the
organism, but are not necessarily voluntary emissions. Consider, for
example, what happens when we hear someone laughing and, against our
conscious will, we find ourselves laughing. Here a signal received has
triggered a response which is outside the control of our conscious intent.
 Figure 11.5 Examples of stridulatory patterns in crickets.

There is, however, an even deeper division between the functions of

acoustic emissions.
 Figure 11.6 Morphological classification of cricket sounds.

No cry leopard frogs emit in the laboratory, however, is comparable in intensity to the scream
heard near midnight on one occasion when a mixed chorus of over a dozen species of frogs
called from a single pond. Microphones were being disconnected when a piercing scream
came from a smaller pond nearby. A beam of light disclosed a raccoon scarcely twenty feet
away carrying a leopard frog in its jaws. The raccoon had seized its prey in shallow water
where numerous other frogs continued to call as though oblivious of its presence.
(Bogert 1960: 204)

In this particular situation the frog's utterance obviously can be taken as an

indicator of its state (terror). However, the indicator does not appear to have
been received by its fellow frogs as a signal.
Clearly in any system of communication at least two features are
involved. We can discuss the sound emissions of one individual in terms of
their being indicators, signals, symbols or signs, only in relation to how
they are perceived by the listening creature. Clearly to us the scream of the
leopard frog is a signal, we recognise its terror. To its fellow frogs, however,
(or so it would seem) the indicator is not a signal. Assuming, however that
we are producing sonic art for human beings, indicators emitted by beings
will tend to be perceived as signals. Thus the sound of a metal rod falling to
the ground is an indicator of the presence of a metal rod. The scream of a
frog, however, not only indicates its presence, but also signals its internal
state to us.
In fact, natural sounds may act as indicators, so that thunder or the
relative pitch of the wind blowing through telegraph wires indicates the
state of the weather. In some cultures humans have imputed intentions to
such sounds and have read the indicators as signals (or even signs)—the
voice of God or the spirits of nature. In the virtual acoustic space of
loudspeakers, we can play with the ambiguity between indicator and signal.
In the recording of a person speaking, for example, we have a nested system
of signalling which may be represented by the schemata: (signal recorded
and presented by recordist (signal emitted by vocalist (utterance))). In
normal listening we bracket out the signalling intention of the recording
engineer and we are left merely with our interpretation of the signalling
behaviour of the recorded voice. In an electro-acoustic work we may play
with the recognition of the sound-objects as utterances by beings or mere
indicators of inanimate events. As a simple example, the anthropomorphic
interpretation of thunder may be hinted at by a formant-like articulation of a
recording of thunder.
Returning to the case of crickets how are we to distinguish an
imposed but involuntary morphology from an intrinsic morphology of the
sound-object? Here we begin to get into deep philosophical water
concerning the nature and degrees of differentiation of voluntariness and
consciousness. It is interesting to note, however, that, even in higher
animals certain emitted sounds are involuntary indicators of an internal
state (sneeze, cough, belch) and some of these may be involuntarily emitted
and received signals (laughter, screaming, both of which may elicit the
same response in a human being without passing through the process of
conscious decision to emit a signal). In most of these cases also our analysis
would be similar to that for the cricket. The ‘intrinsic morphology’ of the
sound-event would be a function of the involuntary conformation of the
organism (resonance of the oesophagus in belching, tense glottis and wide
open vocal cavity in screaming) and the ‘imposed morphology’ would be an
involuntary kind of articulation (such as the deep ululation of breath flow in
laughter).
Of course, for human beings it is possible to utter all these sounds
voluntarily. The contention here is, however, that such basic
indicators/signals always retain some of their primeval communicative
power. Such universal indicators not only transcend our attempts to
formalise them, they even transcend the barriers between species. We may
divide them roughly into three classes:

(1) involuntary physiological indicators;

(2) extremal indicators;
(3) other.
Amongst the involuntary physiological indicators we may cite coughing,
sneezing, vomiting, yawning, biting, chewing, belching, etc. which are
produced not only by humans but by many primates. Furthermore, the
general physiological tone of the creature may be determined from the
nature of breathing. We may distinguish tiredness, ill-health, sexual arousal
etc.. Darwin, in The Expression of the Emotions in Man and Animals
remarks:
As fear causes all the muscles of the body to tremble, the voice naturally becomes tremulous,
and at the same time husky from the dryness of the mouth, owing to the salivary glands
failing to act.
(Darwin 1965: 92)

Tremulous breathing is therefore indicative of fear. Note also that such

indicators may be carried over into the articulation of conventional wind
instruments. Crying and laughing which, as Darwin pointed out, can also be
observed in chimpanzees, orang-utans and baboons are semi-involuntary in
nature.
The main extremal indicator is the scream. This may be characterised
as a continuous, high-frequency, loud, broad-spectrum emission. It usually
indicates a state of extreme terror or pain and may be heard in humans,
chimpanzees, frogs, birds (see Figure 11.7), pigs and even a normally silent
hare when it is torn to pieces by hunting dogs. In birds and mammals, at
least, the similarity of the sound-object ‘scream’ may be related to the
physical properties of a windpipe and throat and a wide open mouth
aperture, an oscillator such as the larynx or syrinx and its activation by a
maximally high energy sustained stream of air.
In fact this indicator is so universal that we may assume that any
sustained high-frequency, loud (and usually broad-spectrum) signal will
carry the connotations of terror. Even in the highly-formalised musical
context of Schoenberg's Erwartung the sustained, high-frequency, loud but
pure-toned pitches which are sung at certain points retain the ‘resonance’ of
screaming (Example 11.1).
There do, however, appear to be trans-special universals. Comparing
the signals of the Great Northern Diver (or Loon), the wolf, the whale and
the red squirrel (Example 11.2) we notice a striking similarity. The signal
initially rises (often a harmonic series step), is sustained and then falls away
in pitch as it ends. It may also include a secondary upward pitch motion
during the pitch-sustained section (see Figure 11.8). These signals should
also be compared with certain human cries. In none of the cases is this
particular utterance the sole utterance of the creature (quite the contrary). Is
there, however, some shared, internal, gestural experience which creates
this particular sustained contour?

Totemics

The analysis of indicators is yet more complex. Clearly, some utterances

indicate the internal state of the being emitting them, regardless of whether
fellow-creatures interpret these as signals or not. Indicators or signals may
have another function, however; they serve to differentiate one species from
another. Thus different members of the same species of cricket produce the
same set of signals but the set of signals is different from the set of signals
produced by a different species of cricket. This may be the case whether or
not the signal is involuntary or (at least semi-) voluntary. Thus, certain
species of birds sing a genetically determined song which characterises the
species and differentiates it from other species. The song is achieved by the
adult bird even where it has no opportunity to hear other birds of the same
species singing the song. In other cases it can be shown that such species-
specific song is learned by the young birds from the adult (and there are
various stages in between the two). Whether or not the creature intends to
emit the sound-signal of its species, that signal serves two important
functions. First of all it can be a territorial indicator and secondly it can
serve to attract a mate of the same species. It is important, however, to
realise that it can fulfil both of these functions without any ‘intention’ to
fulfil either. This I would describe as a kind of involuntary totemic
communication. The sound, in a sense, stands for the species and is
understood to do so, whether or not it is intentionally emitted with that
purpose.
 Figure 11.7 Scream spectrograms of five small birds.

Figure 11.8 Pitch contours of one type of ‘universal’ call.

This sound-totemism can be observed in situations where the

involuntariness of the structure of the signal can no longer be assumed.
Thus, in birds such as the British blackbird which has a repertoire of song-
phrases which it can articulate and sequentially combine in multifarious
ways, the song ‘style’ as a whole still serves as a territorial and mate-
attracting totem (it is wrong to assume that this is a sufficient explanation of
the intent and format of song-production in this case). Even with human
beings, certain musical styles (for example styles of rock-music also
associated with particular types of clothing, hair fashion and even speech;
even the Western classical repertoire in middle-class white society in South
Africa1) can fulfil a basic totemic function, indicating membership in a
particular and exclusive social group and binding one to others of the same
‘species’. At the level of nations, particular ‘song-formulae’ known as
national anthems may serve as sound totemic symbols. In Stockhausen's
work Hymnen an important aspect of our perception is the distancing from,
and commenting upon, this national totemism through techniques of
electronic transformation of the ‘song-formulae’. The use of widely-known
melodies in mediaeval practice (L'Homme Armé) and more recently in
concert music (‘BACH’) and jazz may be viewed from the perspective of
sound totemism whereby religious, intellectual, or merely ‘in’, groups serve
to identify themselves.

Repertoire: gesture and sequencing

As we ascend the cerebral hierarchy of the animal kingdom, our simple

‘cricket’ model rapidly becomes inapplicable. The sound-producing organs
of higher creatures are more complex and are capable of producing a
greater range of sound-objects. There is in fact a repertoire of sounds. At
the same time, and intrinsically tied up with this development, the means
and ability to articulate this repertoire develops—but, as noted before, one
can make no hard and fast distinction between the vibrating media and the
muscles involved in controlling these in the vocal tract of most higher
organisms.
It begins to appear that a repertoire of from about 10 to about 15 basic sound-signal types is
rather characteristic of nonhuman primates as a whole. In some it may prove to be smaller or
larger, but it is doubtful if the limits will be exceeded by very much.
(Marler 1965: 558)
This assertion is, however, misleading (as will be discussed further) as it
depends essentially on where human beings choose to draw lines of
differentiation between different types of signal. Given this qualification,
certain creatures such as birds, dolphins, killer-whales and human beings
display a wonderfully replete repertoire.
The existence of such a repertoire brings other considerations into
play. Apart from the gestural articulation of the individual sound-objects,
we must also now consider the concepts of ambiguity and sequencing.
Whereas with a single sound-source we may talk about a simple iconic
relationship between the state of excitation of the organism and the sound-
emission (i.e. more excited, afraid etc. equals louder or faster), we now
have a multi-dimensional situation. Most sounds of higher animals are of
complex spectrum and dynamic morphology (see previous chapters). There
are therefore a large number of dimensions in which expression of these
may be varied and, in fact, by such variations we may transform one type of
sound-object into another.
For example, in the case of the rhesus monkey (see Figure 11.9), nine
signal types have been described but —
Figure 11.9 Spectrograms of rhesus monkey sounds (after Rowell (1962) and Rowell & Hinde
(1962)).

Rowell (1962) has been able to show that these nine sounds actually constitute one system,
linked by a continuous series of intermediates. Moreover there is one example of multi-
dimensional variation, the pant-threat grading independently into three other calls [...]
Rowell's descriptions also demonstrate correlations with a continuously varying set of social
and environmental situations [...]
(Marler 1965: 561)
Hence, just as with a sound-object of dynamic morphology we are able to
make gestural articulations in different dimensions simultaneously, the
rhesus monkey is able to ‘present’ its internal state in a multi-dimensional
field of utterance. We might imagine a single articulated signal carrying
information simultaneously about, for example, dependency, sexual arousal,
aggression and territoriality, a rich communication medium without the
benefit of the arbitrary sign of a language system. This state of affairs also
reflects the fact that the ‘internal states’ of such organisms have themselves
become multi-dimensional, complex worlds and further underlines a point
made in an earlier chapter about the inadequacy of a discrete verbal
vocabulary of ‘emotional states’ as a means of describing what is going on
within a being.
In addition, however, the repertoire of sound signals may be
sequenced in particular orders and gestural information conveyed through
this sequencing. Most animal studies are heavily concerned with the
development of denotative and arbitrary signs; they are searching for the
roots of human language and there is a tendency to assume that syntax
(rules of sequencing) implies semantics (in the sense of language). As every
musician knows, however, this does not follow. Just as previously we were
able to make a distinction between an indicator and a signal, we must now
note that there is some confusion between a signal and a symbol. Therefore
a certain combination of fear, aggression and sexuality may produce a
particular level of arousal and a particular articulation of the internal state
causing the vocal apparatus to emit sounds of a particular form. As another
organism of the same species will recognise these sound-objects as if it
itself had emitted them, they may be taken to symbolise the particular state
of the first organism. However, we cannot therefore assume that the emitter
intended this symbolisation. Apart from the bringing into action of the
vocal apparatus as a whole, the resulting evolution of the sound-objects
may have been substantially involuntary, a direct utterance.
For the emission to be a symbol to the producer, an act of self-
mimesis is necessary. Mimesis (the imitation of sounds external to the
organism) can be observed in a number of animal and bird species. For
example Indian Hill Mynahs can mimic almost any noise presented to them,
mocking birds incorporate an enormous variety of other bird-sounds into
their repertoire and parrots may be taught to imitate human speech (mimetic
factors in fact enter into natural human languages; see the next chapter). We
may imagine now that a creature may mimic the fear cries of its fellow
creatures and finally that it may mimic its own fear cries, in other words
that it may pretend fear. At this point the signal “fear” becomes a symbol
for fear. This, however, is a difficult point to define. On seeing a predator a
monkey might emit a particular sound which we might take to be a signal of
fear or a symbol of the predator. If the predator is nearer the emission may
be louder. Does this mean, therefore, more-fear or predator-nearer? Where
does the signal end and the symbol begin? We may assume in fact that at
least until the emergence of the arbitrary linguistic sign, there is no absolute
separation. All symbols carry with them an element of physiological-
intellectual signalling to which other creatures respond in a very direct
manner. Once, however, we are able to sequence different signals, we may
convey gestural information in the overall sequence and contour of the
expression. The more this is the case, the less need the individual sonic
gesture units carry immediate signalling information. We generate a
separation (not unknown in music) between the microstructure and the
macrostructure of an utterance. Here, however, nothing is referential in the
sense of the arbitrary sign of language and although the microstructure
gestural units no longer carry such intense and immediate internal gestural
information, there is still a remnant of that original physiological-
intellectual response which allows us to differentiate and respond to them.
We have thus created an articulate gestural syntax which exists on (at least)
two levels.
In such a way we can evolve a multi-level syntax without the
linguistic ‘arbitrary sign’. We can even imagine referring to hypothetical
situations within the context of a real situation, once this operation of
syntactic levels is generated. I am not suggesting here that I know of any
non-human animals which have evolved such a system! Only that such a
system, without language, is conceivable.

Invention; convention

As noted previously, birdsong which serves the function of defining

territory or attracting a mate may arise out of a genetic programme or
through learning from parents or some combination of these two. Certain
birds have a repertoire of possible song phrases which they may articulate
and sequence more or less freely. Given that the song has the totemic
function of representing the species of the bird in question, what more can
we say about it? We have here the first example of a conventionalised
utterance structure which is, however, not language. This is possible
because of the redundancy involved in the system. Thus each phrase of the
blackbird's song might be taken to mean ‘blackbird’ or ‘blackbird territory’
etc. but this gives us no linguistic reason for ordering these utterances. We
might perhaps read from an overall string of phrases ‘self-confident,
ebullient blackbird’ (!) but we do not associate with the sequencing of
phrases the sequencing of immediate gestures (‘happy blackbird’, ‘wakeful
blackbird’, ‘unsure blackbird’, ‘frightened blackbird’ etc.). The individual
units have become distanced from any immediate gestural implication. The
syntax has become conventional.
We may make a similar argument about (at least certain types of)
music. Thus whereas an overall sequence of events may convey a sense of
exaltation, defeat, distancing or whatever we do not therefore associate
every microstructural musical gesture with a particular nameable internal
state. Nevertheless, the microstructural, gestural articulation remains more
closely tied to our visceral/physiological response system than the
detachment of the linguistic sign. Without this remnant of ‘direct
utterance’/‘immediate response’, our response to music at the
microstructural level would be arbitrary and the details of our musical
experience would cease to have any significance.
The fact, therefore, that structured musical utterances are
conventional and, even on a large scale, refer to (at the most) hypothetical
internal states does not mean that they are detached, formalistic exercises in
the arrangement of sound-objects. In particular, a conventionalised
utterance structure (such as music or birdsong) (a) contains the traces of
direct utterance at several levels, (b) can integrate aspects of direct
utterance and (c) can be confused (intentionally, unintentionally or
inevitably) with direct utterance.

Virtuosity, sincerity, acting

The Western classical art music tradition is often noteworthy for its
rejection of the concept of utterance. This may in some respects be traced to
the totemic function of music within society to emphasise group solidarity
in various ways. In situations where the activities of a large group of
musicians are co-ordinated to fulfil a certain predetermined musical end
(for example, fixed ritual observances associated with religious practice),
individual utterance is intentionally negated for the furtherance of a group
utterance manifested through the organisation of the musical materials re-
presented. Here also a second level of conventionalisation arises. Not
merely is a conventionalised structure of musical gesture used, but our
attention is directed away from the personal intent of the performers. In
music of the standard Western repertoire, the conventionalisation of
utterance is many-layered. The composer, conductor and the individual
performers each contribute a level of conventionalised utterance to the
overall sonic experience.
Utterance in the special sense in which we have used it in this chapter
can occasionally come to the forefront of our perception of the musical
experience. The display of virtuosity draws our attention to the technical
expertise or ebullience of the individual performer. The veil of convention
is broken. In the gospel singing of Mahalia Jackson or Aretha Franklin,
conventional musical syntax is gesturally articulated in an extremely
elaborate way which suggests a sense of immediacy (rather than
hypothesis) in the utterance. In contrast a typical Lieder recital normally
presents a sense of distancing; the utterance is clearly of a hypothetical
nature. The singer is not directly involved in the actions or internal states
suggested by texts or musical architecture. In the case of opera singing,
however, a further state is reached. Here (just as an actor adopts a persona
in a play), the singer attempts to present musical material as if it were the
direct utterance of the character represented. In fact the situation is even
more complex, because we of course know that the character represented
would probably not sing about his or her grief, joy, etc. but would more
likely speak about it. We might then schematically represent the situation as
follows: (conventional utterance—opera singer (direct utterance—character
represented (conventional utterance—singing))).
Even without the further level of characterisation we have in opera,
music may be presented in such a way that the utterance aspect is played
down, for example, Xenakis’ Pithoprakta, where the large-scale structure is
dominated by slowly-evolving events, many of which seem gesturally
neutral while the activities of individual players are amassed in dense or
semi-random textures which negate the possibility of individual
articulations emerging through the total structure, or pushed forward, as in
Schoenberg's Erwartung, a monodrama about a frightened woman lost in a
wood sung by a single female singer. The use of the voice in modern
Western popular music presents an interesting case. Whereas in the classical
tradition the singer strives towards the perfection of a particular kind of
voice which is a social convention and is felt to be transferable from one
work or one expressive context to another (liturgical, concert etc.), popular
music projects the idiosyncratic features of the individual singer's voice.2
The audience is assumed to be more interested in music as a personal
utterance rather than as a socially conventionalised utterance. We are
clearly not dealing here with direct utterance in our original sense of the
term. The popular singer adopts many levels of conventional utterance-
structures in order to communicate with an audience. It is usually assumed,
however, (whether or not it is justified) that at some level the singer is not
‘acting’, that the conventionalisation stops and that the singer is presenting
his or her personal utterance. This is pretty obvious in the case of protest
singers, but may be much more indirect. For example, the idiosyncratic
features of the particular voice may be felt to carry the mark of personal
tragedy, grief or difficulty (if in a somewhat distilled format), for example,
Edith Piaf, Judy Garland or Janis Joplin.
Often such personalised utterances will be expressed through widely-
known popular and often totemic song-structures. This is taken to an
extreme in the case of blues, where an almost claustrophobically rigid
structure of music and text is used as a vehicle for sophisticated gestural
expression. This is akin to the highly articulate gestural articulation of
‘stock phrases’ in vernacular speech where the linguistic content can be the
least significant communicative element. The concept of ‘sincerity’ in the
world of popular music can only be understood in relation to these ideas
about utterance.
A more interesting interrelation between conventional and direct
utterance can be observed in ecstatic behaviour. The state of ecstasy
achieved in various religious rites and sometimes in music or dance
improvisation is experienced as a loss of conscious control. In glossolalic
speech, possessed dance, ecstatic gospel music, etc. the performer is able to
articulate the voice or the body to a degree or extent and with a fluency
which is not possible where the conscious mind retains control over the
intellectual-physiological sphere. However, this articulation usually takes
place over a field of conventionally-established possibilities (phonemic
strings, dance movements, musical scales) an intense and immediate
utterance swirls upwards through the conventional structures. How can we
explain this?
When a child begins to walk, it must learn how to do so. It begins
with difficulty to co-ordinate the necessary muscular movements and the
signals about balance and posture received from the ears by the brain.
Eventually, however, all these activities become ‘second nature’ and we are
able to do all sorts of intricate tricks (avoiding objects, hopping over things,
changing our pace to match another person) without consciously thinking
about any of these. Although there is probably a lot of genetic input into our
development of the walking skill, the development of second-nature skills
does not stop when our ontogenetic development ceases. Thus the motions
of the fingers and the fingering patterns required by an experienced concert
pianist are not normally thought about as such in detail during a
performance. They are second-nature. Furthermore, it can be argued that
speech itself (except perhaps among heavily contemplative intellectuals) is
a second-nature activity. High level conscious control is only required at the
most general semantic level. Once this is released, ecstatic glossolalic
speech becomes possible.

Confidence tricksters and psychopaths

One recurring trend of Western art is the movement away from any kind of
direct and ecstatic utterance towards the conventionalisation of all
parameters of the event. The conventionalisation may be an aspect of social
distancing where the conventions are generally understood and accepted as
a medium through which social messages may be transmitted. They may
also, however, be personal conventionalisations of the artist, ways of
distancing himself or herself from the social conventionalisations and even
the implications of direct utterance. Thus the sound poet may plan and
execute a sequence of rhythmic screams or sobs during the performance.
Similarly, a composer like Berio may sit in the studio and coolly edit
together segments of tape carrying verbal gestures which are erotic, funny,
terrifying and so on. Although we know of the artist's detachment from such
utterances, we do not normally distance ourselves entirely from the
utterance-implications of the sounds involved. There is, however, a fine
balance to be preserved between distancing from and involvement in the
utterance whether by the artist or the listener. To quote Gregory Bateson:
I suggest that this separate burgeoning evolution of kinesics and paralanguage alongside the
evolution of verbal language indicates that our iconic communication serves functions totally
different from those of language and, indeed, performs functions which verbal language is
unsuited to perform. [...] There are people—professional actors, confidence tricksters, and
others—who are able to use kinesics and paralinguistic communication with a degree of
voluntary control comparable to that voluntary control which we all think we have over the
use of words. For these people, who can lie with kinesics, the special usefulness of non-verbal
communication is reduced. It is a little more difficult for them to be sincere and still more
difficult for them to be believed to be sincere. They are caught in a process of diminishing
returns such that, when distrusted, they try to improve their skill in simulating paralinguistic
and kinesic sincerity. But this is the very skill which led others to distrust them.
It seems that the discourse of non-verbal communication is precisely concerned
with matters of relationship—love, hate, respect, fear, dependency, etc.—between self and
vis-à- vis or between self and environment and that the nature of human society is such that
falsification of this discourse rapidly becomes pathogenic. From an adaptive point of view, it
is therefore important that this discourse be carried on by techniques which are relatively
unconscious and only imperfectly subject to voluntary control. [...]
If this general view of the matter be correct, it must follow that to translate kinesics
or paralinguistic messages into words is likely to introduce gross falsification due [...]
especially to the fact that all such translation must give to the more or less unconscious and
involuntary iconic message the appearance of conscious intent.
(Bateson 1968: 615)

The point at which the artistic manipulation of materials collapses over into
formalism (in the listener's perception) is very difficult to judge. The type of
artist we are discussing needs to be sufficiently removed from the
immediate utterance implications of his or her materials to explore new
areas of statement or expression. If these implications are ignored, however,
the artistic result is likely to be perceived in some way not intended by the
artist or, worse still, it will be rejected as the artistic equivalent of a
‘confidence trick’.
The problem of detachment has particular significance in Western
society. As an aspect of a professional pursuit, particularly the pursuit of
science, it has proved highly socially fruitful. A detachment from the social
sphere, however, is normally (except in the case of politicians and military
personnel) regarded as a form of mental illness. Mental detachment in
science is useful because it enables us to develop instruments which may
then be useful to the social body. Social detachment in the research which
precedes an artistic work may also be useful in that it enables us to look at
our materials in new ways. Social detachment in the artistic work itself,
however, makes it intrinsically meaningless except as a solipsistic activity
for the artist or an interesting intellectual game for analysts. There is a
certain psychopathology in the scientific method where it is applied to other
beings such as in the pseudo-science of behaviourism and in the pseudo-art
of the notational formalists.

Towards language

Retracing our steps somewhat we have established that a hierarchic

structure of groupings—a syntax—can be established for a stream of
utterance. Birdsong, human music and human language are three examples
of utterance systems using syntax. As explained previously, certain kinds of
birdsong and music can be described exclusively in terms of a hierarchical
structure of gestures from the level of the single event to, for example, the
motif, the phrase, the line, the section, the movement or the work. With
human language, however, another quite separate element enters into our
description: the arbitrary sign. Linguistic signs differ from the symbols we
have discussed previously in that they need not be related in any way (either
causal or mimetic) to the objects, activities, or just syntactic operations,
which they represent.
From the standpoint of this book we are only interested in the sonic
implications of this fact about human language. First of all we should note
that although the linguistic sign need not mimic or otherwise relate to what
it represents, it may do so. Secondly, linguistic signs are only defined up to
a certain point. There are always dimensions of the sound-object which do
not enter into its definition as a sign. These other dimensions may be
articulated, either using independent conventions (the organisation of pitch
in song), interrelated conventions (conventional aspects of paralanguage) or
in other non-linguistic imposed or involuntary ways.
Finally, and most importantly, we now find that the repertoire of the
human voice is divided in a new way. We now have a special class of
sound-objects called phonemes which are used to make meaningful
linguistic utterances. All other sounds, whether voluntary or involuntary,
form a separate category. Strings of phonemes will also have a particular
kind of imposed morphology, including the conventional paralanguage
associated with the particular language (for example the rising tone at the
end of a sentence in English to indicate questioning). Some aspects of this
morphology will be more or less involuntary, such as the vocal transitions
involved in phonemic connectedness. Excluded from this class, however, is
a whole set of alternative imposed morphologies some of which may be
involuntary (such as the contour of a yawn or the physiological
idiosyncrasies of the individual speaker) while others may be intended.
Finally there will be patterns of sequencing of phonemes which will be
syntactically and semantically valid, standing in opposition to all other
types of sequencing. From the sonic artist's point of view, these divisions
may be transgressed in multifarious, multi-dimensional and subtle ways. In
the following chapters I shall try to discuss many of these, but the number
of possibilities and their aesthetic implications are (fortunately) limitless.
Furthermore, all these aspects may enter into our construction and
interpretation of a landscape or any musical structure—in whatever way we
wish to define this term—in which a vocal source is present.
In the following analysis of human utterance we may divide the
subject on the basis of these initial investigations into a consideration of
repertoire (in the sense discussed in this chapter), phonemic objects,
language stream and paralanguage phenomena, and utterance interaction
(duet, antiphony, discourse, chorus, etc.).
Finally, a special word about electro-acoustic music. In the real
physical world we are able to say quite clearly that the sound of a metal bar
falling to the ground is not an utterance whereas a sound produced by a
being is. In the virtual space of loudspeakers this sort of distinction may
become difficult to make. An artificially created sound-object may be
articulated in such a way that we pick up cues of an utterance or not. We
may, in fact, play with the ‘utteranceness’ of a sound-object, just as we may
play with its landscape interpretation. This aspect of the electro-acoustic
medium is another feature contributing to its potentially dreamlike quality,
the creation of an artificial universe in which our conventional
presuppositions are called into question and where we may be brought to
see the world from an entirely different perspective.

1 At least at the time of writing (1983) during the apartheid era [Ed.].
2 See Roland Barthes’ essay ‘The Grain of the Voice’ (Barthes 1977) for a parallel view (Ed.).
 Chapter 12
THE HUMAN REPERTOIRE

Voice is the original instrument.

(Joan La Barbara 1976)

We are now at the point where we can describe the human repertoire,
the fund of possible sonic objects and their articulations which is available
to the human utterer. The following taxonomy is based on an earlier
analysis of mine published in Book of Lost Voices (Wishart 1979). Recent
discussions with the lettriste poet Jean-Paul Curtay have clarified a number
of physiological and other distinctions enabling me to present a more
systematic classification of the sound-materials discussed in that
publication. Curtay has presented a physiological analysis of vocal
technique (Curtay 1981) whereas my own approach is oriented to a
description in terms of sound-objects. This leads to two different notational
approaches to the sounds (both of which will be discussed below).
Furthermore, the analysis here will be confined to sounds related to the
vocal tract. Body slaps, hand rubbing and other sounds may of course be
produced by human beings (and these are discussed in Curtay (1981)) and
no aesthetic preference is implied by their omission here. Furthermore, I
will not claim that this is a complete analysis as I have been discovering
new sounds almost every week since the completion of Book of Lost Voices.
This listing does include several sounds not mentioned in the earlier
publication.
The analysis does treat sounds of intrinsically short-duration
separately from sounds which can be sustained. Of course any of the
sustained sounds can be made into a short sound merely by curtailing its
duration. Furthermore I have not made the distinction between ‘iterations’
and ‘vibrations’ simply because any iterated sound becomes a vibration if it
is sufficiently speeded up. Some sounds (e.g. rolled-r, lip-flabber) appear at
first sight to be intrinsically iterative as we normally produce them in a
range where we can hear the individual pulsations; however, as will be
demonstrated, these sounds can be made to rise into the normal audio
vibration range (the first by increasing tongue pressure on the roof of the
mouth, the second by hand tensioning the lips) and even at their normal rate
of iteration a pitch can be perceived (particularly if it is stabilised). Slow
iteration may be looked on as sub-audio vibration, for example, the glottis
(vocal chords) can be made to vibrate in a sub-audio mode and even to emit
individual impulses (particularly when activated on inhaling).

Oscillators and other sources

For the moment we will assume that the flow of air is outwards from the
lungs (exhaled). The principal sound-sources in the vocal tract are physical
oscillators in the sense that they produce sound by physically moving to and
fro (just like the reed in a reed instrument). Certain other parts of the vocal
tract can be made to resonate and hence produce pitched sounds.

(1) The glottis is the source of the normal human singing voice. Sounds are
produced by oscillations of the vocal cords, and we will refer to these
sounds either as glottal vibrations, or vibrations of the larynx. Vibrating
the larynx produces an impulse which is normally iterated (Example
12.1). At normal rates of vibration this is heard as a pitched sound
(Example 12.2). At least two separate registers can be perceived within
the range of pitch produced by the larynx. For the male voice these are
usually referred to as normal voice and falsetto voice. A completely
seamless transition can be made between the two registers. It is also
suggested that there is a further break in the voice permitting an even
higher range of pitches to be obtained. My own (male) voice will at the
 moment reach to the G two octaves and a fourth below middle C (the
lowest part of the range is very relaxed and quiet) and as far as the G
one octave and a fifth above middle C (in falsetto).
(2) The windpipe. If air is expelled very forcibly from the lungs a low pitch
is produced which does not originate in the larynx but somewhere
below it (Example 12.3). The sub-glottal vibration may be stabilised on
a definite pitch. This pitch, as far as I can tell, cannot be altered and
may be combined with glottal pitches. Note that this sound is quite
different from sub-harmonics (see below) and is the basis of the famous
‘Satchmo’ gravel-voice. A similar effect may be observed in the
windpipe above the larynx (Example 12.4). Although I have no direct
medical evidence that these sounds are produced where I suggest, they
are quite distinct from glottal vibrations because it is possible to
combine them with glottal vibrations. Examples 12.5–12.8 illustrate the
sound below the larynx, the same combined with the sound of the
larynx, the sound above the larynx and the same sound combined with
the sound of the larynx.
(3) Subharmonics. If glottal production is made very relaxed, it is possible
to instigate a note one octave below the original glottal note and
sounding simultaneously (Example 12.9). This note varies in pitch
along with the original glottal note, always remaining at the interval of
one octave and is produced either by the larynx or by the false vocal
folds resonating (half) in step with the larynx. With practice, a note one
octave and a fifth, or even two octaves, below the glottal note can be
produced.
(4) The oesophagus is resonated during belching. The sound is used as a
basis for speech by people who have had their larynx removed for
medical reasons.
(5) The tongue may be vibrated against the roof of the mouth. This may be
done using the tip of the tongue towards the front of the mouth
(Example 12.10), upwards onto the soft palate (Example 12.11) or
strongly retroflexed (Example 12.12). Alternatively the tongue may be
arched so that the middle of the tongue is in contact with the roof of the
mouth, producing the characteristic French ‘r’, either in the middle of
the mouth (Example 12.13) or towards the back of the mouth (Example
12.14) and finally against the uvula, the sound associated with snoring
(Example 12.15). By suitable use of tongue pressure and placement
 these sounds can be brought into the range of normal sung tones,
particularly the arched tongue type (Example 12.16).
(6) The lips may be made to vibrate either in normal position (Example
12.17) or strongly folded inwards towards the teeth (Example 12.18) or
strongly pouted outwards (Example 12.19). Pitch formation with the
lips may be assisted by using the hands to stretch or relax the lips. This
also helps to stabilise the vibrations so that lip notes can be sustained
for long periods (Example 12.20). The available aperture and tension of
the lips can also be manually controlled to produce a variety of
different kinds of oscillations (Example 12.21) and in particular two
independent sets of vibrations can be set in motion at different corners
of the mouth simultaneously (Example 12.22).
(7) The cheeks may be vibrated independently of the lips and the pitch
controlled by varying tension by use of the hands (Example 12.23).
These vibrations may be sub-audio (Example 12.24). The two cheeks
may also be set in vibration independently (Example 12.25) with
possibilities such as producing patterns of beats between two sub-audio
frequencies.

Filters

Sounds produced within the vocal tract have not only a fundamental
frequency but also formants, frequency areas within the spectrum where
energy is concentrated. Formants are generated by various resonances
within the oral and nasal cavities and it is quite possible to vary these in a
continuous fashion. This may be done in four ways: by varying the size of
the oral cavity, by varying the position of the tongue's arch, by greater or
lesser rounding of the lips and by greater or lesser nasalisation (i.e. varying
the amount of sound which is passed by the nasal passages). The latter will
be discussed separately as it is of less importance and applies only to glottal
and windpipe sounds. (Vowels, Example 12.26.)
Sounds produced prior to the oral cavity (glottal, oesophageal and
windpipe sounds) may have their formant structure altered by any of these
four methods. The oral (nasal) cavity may thus be regarded as a complex
filtering device. As an initial approximation we will omit lip-rounding from
our analysis. Fortunately there is a notation immediately to hand for
specific formant types as the vowels of human languages are determined by
their formant colour. As a first approximation we may draw a two-
dimensional map (see Figure 12.1) of the ‘formant space’ available. Note
that we may move from the open a sound of English ‘father’ to the small
aperture vowel with the tongue arched against the front of the mouth
(German ü) by either first closing down the mouth aperture to reach the
vowel u as in North of England ‘mud’ and then moving the tongue arch
position forward, or we may begin by moving the tongue arch forward,
passing through the vowel e of English ‘red’ and arriving at the vowel e of
‘she’ and then closing down the mouth aperture. Between these two
extremes there are any number of roots from the a to the ü through the
vowel space (Example 12.27).

Figure 12.1 Map of the ‘formant-space’ of vowels.

In this analysis in fact we have compounded lip-rounding with

aperture effects. The effect of lip rounding on various formant types can be
heard in Example 12.28. A further dimension is also added by moving the
tip of the tongue into the r or l position, producing a clear changing
colouration of the formant spectrum. Finally by using the tongue arch to
stop off the flow of air through the oral cavity (the ng position) the sound is
projected entirely through the nasal cavity; varying the degree of contact of
the arch with the roof of the mouth will vary the proportion of the sound
which is projected through the oral cavity.
The international phonetic alphabet provides a concise and fairly
detailed means of notating particular formant structures (Figure 12.2). It
also has the advantage over a physiological notation (see later) that we may
establish links with language utterances quite easily. However, the notation
does need to be modified for use in sonic art. For example, if we wish to
produce a formant which has a mouth formation for e but in which the
tongue-tip is half moved towards the l position, or a sound with the mouth-
shape of i but the tongue-tip halfway towards the r position, an extension of
the notation is required (as illustrated in Figure 12.3). The same thing is
also true of the notation of consonants. When we produce the phoneme le
we normally assume that the mouth is already in the e formant position
during the production of the l apart from the tongue position for the l. This
need not be the case. We may for example put the mouth in the u position
for the l, changing with the articulation of the l to an e formation (Example
12.29). A notation for this is illustrated in Figure 12.4.

Figure 12.2 Some vowel signs from the International Phonetic Alphabet.

Figure 12.3 Examples of extensions of phonetic notation.

Figure 12.4 Indication of specific formant-structure for consonant.

The filtering techniques we have been discussing can be applied not

only to exhaled pitched sounds but also to inhaled and/or pulsed sounds,
e.g. inhaled glottal clicks (Example 12.30).
These filtering techniques can also be applied in modified form to
both tongue and lip vibrations. Vocal aperture and lip-rounding variations
may be applied to tongue vibrations. In fact, with practice an analogue of
the Shepard tone effect can be generated vocally where the fundamental
pitch of the rolled-r vibration falls while the pitch area stressed by the
filtering rises (Example 12.31). In fact the position of the tongue's arch can
also be moved smoothly from one position to another, but this is more
easily classified as a sound transformation (see below).
Because lip vibrations take place in front of the oral cavity filtering
effects are less effective with these sounds but they are in fact possible
(Example 12.32). Cheek vibrations may be filtered in an altogether different
fashion by hand manipulation of the cheeks. What appears to happen is that
the cavity between the cheek and the gums is varied in size, changing the
pitch of the resonance (Example 12.33).
In addition, a second level of filtering may be added by the hands
which may either form a cup external to the mouth or a funnel-shaped
aperture (made with one hand) partially closed off (with the other hand).
Any sounds whatever produced in the oral cavity may be filtered using
these techniques. A particularly interesting effect can be produced with the
funnel technique if the other hand is waved very rapidly backwards and
forwards away from and towards the mouth of the funnel. With noise-based
sounds (see below) this produces an effect akin to electronic phasing
(Example 12.34).
The technique of formant filtering may also be used to emphasise
particular harmonics of a tone. In this way it is possible to play the
harmonic series components of (for example) a glottal tone (Example
12.35). The production of very high harmonics allows the continuous
sliding of the upper partials as they are close enough together to meld into a
continuum. Harmonics may also be ‘played’ above sub-harmonic
production and even (though the effect does not seem to be particularly
strong) over high-pitched tongue vibrations (Examples 12.36, 12.37).

Noise

Noise-type sounds may be generated by setting up turbulence in the air-

stream travelling through the oral (or nasal) cavity. Such sounds are most
usually indicated by the various noise-based consonants (see Figure 12.5).
Most of these are distinguished from one another by the particular
placement of the tongue except for the f which is produced by a contact
between the teeth and lips. Given the particular conformation of the tongue
each consonant corresponds to a particular formant structure but these may
themselves be varied in average pitch by changes in the size of the mouth
cavity (Example 12.38). In addition, by changing the formation of the
tongue we may pass from one formant formation to another and these two
types of filtering motion may be combined (as was the case with vowel
sounds) (Example 12.39). We may also extend the phonetic notation to
indicate noise-states which lie halfway between states to which the phonetic
signs refer or are in fact combinations of these states (as in the case of x and
f; listen to Example 12.40) (see Figure 12.6).
With noise sounds, however, the filter may be made sufficiently
narrow that it ‘rings’ and we produce the characteristic sound of whistling.
Because whistling is essentially a filtering effect it is difficult to alter the
formant spectrum of a whistle sound (though not totally impossible).
Normal whistling is produced by a filtering at the lips. Pitch is varied
by altering the size of the oral cavity and two registers may be distinguished
according to whether the tongue arch is to the back or to the front of the
mouth cavity. A third and higher register can be produced by filling the
cheeks with air and pushing the tongue forward (Example 12.41). Whistling
may also, however, be produced over the tongue. This may be with the
tongue arch in its rear position (Example 12.42) which does however not
produce a very strongly focused pitch like normal whistling. Alternatively
the tip of the tongue may be used either in the sh position (Example 12.43)
or in the s position (Example 12.44). Finally, whistling may be generated
with both the tongue and the lips, producing a double whistle (Example
12.45). In this case, as pitch is produced almost entirely by varying the oral
cavity aperture, the two pitches almost inevitably move in parallel motion.

Figure 12.5 Phonetic notations for various noise-band consonants.

Figure 12.6 Examples of extensions of phonetic notation for noise-consonants.

Double and treble production

Within the vocal tract, it is clearly possible to set in motion two oscillators
or other sources simultaneously. This may of course produce merely a mix
of two kinds of sound, but in many cases the two vibrations interact with
each other and intermodulation effects are produced. Figure 12.7 charts all
the double production sounds. In certain cases I am not sure whether a
certain sound can be produced or not and I have therefore entered a
question-mark in the chart. The numbers in the chart refer to sound
examples on the tape (Examples 45–66).

Figure 12.7 Double production (numbers in the boxes refer to Examples).

The types of sound available are not as simple as this chart might
suggest; the particular nature of the individual oscillations, their relative
amplitude and pitch and the extent to which they interact may vary or be
varied, thus producing quite distinct classes of sound-objects. Thus, when
we combine glottal vibrations with windpipe vibrations, the latter may be of
the stable low-frequency type which gives us the ‘gravel voice’ or the less
clear and more forced ‘air’ type which interact with the larynx sound to
produce a sound-complex like a roar or bark (Example 12.46).
Where glottal sounds are combined with tongue oscillations in speech
discourse, the amplitude of the glottal vibration is usually much higher than
that of the tongue oscillation which thus appears as a mere colouring of the
glottal pitch. However, if the strength of the tongue vibration is increased
we become more strongly aware of the amplitude modulation which is
taking place (the tongue is an audio rate amplitude modulator of the glottal
pitch stream). If the tongue vibration is made even louder and the glottal
vibration reduced in level the glottal pitch becomes a mere decoration of the
strong tongue vibration. In addition the two oscillators may be tuned to one
another. Where they are of equal strength and in similar register this tends
to take place ‘naturally’. As the physical system of the oral cavity seeks
positions of lowest energy, we discover ourselves producing octaves and
fifths almost automatically and it is much more difficult physically to
produce other intervals. By slightly detuning one of the oscillators, beating
effects can be produced which may be physically felt within the mouth (or
on the lips) (Examples 12.47–12.50). In glottal/cheek and glottal/lip
vibrations there may be a remarkable difference between ‘rounded’ slow-
stream impulses with glottal pitch colouration and chordal effects produced
by the inter-modulation of mid-frequency glottal and lip or cheek vibrations
(Examples 12.51–12.53).
If a highish glottal note and a whistle are tuned to an octave and then
the whistle tone moves slightly away from the octave intermodulation will
produce chord-like colouration of the resulting sound. This process is
exactly analogous to what takes place (electrically) in a synthesiser. If on
the other hand a very high s-whistled note is produced fairly quietly against
a deep (male) glottal note, an almost grain-like amplitude modulation is
imparted to the note (Example 12.54).
Windpipe sounds can be combined with other articulations but
currently I would not recommend anyone to try these too much! It is also
possible to vibrate the tongue in two modes (rear arch and tip)
simultaneously (Example 12.55). The forward vibration tends to be a
subharmonic of the rear vibration. In a similar fashion the lips and tongue
can be simultaneously vibrated. Tongue-tip vibrations can be easily
combined with loose-lip and manually-stretched-lip vibrations, often
producing intermodulations which can be felt in the lips (Example 12.56)
and either low register or high register pitches may be tuned. Noise bands
are easily and effectively amplitude modulated by low-frequency tongue,
uvula and lip vibrations and noise may also be made to colour the more
high frequency vibrations of these types (Examples 12.57–12.64 give some
examples of these combinations). Even whistling can be amplitude
modulated by tongue-vibrations (easier with rear-arched tongue than with
tip of tongue). The lip modulation of whistling is the basis of the
‘trimphone’ imitation which became a craze in Great Britain in the early
1980s (Example 12.65).
We can go beyond this and generate three or even four sounds at
once, for example noise turbulence, tongue vibration, lip vibration and
glottis vibration may all be activated simultaneously and controlled
independently of one another (Example 12.66). Some special cases can be
observed, for example, if a glottal/tongue-tip vibration in the male low
falsetto register is passed through pouted lips, they may be caused to vibrate
in a very subtle way, producing a trumpet-like colouration of the sound
(Example 12.67).

Air-stream and other effects

Many of the sounds discussed so far can be altered or articulated by

modifications of the air-stream which initiates them. The manual and
physical control of the air-stream (discussed below) in particular may be
applied to lip and tongue vibrations as well as glottal vibrations. The other
techniques apply to glottal vibrations except where otherwise stated. If the
glottis is activated by a sudden rush of air, we produce the roaring sound
discussed previously. If, however, the glottis is activated very quietly by a
strong and fairly constant stream of breath we first produce a mix of pitch
and air sounds (‘half-wamp’). If the glottal sound is made even quieter it
begins to destabilise and the pitch content breaks up (‘quarter-wamp’)
(Example 12.68).
If the glottis is fairly relaxed and the air-stream is restricted at (I
believe) the epiglottis a sub-harmonic is produced which differs from the
usual sub-harmonics in being produced in this state of tension (normal sub-
harmonics need great relaxation of the vocal apparatus). If only a very small
amount of air is allowed to pass the epiglottis and much pressure is applied
to make this pass we produce a multi-phonic. Such production may be
called half-lunged (Example 12.69) (these sounds are called ‘constipation
multiphonics’ in the Book of Lost Voices). Half-lunged sounds must be
differentiated from unlunged sounds which are produced when no air passes
through the oral cavity. In the case of short pulses (discussed below) these
may be difficult to distinguish. The surest indicator is that during the
production of an unlunged sound it is possible to breath normally through
the nasal passages as if no sound were being produced. Half-lunged sounds,
however, need sufficient air and ‘back pressure’ that independent breathing
during their production is impossible. An example of an unlunged sound is
the unlunged whistle (Example 12.70) which can be produced by sucking
air into or pushing air out of the mouth cavity by a movement of the tongue.
Notations for these are given in Figure 12.8. These are not entirely
consistent with my previous use (in Anticredos and Vox-I) but are more
systematic.
Various (tremolo-type) amplitude modulations of the breath-stream
can be achieved manually or physically (flutters). Manual pressure on the
diaphragm can be used (Example 12.71). If the glottal signal is quiet and
the formants are high whilst the diaphragm is drummed rapidly with the
fingers, the stream of glottal pitch can be broken up into a series of short
staccato sounds (Example 12.72). Shake-head flutter (Example 12.73),
drum-glottis flutter(Example 12.74), shake-body flutter (Example 12.75),
drum-cheeks flutter (Example 12.76), strum-lips flutter (Example 12.77),
strum-nose flutter (Example 12.78) and hand-cup flutter are all that they
seem. Some of these may be applied to tongue vibrations (Example 12.79)
and lip vibrations (Example 12.80).

Figure 12.8 Some suggestions for notation conventions.

 A natural oscillation is that known as ululation (Example 12.81). This
may be used across a break in the voice to produce an alternation in pitch
(Example 12.82). The ‘depth’ of the ululation may be increased, producing
a sound more like laughter (Example 12.83).

Water effects

Saliva (or externally introduced water) may interact with the articulation of
vibrations. Gargling is the most obvious example of this. Saliva often
affects the sound quality of arched-tongue vibrations (Example 12.84). In
particular, the noise sound x has a great number of possible modes when it
is combined with saliva-water sounds (Example 12.85). The sound may be
filtered in various ways (Example 12.86). It may be half-lunged and then
filtered again (Example 12.87). It may be plosively attacked with a k and a
short rush of air to produce the children's ‘gun’ sound (Example 12.88). It
may be half-lunged, filtered to produce very high partials and produced
staccato and plosive (Example 12.89). A rational notation for the
distribution of harmonics in this sound is very difficult to achieve because
although the high partials are strongly emphasised, it is clearly still possible
to vary the resonance of the mouth cavity that is produced. The notation
shown in Figure 12.9 uses the ‘stave of harmonics’ to indicate that this is a
high partial sound but a simpler mnemonic is proposed. This sound
incidentally can be combined with tongue-tip vibration (Example 12.90).
Saliva-water effects also account for the pitch content of the sounds
indicated in Figure 12.10 (Example 12.91).
Most of these half-lunged water sounds can also be produced inhaled.
Inhaling, however, also generates a number of other sounds such as the
modifying of inhaled lip vibrations by water held behind the lips (Example
12.92) and various sounds around the sides of the tongue (Example 12.93).

Figure 12.9 Stave notation for harmonics.

 Figure 12.10 Notation for unvoiced sounds with water content.

Transformations

Within this huge class of exhaled sustainable sounds it is possible to make

transformations from one type to another. The simplest transformations are
those of formant change demonstrated earlier. We may also make
transformations between the different kinds of oscillations produced by the
tongue (Example 12.94). We may add or subtract water (Example 12.95).
We may move from the lunged to the half-lunged or vice-versa (Example
12.96) or from the normal to under- or over-breathed (Example 12.97). We
may make transformations from normal to fluttered production (Example
12.98) and we may also make a transformation between two sounds by
passing through an intermediate mix, as for example moving from a tongue
vibration to a lip vibration (Example 12.99).

Inhaled sounds

Sounds may also be produced when air is inhaled. In many cases these are
the same or quite similar to those produced on exhaling, but in a number of
cases quite different sounds are produced. Many of the sounds produced in
this way exhibit instabilities, either in pitch, spectral content or sub-audio
attack rates.
Vibrating the lips by inhaling can produce pure tones, trains of pulses
or multiphonics (Example 12.100). The vibrations may be controlled and
articulated by using the heels of the hands (Example 12.101). All these
sounds are unlunged. The tongue may be made to vibrate, both in retroflex
position unlunged (Example 12.102) or at the uvula as in snoring (Example
12.103).
Finally, the glottis (and possibly the windpipe) may be made to
vibrate while inhaling. If a lot of air is drawn inwards the effects produced
by (I believe) the larynx and windpipe are heard. A better method of
production, however, is to draw air in regularly and slowly (as air would be
expelled during normal singing). By varying the tension of (I believe) the
larynx and the filtering in the oral cavity a great number of different kinds
of sounds can be produced: from pure tones (Example 12.104) which may
be outside the normal range, click trains (Example 12.105), sub-harmonics
(Example 12.106), more complex multiphonics (Example 12.107) to
complex and unstable oscillations (usually produced at the end of a long in-
breath when the pressure inwards is difficult to maintain) (Example
12.108). The instabilities in these latter sounds can be felt as a kind of
irregular beating in the larynx. In the various complex sounds different
aspects of the complex spectrum can be emphasised by the filtering process
(Example 12.109).
As some of these inhaled sounds are unlunged it is possible to
simultaneously produce inhaled and exhaled sounds. For example, one can
produce inhaled lip vibrations while projecting glottal vibrations through
the nose. Furthermore, various of the flutter techniques can be applied to
the air stream.

Pulses

Short sounds or pulses may be produced in a variety of ways in the vocal

tract. The pulse may be released or initiated by the epiglottis (Example
12.110), the arched tongue which may touch the roof of the mouth at the
rear, in the centre or at the front (Example 12.111), the tip of tongue
(against the roof of the mouth ) which may be retroflexed or further
forwards (Example 12.112), the tongue and the teeth, the tongue and the top
lip (Example 12.113), the teeth and the lips (Example 12.114) or the lips
(Example 12.115). Pulses may be lunged, half-lunged or unlunged
(Example 12.116).
Pulses may be either normal or plosive. If they are lunged, plosive
production tends merely to make them louder but if they are half-lunged or
unlunged plosive production may produce a quite different sonority by
slight modification of the process of production. Thus the half-lunged sound
d can be made plosive by retroflexing the tongue and releasing it plosively
forwards (Example 12.117). Similarly the half-lunged p may be made
plosive by folding the top teeth and lip over the bottom lip and releasing air
plosively (Example 12.118). The same technique can be applied to the
unlunged p (Example 12.119).
If pulses are lunged they may be voiced or unvoiced. The voicing
may be dominant as in the normal production of voiced consonants in the
language stream or secondary, and various amounts of breath may be added
(Example 12.120). The ‘voicing’ need not be a glottal sound but may come
from the windpipe, tongue vibrations or even whistling (Example 12.121).
If lunged they may have more or less air-stream content (Example
12.122) and of various kinds (Example 12.123). The air-stream may also be
glissandoed (Example 12.124). In fact the pulse itself may be suppressed to
produce a plosive air-stream effect (Example 12.125). If lunged or half-
lunged they may be stopped. In this procedure the air-stream is cut off
abruptly, almost as soon as it has been initiated. This may be achieved with
the epiglottis, by moving the tongue into the g, k or t position but not
releasing it from the roof of the mouth or by plosively closing the lips. The
stops may be used to make the pulses exceedingly short and this alters their
character. A g may be used to alter the character of a plosive d (d(g)!)
(Example 12.126). When a p is used as a stop its slight buzzing may alter
the character of the pulse (Example 12.127). Such very short stopped
sounds may also be used as envelope shapers for glottis vibrations,
producing for example (with d(g)!) extremely loud impulses or, with a
quiet, low frequency and slightly glissandoed glottal vibration, an excellent
drum imitation (Example 12.128).
Stops may also be used to produce end-pulses, such as those
produced by a ch- or a k-stop on an h-air-stream (Example 12.129) or a p-
stop on the end of a s-stream (Example 12.130). The slapping of the tongue
into the ‘t’ position on the roof of the mouth (unlunged) and the slapping
together of the teeth provide two other pulses (Example 12.131).
If pulses are produced by the lips, the lips themselves may be buzzed
whether the pulse be lunged (Example 12.132), half-lunged (Example
12.133) or unlunged (Example 12.134). During an unlunged lip-buzz of this
type air may be pushed out of the mouth with the tongue, producing a
glissando as the mouth opens (Example 12.135). Lip pulses may also be
manually initiated. In this way quite loud unlunged and unbuzzed sounds
may be produced (Example 12.136).
In all cases the pulses produced can be filtered by altering the shape
of the oral cavity (Example 12.137). This applies even to unlunged p
sounds where the position of the arched tongue and the degree of pouting of
the lips can alter the resonance of the mouth cavity (Example 12.138). Even
with such very short sounds the filter may be made to glissando, for
example, with the plosive click (discussed below) (Example 12.139).
 Certain pulses may be produced simultaneously, for example g and k
(Example 12.140) or t and k (Example 12.141). Pulses may also be paired
and iterated as in tktktktk which may be lunged (Example 12.142), half-
lunged (Example 12.143) or, if produced by a lateral movement of the
tongue, unlunged (Example 12.144). The physiological limits of this
iteration may be extended to produce g and th (Example 12.145). Lip and
tongue pulses may be thus iterated as in ptptptptp which may be lunged
(Example 12.146), half-lunged (Example 12.147) or unlunged (Example
12.148). Lip pulses may also be iterated manually and may be lunged
(Example 12.149) or unlunged (Example 12.150).
Pulses may also be produced during inhaling at the epiglottis
(Example 12.151) and variously with the tongue and lips. The most
interesting of these are perhaps the unlunged clicks such as the lateral
tongue movement (Example 12.152), the vertical downwards tongue
movement (kl) (Example 12.153), the plosive, pure-resonance, vertical
downwards tongue-click (Example 12.154), the t (or tut) click (Example
12.155) but also the th-click (Example 12.156) and the ‘kiss’ with various
degrees of lip pouting (Example 12.157). As with exhaled pulses some of
these may be combined. For example, the plosive and the kiss (Example
12.158) or the lateral and the kiss (Example 12.159).

Transitionals and percussives

A number of sounds used as consonants in language are essentially

transitions from one formant state to another, produced by movements of
the tongue or lips. The consonants l, m, n, ng and ‘deep’ ng (i.e. with the
arch of the tongue towards the rear of the mouth) are of this type. By
controlling the oral aperture it is possible to use these consonants to make
transitions between harmonics of a given glottal pitch (Example 12.160).
Transitionals may also be iterated as in mnmnmn and n-ng-n-ng (Example
12.161).
A number of short sounds may also be generated by the use of the
hands. For example, hand-clapping in front of the mouth used as a variable
resonator (Example 12.162), tapping the teeth with the mouth used as a
variable resonator (Example 12.163), popping the finger out of the closed
lips (Example 12.164), filling the cheeks with air with the lips tightly closed
and then striking them with the fingers to expel the air (Example 12.165)
and the now-famous ‘water-drop’ discovered by the San Diego Extended
Vocal Techniques Group (see 1974): an unlunged sound where the finger
strikes the cheek with the teeth parted and the mouth open and as it does so
the tongue pushes air out of the mouth in an unlunged whistle (Example
12.166).

Multiplexes and complex articulations

Using the multiplex notation described in Chapter 5 (an example is shown

in Figure 12.11) we may combine various short sounds into a more
complicated stream. In Figure 12.11 the iterated string pk∫lgr is combined
rapidly and randomly with lunged and half-lunged X+ sounds and lip-
flabber while the box containing vowels is a mnemonic for ‘maxvary’
mouth vowel shape (Example 12.167). The horseshoe shaped symbol
indicates that the vowel formants are to be varied as rapidly as possible. We
may now further articulate this multiplex by putting the tongue in and out of
the mouth as rapidly as possible (Example 12.168).This sound may now be
super-imposed, for example on a glottal pitch which glissandos rapidly and
at random over its maximum range (Example 12.169). Using these symbols,
together with the continuum and transformation signs (see Figure 12.12)
and our general knowledge of the sound repertoire of the vocal tract and the
individual physiological parameters, we can develop rapidly articulated and
evolving sound-streams of great complexity.

Figure 12.11 An example of multiplex notation.

 Figure 12.12 Further examples of extended vocal technique notation.

Notation

In Figure 12.12 various other aspects of the notational system for voice-
sounds are indicated. The vowel and consonant symbols are derived from
the international phonetic alphabet (Figure 12.13). In assembling a score a
three-level representation is used (see Figure 12.14). At the upper level
durations and loudness are indicated in the traditional musical fashion, at
the bottom level detailed phonetic and extended-phonetic notation is used to
indicate the details of the sounds. In the central level these sounds are
notated using graphic symbols which allow us to indicate pitch, pitch
motion, transformation, intermodulation and so on.
The international phonetic alphabet has been developed for linguistic
analysis of phonemes. In a performance situation, however, such diversity
of symbols may become confusing and it seems more practicable to use a
smaller set of symbols and methods of combining or modifying them (see
previous diagrams). In addition the phonetic alphabet has been derived from
natural languages and does not cover the whole human repertoire, therefore
modifications and extensions are required.
 The system of notation developed here has a degree of redundancy. In
particular information is given both in a ‘phonetic’ format and in a graphic
format. This redundancy is useful when it comes to reading notation during
a rehearsal or performance. The notation is also somewhat eclectic, using
devices drawn from standard repertoire music, contemporary music and
phonetic research. This, however, has the advantage that we are able to pass
over into conventional musical or conventional linguistic use of the voice
without any abrupt change in the way we represent the sounds.
It is also possible to present a systematic physiological notation for
the sounds of the human vocal tract (and the body in general). This has been
developed by Jean-Paul Curtay (Figure 12.15 which is from Curtay (1981)).
This notation is in fact more systematic and is used by Curtay in his
performances. However, I would still tend to prefer the eclectic method
which retains the links with language and conventional music and allows us
to notate complex sounds such as multiplexes while using physiological
descriptions as a very useful aid to performance practice. Furthermore, just
as traditional music notation tends to channel the aesthetic possibilities
(Chapter 2), even these extended notations have some bias towards a
physiological and a sound-object-oriented perspective respectively.

Natural morphology

In the next chapter we will return to some of our ideas about a natural
morphology of sound-objects in relation to phonemes. However, even at
this level we can note certain distinctions which may relate to our natural
morphology classification. Curtay has suggested a gross classification of
the human repertoire into gaseous, liquid and solid.1 This might be given a
more general interpretation as an aspect of natural morphology. Sounds of
solid objects are generally of stable mass (or pitch)—this includes the air
resonances within fixed-shape objects, e.g. a flute. Liquid sounds on the
other hand will often have changing mass, tessitura, spectrum and other
features but this change will exhibit a specific class of form (like a bubble
archetype). Gaseous sounds however, will be varying continuously in
various parameters (particularly mass) without a definite class of
morphological forms emerging. Clearly there is very much more to be said
about this. Air columns or liquids vibrating within solid objects (the water-
in-saucepan effect) or air passing through liquids would need to be
considered but there is certainly an interesting natural morphological aspect
here. For example, the sounds of a classical synthesiser can be very stable
in their spectral properties, implying a ‘solidity’ of the source. The sounds
of the human voice, on the other hand (and of course of musical instruments
articulated by human beings), even when they attempt to be stable, in fact
contain micro-fluctuations of pitch, dynamics etc. partly because the
musculature acting as a physical source or articulator is not a rigid object.
We tend to prefer even in normal musical practice a certain small degree of
‘liquidity’ in our musical objects.

The International Phonetic Alphabet

Figure 12.13 The International Phonetic Alphabet.
Figure 12.14 Use of three-level notation (used in three voice parts) in Anticredos.
 Figure 12.15 Examples of Jean-Paul Curtay's iconic notation (after Curtay (1981)).

Moving outward

From this repertoire of human sounds we may lay out areas of sonic
discourse. Focusing on the repertoire as physiological acts and perhaps
complementing them with visceral sounds recorded from within the body
(flow of the blood, etc.) we may evolve sound-structures which become a
kind of physiological diagnosis of the state of the organism. Something of
this feel is achieved in Curtay's Abridgement (Curtay 1981) where the
physiological landscape becomes the basis for sonic exploration. We may
treat the repertoire as a class of (intertransformable) sound-objects and
organise music accordingly (for example in my Anticredos) though we
cannot entirely avoid physiological (and para-linguistic) aspects of the
landscape adhering to the events at least on a first hearing. The sound-
objects may be extended into the electronic (as in McNabb's Dreamsong) or
into the world of recognisable sound-objects such as the transformation
from ss to birdsong in Red Bird. Given a good computer model of the
human voice as in the language Chant we may manipulate the form and
structure of the voice beyond that which is physiologically possible so that,
for example, individual glottal pulses may become bells (the formants are
narrowed and ring) or the energy in the formants may be refocussed and the
pitch articulated in such a way that we produce birdsong.
We may focus upon the mimetic abilities of the human voice made
possible by its enormous repertoire. The imitation of instruments such as
drums or trumpets has been touched upon earlier. Natural morphology in
relation to phonemes will be discussed in the next chapter as will the idea of
phonemic analogy (phonemic objects which are akin to but not exactly
mimetic of other sounds). We may extend the human repertoire by the use
of external resonances; thus brass instruments amplify and stabilise lip
vibrations extending their range of loudness and controllable pitch. We may
model the voice on musical instrument technology, separating out pitch as a
parameter and developing the field of song.
We may focus upon the paralinguistic articulation of the sound-
objects. Such paralanguage may be based on (originally) involuntary
physiological states, gestural expression or linguistic conventions. In this
way we produce a kind of phoneme-free poetry. Such paralinguistic
articulations may enter into instrumental practice, particularly in relation to
pitch and timbre control on the trombone or pitch, timbre, breathiness
control on the saxophone. Finally we may select specific sound-objects
from the repertoire and combine them in particular ways to produce
phonemic objects. We then enter the sphere of language, of linguistic syntax
and semantics. In the following chapters we will look more closely at this
world from a sonic art viewpoint.
Figure 12.16 indicates some of these many possibilities. (Fine arrows
indicate areas between which there is a continuum of intermediate
possibilities or an interaction of perceived categories.) Note, however, that
these cannot truly be represented on a two-dimensional surface. The
implications of human vocal utterance are multi-dimensional. The
biological, paralinguistic, linguistic, mimetic and musical may all be
present in an utterance. In sonic art we will structure this material in order
to focus in upon one or several aspects of this amazing universe of sounds.
Although the archetype of the keyed musical instruments, fashioned in the
image of a theory of pitch, has been the dominant focus for musical
thinking in the West for at least 300 years, at this moment of enormous
technological and musical change there can be no doubt that we shall return
to the human voice for our inspiration as “voice is the original instrument”.

Figure 12.16 The relation of the elements of the human repertoire.

1 Although mentioned in the spoken introduction to ‘body music’ on the cassette (where solid is
strictly referred to as tissue), Curtay (1981) concentrates on a discussion of method of production
which he divides into three levels: excitation, emitter and modulation (or resonance) indications.
Wishart interviewed Curtay at the time of his visit to London in 1981 and his material has been
elaborated through this personal communication and through Curtay (1983) (Ed.).
 Chapter 13
PHONEMIC OBJECTS

gadji beri bimba

glandridi lauli lonni cadori
gadjama bim beri glassala
glandridi glassala tuffm i zimbrabim
blassa galassasa tuffm i zimbrabim [...]
(Hugo Ball 1974: 70)

We must return to the innermost alchemy of the word, we must even give up the word
too, to
keep for poetry its last and holiest refuge. (Ball 1974: 71)

Sequence and Morphology

From the vast array of possible sound-objects available from the human
repertoire any natural language selects only a small proportion and
combines these phonemes into phonemic objects. These are the basic sound
units of any language and correspond roughly to the notion of the syllable.
Phonemes themselves are then combined sequentially to form morphemes
(words), the smallest independently meaningful units of language. We will
not go into this in greater detail, however, as this is not a linguistic analysis
but an attempt to explore the world of phonemic objects as a special class of
sound-objects for the purposes of sonic art.
 Any particular language will exclude a large number of possible
human utterances from the sphere of the phonemic. If, however, we scan all
existing (and extinct) languages, we will discover that quite a large
proportion of the sounds in the human repertoire are being (and have been)
used in human language discourse. For example, inhaled clicks are used as
consonants in a number of Southern African languages but not in any
European languages.
Phonemic objects are almost paradigmatic examples of sound-objects
of complex morphology. For example, if we consider the word ‘when’, its
written form suggests that it contains four consecutive objects. A superficial
listening might suggest that it contains three separate sound entities (a ‘w’,
an ‘e’ and an ‘n’). If, however, we speak this word very slowly we will
discover that it is one coordinated sound articulated by the opening of the
lips (and associated widening of the oral aperture) and the coordinated
motion of the tongue concluding where the tongue tip reaches the roof of
the mouth and cuts off the air stream. As a sound-object, therefore, this
event consists of a complex but continuous motion through the formant
space, most likely simultaneous to a slight movement of the fundamental
pitch. This kind of continuum exists in most verbal objects, except where it
is explicitly chopped up by stops and pauses. Thus, in the word ‘say’ it may
seem superficially that we have two distinct objects ‘s’ and ‘ay’. If,
however, we speak these two objects (even very carefully) and record them
onto tape and edit the two together we will not reproduce (except
approximately) the word ‘say’. In the speech stream there are subtle
transformations both of formants and into and out of noise turbulence. The
speech stream is thus an archetypal example of complexly evolving timbral
morphology. This will be discussed again in the next chapter.
For the purposes of linguistic analysis it is necessary to separate off
the distinct units which form the basis of the ‘digital’ coding of language.
From a sound morphological point of view, however, this can be quite
misleading. For example, the computer model of the singing voice
encapsulated in the programme Chant had by 1981 very successfully
modelled vowels in terms of definable and fixed formant structures which
could be reproduced by simulating the effect of a related system of filters
on an impulse stream. Modelling many of the consonants, however, proved
to be more difficult as their absolute characteristics varied very greatly with
context, both in absolute formant structure and, for example, onset time of
noise turbulence. They were thus characterised more by second order
characteristics (characteristics of the process of change itself) than by any
absolute properties. It is therefore important not to confuse the economy of
print with the reality of this speech stream.
A similar complexity of timbral morphology may be found in the
utterances of other creatures. For example, various kinds of birdsong which
appear superficially as trills or bubblings of fixed pitch have in fact a
complex internal structure which can be heard when they are slowed down
(Example 13.1). I would not agree, as some observers have suggested, that
these internal complexities do not exist for the human listener. On the
contrary, it is possible to predict with some accuracy what the internal
structure of a sound will be when slowed down if one develops one's ear for
dynamic morphological properties. Even without this degree of
discrimination, however, the listener can usually observe a qualitative
distinction between various songs which derive from these rapid internal
articulations, even if he or she is unable to describe how they arise.
Phonemic objects, then, are interesting sound-objects from the point
of view of the musician interested in sound-objects of dynamic morphology.
Interest has also arisen, however, from an entirely different direction. In
1947, the Rumanian Isidore Isou published Introduction à une Nouvelle
Poesie et à une Nouvelle Musique (Isou 1947) In this book Isou proposed a
new type of poetry based on the letter, to be known as lettrism. Isou's view
of the development of poetry is illustrated in Figure 13.1. The lettrist
movement led to many interesting developments, including aphonic poetry
(see Figure 13.2) and also to a deeper interest in the expressive possibilities
of phonemic sound-objects beneath the level of the word. Curtay's work
(discussed in the previous chapter) develops out of this tradition. This is an
interesting juncture of fields of thought about the world of sound, that
springing from music and that springing from poetry and language. As we
begin to consider larger units of language (such as words, phrases,
sentences etc.), considerations of semantic meaning (or the lack of it) will
enter increasingly into our field of view until we finally arrive in an entirely
different area of human discourse (didactic or scientific prose). At the level
of the phoneme, however, we are still deeply embedded in sonic art.
We may also note that from the human repertoire described in the
previous chapter we can create imaginary phonemic objects and in fact
construct imaginary languages and linguistic streams from these. (We can
also construct imaginary linguistic streams from ‘valid’ phonemic objects,
as we shall see.) This kind of imaginary language retains our material for
the field of sonic art as questions of semantics cannot enter into the
listener's perception (though paralinguistic and other signs and signals may
remain part of the experience). The four-voice piece Vox-I concludes with
the peroration of such an imaginary text.

Mimesis and phonemic analogy

Language divides the human repertoire into two distinct fields. Those
sounds (or sound combinations) which may enter into the construction of
language sounds and those sounds (or sequences) which may not. If we do
not make any restriction on the sounds we can use, the immense pliability
of the voice makes it able to mimic an enormous variety of sounds. In the
previous chapter we discussed how it was possible to use plosive
consonants and stops to imitate drums very accurately. Similarly, various
entertainers and serious investigators are able to imitate birdsong
sufficiently accurately to fool other birds, to imitate the idiosyncratic
features of another person's voice, or to imitate the sounds of natural objects
or machines. Even approximate mimesis makes the construction of the kind
of transformations into other recognisable sounds used in Red Bird a
possibility. Once, however, we restrict ourselves to those sound-events
deemed suitable for use in a particular language this type of mimesis
becomes more problematic.
Figure 13.1 Isadore Isou's view of the development of poetry (adapted from Isou (1947)).
 Figure 13.2 Example of notation from Roland Sabatier aphonic poem Histoire.

Instead, we must substitute the technique of phonemic analogy. Here,

sounds may be imitated by a closest possible approximation constructed out
of sounds available as phonemes for language. Thus we might indicate a
drum sound by the syllable du or bom. This imitation retains certain
characteristics of the sound it imitates. It has the initial attack, though now
somewhat soft-edged and the resonance of the drum skin is suggested by
the fixed resonance of the vowel formant. In the case of bom the slight
natural reverberation of a deep drum with a loose skin is suggested by the m
continuation of the vowel. As the mouth moves into the m formation the
formants of the o are rapidly filtered out from the top downwards, a process
similar to the rapid damping of the drum skin. There is no way, however, in
which any listener would mistake these syllables for the sound of a drum.
Phonemic analogy need not, in fact, use the phonemes of the language. For
example the children's gun imitation kX+ uses the sound X+ which does not
occur in the English language. Nor, in fact, does the sound-object sound at
all like a gun but more like a small explosion. The object is also interesting
because it is clearly not based on the restricted phoneme set but the use of
such sound analogues may be seen to relate to the kind of vocal set induced
by speaking a particular language, i.e. particular settings and articulations of
the vocal tract muscles become ‘second-nature’ to the speakers of a
particular language and they will therefore seek to make such analogies
using these second-nature articulations. kX+, though not within the
phonemic set of the language, is not outside the second-nature vocal set.
A second aspect of phonemic analogy is illustrated by ornithologists’
attempts to write down birdsong using specially created syllables. Figure
13.3 (adapted from Schafer 1977: 30 citing Nicholson and Koch (1946))
shows a number of examples. Their most interesting feature is the way in
which formant movement is used to ‘track’ pitch movement; thus in the
great titmouse, the syllable ‘tsoo-ee’ clearly indicates a figure which moves
upwards in pitch. In fact, if this is a good imitation the birdsong should
slide up in pitch as the speech stream transition from oo to ee involves a
spectral glide from low formants to high formants. The formant tracking of
pitch is not confined, however, to the vowel. In the greenfinch, for example,
wah-wah is perceived as spectral glides up through the formants and
presumably represents a rising pitch figure. At the same time other
consonants, as in tchee or tic, may indicate spreading of the spectrum
during the attack (or decay) portions of the bird tones producing ‘impure’
tones which cannot thus be represented simply by a formant-tracking
procedure. Similarly, iterated sounds such as the rolled r, may be used to
indicate rapid streams of impulses in the song.
 Figure 13.3 Phonemic analogues of birdsong used by ornithologists.

These formant-tracking and spectral-analogue procedures were also

used in an art context by Raoul Hausmann in his poem Birdlike (1946), for
example in this extract:
Pitsu puit puittituttsu uttititi ittitaan
piêt piêt pieteit tenteit tuu uit
ti ti tinax troi troi toi to
Iti iti loi loi loiouttouto!
(Motherwell 1989: 316)

There may of course also be direct analogies with birdsong as certain birds
(such as parrots) are able to articulate clear formant structures (such as in
their imitations of human speech).
Phonemic analogy and formant tracking of pitch can also be found in
human names for animal sounds. These vary in the ‘goodness of fit’. For
example, a cow, which produces a low-pitched sound, ‘moos’. A mouse,
producing a high-pitched sound, ‘squeaks’; a wolf, which sings a sustained
pitch which then gradually falls, is represented by a formant structure which
falls as in ‘howl’ (say it slowly). On the other hand, the low frequency
glottal/windpipe multiphonic (which humans can produce) of the pig or the
lion is only loosely represented by ‘grunt’ (where the ‘gr’ hints at the
subaudio oscillation of the windpipe) and ‘roar’. Such phonemic analogy
may breach the distinctness of natural languages, such as the various words
for the sound of the cockerel (cock-a-doodle-doo in English, kikeriki in
German, kokke kokko in Japanese, kio kio in the language of the Lokele
tribe of the Congo) or of sneezing (kerchoo in American, atishoo in
English, atchum in Arabic, cheenk in Urdu, kakchun in Japanese and ach-
shi in Vietnamese).

Natural morphology; mouth symbols

We may go one stage beyond the concept of mimesis and ask whether we
can apply the criteria of natural morphology to the sound-objects produced
in the vocal tract. Clearly these sound-events are generated by processes
which may be physically described (turbulent air-streams, plosions, opening
and closing of apertures, etc.). Is there a natural morphological description
of certain kinds of phonemic (and other) vocal objects?
Jean-Paul Curtay has approached the same problem from a slightly
different point of view. Hence he considers both the motions and shaping of
the vocal tract organs in his conception of mouth symbols. These two
conceptions are very close indeed and it is worth considering the slight
difference that does arise. In considering, for example, the word stop,
Curtay talks of “st- evoking a sudden interruption of movement in a rigid
vertical posture” (Curtay 1983).1 The rigid vertical posture is suggested by
the tight downward movement of the tongue and this particular st-
formation can be associated with a number of words (stake, stalk, stand,
stare, statue, staunch, stick, stiff, stop). Another way of describing this
would be that we hear a continuously sustained sound which is suddenly
interrupted by an impulse. The feeling of interruption of flow is equally
apparent from such a description and in fact, of course, the sound
morphology arises from the physical morphology of the sound production
process. If, however, we now consider the phonemic object sp-, Curtay
states that this is “evoking a circular movement” (spin, spiral, spool). The
circular movement is presumably suggested by the circularity of the lips, a
kind of spatial metaphor. Looking at it temporally (and in terms of what is
heard) we would suggest that the (air) flow of the s- is momentarily
interrupted by the constriction p- and then released into the rest of the word.
This is more evocative of the whipping of a top where a sweeping motion
of the whip is applied suddenly (the strike) to the stable spinning motion of
the top. Similarly, when a dancer spins it is necessary to tense the
musculature in a particular way and then suddenly release this energy,
allowing the body to spin as a result. The sense of stored energy and sudden
release into a stable motion is evoked by the time morphology of the word
spin.
It is interesting to consider some other examples. The phonemic
object sl- consists of a stream (or store) of energy (s) which is gently
released (l) into the stable motion of the vowel. The sense of gentle release
into movement is of course caused by the sliding of the tongue. We may
bring this motion to an abrupt end by the insertion of a stop consonant such
as p, to produce the word slip. This motion is so analogous to someone
slipping on ice, where the move into the continuous sliding motion is
abruptly interrupted by a fall that it seems unlikely to be coincidental.
Curtay quotes the related words sledge, sleigh, slide, slime, slip, slough,
slug, slant, slope, slash, slat, slit, slice, slither, slot, slender, slim. Some of
these associations are undoubtedly metaphoric or tangential (for example,
slope from slip and possible also slap from the act of killing by sliding a
sword into someone). Furthermore, the reader can now think of many words
which do not conform to this archetype. The point being made, however, is
not that all language is somehow made up of symbolic or metaphorical
sound-objects, but that some parts of language may originate in such
symbolism.
Spr-: a flow (or store) of energy (s), passing through a constriction
(p-) and continuing but being broken up into an iteration, (r-), as in spray,
sprinkle and, more metaphorically, sprout, spring.
Spl-: in which a continuous air stream (s-) meets a constriction (the
initiating mouth formation for p-), leading to a double release (pl-), as in
split, splice, splinter, splay, splash and, by metaphor or association,
splendour.
Gr- and scr-: in which an impulse or contact (g-) is followed by a
non-continuous (iterative), or we might say abrasive, motion as in grate,
grind (and possibly grip) and scratch, scrape, scrawl (even perhaps
scream).
Cl-: a double, rather than a clean, attack as in clang (as opposed to
dong), clatter, clink, clash. Certain word endings are also interesting from
this point of view, such as -ng, a resonant extension of vowels in which the
higher formants are gradually filtered off—sound is gradually directed
through the nasal cavity—as in many naturally decaying resonances. This is
found in bang, clang, bing, ding, ping, ring, dong, song. ‘-ash’, a resonance
which breaks up into turbulence, as in dash, smash, clash, splash, bash,
crash and, perhaps as a metaphor for the visual after-effect, in flash. The
word clash thus has an unclear (double) attack onto its resonance which
rapidly dissipates into turbulence.
These sound morphologies (which are illustrated graphically in
Figure 13.4) point in two directions. Curtay has suggested that the linking
of these phonemes with objects and activities in the real world is to a great
extent kinaesthetic, i.e. we feel the formations inside the mouth and thereby
associate them with activities or the shapes of objects in the external world.
This naturally leads us into the sphere of representation and language. At
the same time the sound ‘s’ which we feel as a store of tension because it is
produced by constricting the passage of air with the use of the tongue, is
also indicative of a similar or related physical situation in any natural world
event. These forms, therefore, also point towards a natural morphology of
timbral gesture.

Expletives and paralanguage

Phonemic objects may, however, carry gestural information outside their

timbral morphology. This may be suggested by pitch articulation, dynamic
or rhythmic articulation or physiological indicators such as breath state etc..
An interesting case to consider is that of expletives. If we consider four
words of equivalent meaning which might be used as obscenities: fuck,
screw, copulate, make love, we find that only the first is used regularly as an
expletive. This can be related to its timbral structure. A restricted flow of air
f- is released very briefly u into an abrupt stop k. This intense and brief
release of energy is like a verbal punch, a violent physical action is mapped
into violent verbal action and the tension caused by anger or pain is
symbolically released. Compare for example the words shit and the (slightly
less satisfying) damn.
Figure 13.4 Graphic representation of the morphology of some consonant clusters.
 Figure 13.5 Paralinguistic transformations of phonemic objects in Red Bird.

In general, however, the phonemes of language are not mouth

symbols. Their dynamic morphology or energy contour does not represent
either the object of action they stand for or anything about the immediate
internal state of the utterer (as with expletives). In these cases, however, it is
possible to impose a morphology on certain aspects of the phonemic object
giving (or altering) meaning to it which might not be carried linguistically.
In the work Red Bird the various phonemic objects contained in the phrase
‘listen to reason’ were spoken by a male and a female speaker in a number
of different ways indicated by adverbs, in order to project such implications
into the phonemic objects themselves (see Figure 13.5). In Berio's Visage
similar paralinguistic ‘meaning’ is projected onto phonemic utterances.
It should also be noted that certain phonemic objects already act as
signals, although they are not elements of language proper. Some of these
have been called vocal segregates and examples are ‘mm’, usually meaning
‘yes’ and the ‘tut’ click implying ‘how stupid’ or ‘fancy that’. The kiss,
though not used in the phonetic structure of most languages, does, however,
carry social connotations. Finally, it should be noted before we move onto a
consideration of language-stream that phonemic objects may be organised
sequentially into larger units which are yet not linguistic. A good example
would be the ‘Word-Machine’ in Red Bird (see earlier in text).

1 See Chapter 12 footnote 1. Curtay (1983) is unpublished.

 Chapter 14
LANGUAGE STREAM AND PARALANGUAGE

The fall (bababadalgharaghtakamminarronnkonnbronntonner-

ronntuonnthunntrovarrhounawnskawntoohoohoordenenthur-nuk!) of a once wallstraight
oldparr is retaled early in bed and later on life down through all christian minstrelsy.
(James Joyce, Finnegans Wake (Joyce 1939: 1))

’Twas brillig and the slithy toves

Did gyre and gimble in the wabe:
All mimsy were the borogoves,
And the mome raths outgrabe.
(Lewis Carroll, Jabberwocky (1872) (Carroll 1994: 28))

For a language utterance to convey its meaning, the linguistic signs need
not reflect in any way the properties of the objects or activities to which
they refer. There is nothing in common between the word ‘red’ and the
property of redness. This is the famous Jacobsonian arbitrariness of the
linguistic sign and is the assumption upon which most linguistic research is
based. However, when we say that the linguistic sign need bear no relation
to the signified, we are not saying that it must bear no relation. As we have
seen in our analysis of phonemic objects, such relationships can be
established in different ways in many cases.
The language stream itself conveys meaning in many ways (in many
different sonic dimensions). Taking a minimalist view, we may describe the
significant distinguishable elements of the speech stream purely in relation
to the formants. Roughly speaking, with an unvoiced speech stream, vowels
will be distinguished by specific formant structures and consonants by a
combination of specific qualities and specific structures of change of these
qualities. A typical speech act, however, is also characterised by a number
of other properties relating to its rhythm and tempo and their articulation, its
pitch and pitch articulation, its phonemic connectedness and so on. These
other properties are also capable of conveying meaning and particularly of
modifying the significance of the semantics that might be implied from a
written version of the sentence. Furthermore, we can define sonic properties
of the language stream which have nothing to do with any of this, for
example, aspects determined by the particular physiology of the speaker.
From the point of view of sonic analysis, this distinction between
language and paralanguage (as these other aspects have been called) is
somewhat arbitrary. It is not based on a distinction between what is
semantically meaningful and what is not, but on a distinction between what
is captured in writing and what is not. This is only true of certain writing
systems, however. In the Aztec codices sometimes the ‘speech scroll’—the
balloon issuing from the mouth of a character depicted—is specially
elaborated to indicated paralinguistic aspects. In one instance (see Figure
14.1) ambassadors delivering speeches are shown with knives coming out
of their mouths (left), whilst in another a Spaniard is shown talking to
Aztecs and his speech scroll is decorated with feathers indicating the soft,
smooth words he is using (right). Approaching these things from the point
of view of sonic art, therefore, we will talk in terms of timbre fields and
articulation fields which will be explained more fully below.
Comparing, for a moment, standard repertoire use of language and
standard repertoire music, we may characterise the difference in the
articulation of the sonic stream as follows: the melodic stream is
pitchdisjunct and may be articulated by timbral colouration (either in the
choice of instrumentation or within the internal morphology of the sound-
objects of instruments). The language-stream is timbre-disjunct (bearing in
mind the qualifications on the notion of vocal disjuncture mentioned in the
previous chapter) and may be articulated by pitch inflections. It has been
argued that a music based on the complex articulation of timbre could not
be as sophisticated as that based on the articulation of pitch. However, if we
investigate the language-stream, we will discover that the human brain has
a truly amazing ability to generate and perceive rapid articulations of
timbral quality. At certain points in the speech stream (particularly in
diphthongs or within consonants) the removal of just one thirtieth of a
second of sound is clearly noticeable (as the ear is crucially sensitive to
change-continua). It is often argued that this perceptual ability is crucially
linked to semantic understanding. If, however, we consider text-sound-art
using essentially meaningless phonemic strings and take into consideration
our discussion of mouth symbols and natural morphology of phonemes,
plus our ability to perceive simultaneously a wide variety of characteristics
of the speaker (age, status, regional accent, idiolect, physiological state, and
attitude), it is clear that a sonic art based on the articulation of timbral
characteristics may be quite as subtle as any pitch-lattice-based sonics.

Figure 14.1 Paralinguistic indicators in Aztec codices.

The language-stream may therefore be thought of as a model for such

a timbre-stream music. Such a music need not use the phonemic objects of
languages or even confine itself to the timbral objects within the human
repertoire. It need not be based fundamentally on the voiced/unvoiced
distinction of the speech-stream (but might use some other kinds of timbral
disjunction). It need not be pitch-stable (as the speech stream tends to be,
remaining within a narrow range of pitch), nor need it confine itself to the
tempo, rhythm and tempo and rhythm articulations typical in the speech-
stream (for example we might use such human repertoire sound-objects as s
or X+ as sustained objects, inside which the formant area is being
continuously filtered, or rapid multiplexes with a phonemic distribution
statistically related (or not) to language in general or a specific language).
Using the language-stream as a model, therefore, does not imply we are
going to produce some kind of ‘universal vocoder’ form of musical
discourse.
Conversely, any sound-stream which exhibits the articulation of
formant areas within or close to the range of the tempo and rhythm of
normal speech will be imprinted with the landscape language, hence a
timbral organisation not deriving in any way from the human repertoire
may make reference to and play with the landscape of the human voice and
the notion of utterance.
Thirdly, we may look upon a particular language-string as a sound-
object in its own right. In his commentary to Ommagio a Joyce (Berio
1959), Berio describes classifying the language strings into three types on
the basis of their very general (external) properties. These are the
continuous (e.g. sustained sibilants), the periodic (e.g. ‘thnthnthn’) and the
discontinuous (e.g. ‘Goodgod, henev erheard inall’). We can, however,
work with the sonic properties of language-strings in a much more specific
way.

Timbre field, phoneme field

In listening to (or working with) pitch-lattice music, it is often possible to

define an entity known as a harmonic field. A sustained chord, of course,
defines a harmonic field for as long as it is sustained. However, the crucial
feature of a harmonic field is that we do not need to state all its pitch
constituents simultaneously for us to be aware of its existence. Thus, the
opening section of the Webern Symphonie is based on a harmonic field
(Figure 14.2). The harmonic field is thus a property of a sequence of
sounds, a property of inclusion in, and exclusion from, a set.
In the sphere of tonal music we may slightly redefine a harmonic
field, such that the same note-name (e.g. C, E flat) in a different octave will
be regarded as the same note. Our recognition that we are in a particular key
relates to the definition of our harmonic field (in this new sense)
corresponding to the seven-note scale which defines the key. Chromaticism
within the key involves the expansion of the set of acceptable pitches (the
harmonic field). Modulation involves a change in harmonic field. In this
case, however, we might also consider the harmonic field of the entire piece
which will be the entire set of twelve pitches making up the chromatic scale
(or some sub-set of it). This in its turn is a sub-set of the set of pitches
making up the quarter-tone scale and of the infinite set of pitches making up
the continuum of pitch. We may thus focus in upon a piece on different
time-scales and describe it in terms of the harmonic field defined by the
sequential use of pitch.

Figure 14.2 The symmetrical harmonic field in Webern's Symphonie op. 21.

It is easy to see how the concept of a field may be adopted for the
spheres of timbre, timbre articulation, pitch articulation, phonemes and so
on. At the most general level we can talk about the timbre field of a
particular language. When listening to a language we do not understand, we
will be particularly aware of this feature: in Japanese the extreme (very high
and very low) formant areas used in some vowel production, in Dutch the
salival fricatives (X+), in English the sense of articulatory (consonantal,
dipthongal) continuity (compared with, for example, German). Such
features in fact may often lie at the root of certain kinds of cultural
prejudice where the timbral and articulatory aspects of the language are
taken to indicate something about a spurious ‘national character’ of the
entire group of speakers. This is essentially a confusion of the normal
timbral field of a particular language with the attitudinal ‘modulations’
which may be applied to the normal timbre field of the native speaker's
language.
Just as the definition of a harmonic field on a pitch lattice allows us
to define chromaticism (the inclusion of pitches foreign to the harmonic
field originally established), so the definition of a timbre field allows us to
define ‘chromaticism’ or ‘modulation’ from normal language practice.
Figure 14.3 gives a highly schematic representation of this idea. To do
justice to the concept we would need a multi-dimensional space in which to
draw this figure. However, from it we can see that a particular language will
have a characteristic ‘normal’ set of timbre types, articulations, etc. and
(even without understanding a language) we will be able to distinguish
regional accents or idiolects (ways of speaking characteristic of the
individual speaker) by their variation from this norm. Other variations from
the norm, which might overlap with aspects of accent or idiolect, will
indicate non-neutral modes of discourse (for example, anger or ridicule).
It is more interesting, however, to look at much more specific timbral
fields. Thus, any short verbal utterance contains a particular set of timbral
objects, and these define a timbral field. We may explore the
interrelationships amongst these objects, not only through their reordering
in a linear text (an approach from poetry) but also in simultaneous and
textural orderings (a choral approach). The objects may also be grouped
into ‘timbral-motifs’ (which may, in fact, be words or phrases). Thus, just
as Lutoslawski or Berio will define a harmonic field and a class of pitch-
groupings (melodic motifs) simultaneously, it is possible to do exactly the
same thing with timbre-stream material.
Figure 14.3 Timbre fired of a language and various subsets compared with a similar analysis for
harmonic fields in tonal music.

In Steve Reich's Come Out the phrase ‘come out to show them’ is
used purely as a timbral motif. Several copies of the phrase are played
initially in synchronisation and then gradually de-synchronised. As this
happens, rhythmic and timbral patterns (due partly to phasing effects) are
established, which arise directly out of the timbral properties of the sonic
object ‘come out to show them’ (or rather the specific speech utterance of
this phrase initially recorded on tape).
 Aspects of variation and change amongst timbral fields may be
observed in various text-sound pieces with no semantic content. Thus, in
Schwitters’ Ursonata (Schwitters 1993), for example in the ‘fourth
movement’, we can see first of all that the overall timbral field is confined
to a small number of phonemes—Grimm, glimm, gnimm, bimbimm, bumm,
bamm, Tilla, loola, tee etc.. Next we notice that there are large-scale
groupings, for example, we may divide attack-vowel resonance-m areas
(Grimm, bamm) from l-vowel resonance-l-vowel resonance areas (loola,
luula, lalla) and from attack-vowel resonance areas (Tuii, tee, bee). Within
these areas we may make further subdivisions, for example, in the first
section between areas stressing g and i, and areas stressing b and using a
number of different vowel resonances (u, i and a).
All this differs from our perception of field characteristics in pitch-
lattice music (apart from the obvious pitch-stream/timbre-stream
distinction) in a number of ways which are, however, not intrinsic to text-
sound composition. First of all, there is no counterpoint or chorusing.
Secondly, there is no indication of rhythm (which might, however, be
implied from the printed spacing) or tempo. Adding these, and other,
dimensions we can imagine a sophisticated contrapuntal art based on the
articulation of a multilevelled timbre or timbre-motif (possibly phonemic)
field structure.
These conceptions have a bearing on the construction of standard
poetry. Here the timbral colouration of words may be a crucial aspect of the
poetic form. Such features are usually divided into vowel correspondences
(assonance) and more general correspondences usually involving
consonants as well. From a sonic art point of view this distinction is either
invalid (both are to do with timbral correspondences) or too narrow (there
are many more than two classes of timbral objects). The correspondences
between phonemic objects have a very long history in poetry, mainly in the
form of rhyme. Some poets, such as Gerard Manley Hopkins, have placed
particular stress upon this aspect of their writing and, in the book Phonetic
Music with Electronic Music, Robson (1981) has developed a general
theory of vowel harmony. If we free these considerations from linkage to a
linear solo text, then poetry (even poetry deeply based in semantic
meaning) and a choral or electroacoustic art of timbral articulation meet, as
pitch and harmonic field can be articulated independently of all these
parameters. A vast new area of sonic art opens up before us which has
previously been bypassed by the linguistic or pitch-lattice preoccupations of
poets and musicians respectively.

Linguistic flow

In addition to the intrinsic properties of the timbral objects of a language-

stream, the juxtaposition of particular kinds of objects determines a
particular kind of ‘flow’. Thus, the passage of one vowel to another need
involve no alteration in the rate at which air is expelled from the lungs. A
text which articulates vowels only can therefore be entirely smooth in its
flow (apart from where the speaker needs to breath). Certain consonants
also allow the air stream to continue, restricting or varying it only slightly,
thus m and n momentarily divert the air stream through the nasal passages
(only). W and I partially constrict the air flow for an instant whilst v, z and
others do likewise but the constriction is slightly more intense, introducing
noise turbulence momentarily into the air stream. Text may therefore be
constructed exhibiting different properties of flow. Schwitters’ “Tee tee tee
tee” tends to invoke a set of air pulses, whilst “Bumm bimbimm bamm
bimbimm” produces a kind of sawtooth articulation of the air flow, and
“Tilla lalla” is essentially undulatory. Let us now consider the text extracts
from Joyce's Ulysses which Berio chose for his Omaggio a Joyce1: “Deaf
bald Pat brought pad knife took up” cuts up the air flow between each word.
At the same time the use of the long vowels a in bald, ou in brought, i in
knife suggests sustained, separated units and justifies to some extent Berio's
attribution ‘martellato’ to this phrase (Berio 1959). “Chips, picking chips”
is similarly discontinuous, but in this case all the vowels are short,
suggesting the attribution ‘staccato’. Finally, “A sail! a veil awave upon the
waves” has in general the undulatory character of Schwitters’ “Tilla lalla”
which we might link by analogy with pitch portamento (see Figure 14.4).
 Figure 14.4 Linguistic (or aerodynamic) flow in Schwitters and Joyce.

In some lettrist poems an irregular staccato flow of air is implied by

the stringing together of many consonants as in Improvisations by
Jacqueline Tarkieltaub:
jtrsrndijvakaia. rdnstrklmndrnchtkrissvrichk!
akrt! akrt! kh... kh...srk!
rdinsrikarkdineirksrinirchvrstmnskrdrsismanris!
fmjrkstnrsnichkezriksrmnsrnrguitrnsnierch!
(quoted in Curtay 1974: 232)

We might also point to certain general pitch characteristics of utterances at

this stage: the sustained monotone with sudden articulations of the
American tobacco auctioneer's spiel or the gradual rise in pitch of a
horserace commentary as the race nears its end (the siren effect) spring to
mind. We will leave a more detailed consideration of pitch articulation to
the section on paralanguage.
Semantics and cultural landscape

We are already approaching very close to the sphere of semantics which

forms a universe of discourse separated from sonic art by the concept of the
arbitrariness of the sign. For the text-sound-artist the link with linguistic
semantics is a binding thread, even where it is negated. Sonic art
perspectives will often be used as a foil to illuminate semantic content.
There are numerous subtle ways in which semantic and sonic aspects of the
language-stream may interact in a work of art. I wish to give a schematic
outline of these here, not because I do not consider these things to be
interesting, but because I have to draw the limits of what I write about at
some point or I would be writing for ever!
First of all we should note that individual human languages, dialects,
accents, cultural styles (for example, those indicative of class), special
cultural forms (the archaic in religious observances, poetic forms,
exultatory and ecstatic language), all have particular sonic aspects as well
as cultural reference. Such features may be brought together and organised
(for example, on tape) in terms of their (cultural) landscape properties, in
terms of some kind of supra-linguistic scheme of reference or purely for
their sonic properties (or of course all three or any degree or combination of
these three, simultaneously).
The degree to which we recognise the semantic content of language
will also alter the focus of our perception upon the language stream. Heard
in ‘normal’ contexts, our native language will appear heavily semantic. A
language which we know, but not well, will be heard in both semantic and
timbral focus, whilst a language which we cannot speak or understand at all
will be a largely timbral experience although, of course, mouth symbols and
paralinguistic aspects of utterance may have some semantic impact upon us.
In terms of written text, we may accept extensions of vocabulary and
even grammar as semantic:
Spat—he mat and tried & trickered on the step and oostepped and peppered it a bit with long
mouth sizzle reaching for the thirsts of Azmec Parterial alk-lips to mox and bramajambi
babac up the Moon Citlapol—settle la tettle la pottle, la lune—Some kind of—Bong!
(Jack Kerouac Old Angel Midnight quoted in Kostelanetz 1980: 18–19)

Mike walked in on the: attense of Chjazzus as they sittith softily sipping sweet okaykes H-
flowered purrhushing ‘eir goofhearty offan-on-beats, holding moisturize’-palmy sticks clad in
 clamp dresses of tissue d'arab, drinks in actionem fellandi promoting protolingamations e
state of nascendi; completimented go!scene of hifibrow'n [...]
(Hans G. Helms Fa:m’ Aniesgwow quoted in Kostelanetz 1980: 20–21)

Using standard conventions of grammar we may invent entirely new words

which however appear to have quite clear semantic connotations:
’Twas brillig and the slithy toves
Did gyre and gimble in the wabe:
All mimsy were the borogoves
And the mome raths outgrabe.”
(Lewis Carroll, Jabberwocky (1872) (Carroll 1994: 28))

Although almost half the words in this text are imaginary, it is quite clear
from the context that brillig is a time of day,2 a season or special occasion
and that toves are some kind of creature. Slithy is interestingly constructed
out of mouth symbols (relatable to both slimy and slither) whilst gyre we
might relate semantically to the word gyrate.
In the computer-generated mock-Latin quotation quoted in Chapter 9,
there are almost no real words but a non-speaker will accept the ‘Latinness’
of this text as the frequency of occurrence of certain phonemes and their
typical ordering is reminiscent of Latin. In the poem of Hugo Ball quoted at
the beginning of the previous chapter, although no known natural language
is being used, we accept the utterance as linguistic because phonemes are
organised into polysyllables and their statistical distribution has some
similarities with typical languages. In the Schwitters Ursonata, however,
the typical statistical distribution of phonemes in language is contradicted.
We are beginning to enter into pure sonic art. Finally, the
‘rdnstrklmndrnchtkrissvrichk!’ utterance of lettrism severs our link with the
phoneme and we approach the pure percussive sound-object ‘tjak’ of the
rhythmic Ketchac monkey-chant of Bali.

Paralanguage of pitch and stress

The pitch and loudness of a natural speech act tend to remain quite close to
a mean value. Around this value both pitch and loudness are articulated.
Some of these articulations are conventional and contribute or qualify the
meaning of the linguistic stream. In certain languages (such as Chinese) the
perception of pitch is integral to the recognition of a particular phoneme. A
phoneme otherwise having the same timbral characteristic will have a
different meaning according to whether it is spoken in a low, medium or
high tessitura or with a rising pitch contour or a falling pitch contour. Pitch
characteristics which are integral to the recognition of phonemes are known
as tonemes. Pitch is also used in a less specific way to indicate the end of
sentences (usually rising in French but falling in English), to indicate
questions (rising at the end of the sentence in English) and so on.
Stress, which may be articulated by loudness difference or pitch
difference, is also used as a conventional aspect of the language stream.
Thus, in English, four levels of stress may be recognised: primary (1),
secondary (2), tertiary (3), weak (4). For example: hot food (2, 1); hotter (1,
4); hotel (3, 1); contents (1, 3); operate (1, 4, 3); operation (1, 4, 3, 4).
Stress is also distributed in a semantically significant way within phrases.
Compare the difference in meaning of don't do that with the stress patterns
(1, 4, 4), (2, 1, 4), (2, 4, 1), (4, 1, 2) or (4, 4, 1).
Such aspects of conventional paralinguistic pitch and stress cannot be
completely divorced from even conventional musical practice. Thus, for
example, classical Chinese word-setting must concern itself with the level
or inflection of the tonemes. Certain languages tend to place the principal
stress on the first syllable (and not on the second). Musically we might say
that the language-stream lacks an anacrusis. This is most true of Finnish and
Hungarian and the characteristic stress pattern of Hungarian carries through
the folk music into the work of Bartok (the characteristic falling leap with
the stress on the first note).
Beyond these features we may be able to distinguish certain overall
characteristics of pitch (or stress) which are deviant from the normal mean
value. The general range may be over-high or over-low or the range through
which pitch is articulated may be over-wide. This may be a conventional
aspect of accent as in the perceived ‘sing-song’ of Welsh English. Beyond a
certain limit, however, an unusual tessitura or range of pitch-articulation
will have more universal gestural significance. For example, the expression
of anger is usually associated with high pitch and loudness, whilst low
pitch, quietness (and possibly breathiness and smoothing of the air flow)
may be associated with intimacy.
The melodic implications of intonation patterns may, in fact, be quite
sophisticated. Istvan Anhalt (1984: 159) cites Fonagy and Magdics (1963)
as having described in musical notation “what they perceive to be the
melodic patterns of certain emotions or emotional attitudes: joy, tenderness,
longing, coquetry, surprise, fear etc.”. Even the simple question-pitch-
inflection is, in fact, much more complicated and Kingdon (1958) has
distinguished between “general, particular, alternative, asking for repetition,
interrogative repetition, insistent and quizzical” intonation patterns for
questions (cited in Anhalt 1984: 159).
Such paralinguistic aspects of pitch and stress have a bearing on the
construction of vocal melody (and thence on melody in general). Thus,
where a melodic line proceeds by wide leaps, particularly where this is
outside a simple harmonic field, such as an arpeggio, and/or is associated
with a freer speech-type of rhythmic articulation, we are likely to make the
paralinguistic interpretation that the utterer is agitated, frightened or
disturbed, or at least over-emphatically expressive (the ‘Schoenberg
effect’).
The paralinguistic implications of pitch, tessitura and motion have
clearly always had a strong influence on the practice of singing (consider
for example Japanese joruri singing for the Bunraku puppet theatre, the late
sixteenth century invention of recitative in Western Europe, etc.). Even
where pitch is compositionally ordained to follow the most rigorous
instrumentally conceived pitch-lattice logic, interpretation tends to input the
paralinguistic gestures in the form of small articulations of pitch and stress.
Schoenberg further extended these connections by his development of
Sprechstimme. A schematic analysis of pitch usage (and its combination
with timbral morphology) is indicated in Figure 14.5.
The use of pitch in any vocal utterance which is not clearly
conventional language or conventional pitch-lattice music needs careful
consideration. Curtay, in his lettrist compositions, specifies pitch in only the
four (linguistically significant) registers, low, medium, high and very high,
as he wishes to avoid any over determination of the events by the logics of
conventional music. I would consider this a much too simple view as may
be evident from the analysis above. There is, however, a technical difficulty
in integrating clearly-pitched material into a sonic structure which has
previously contained no (stable) pitches; steady-state pitches tend to ‘stick
out like a sore thumb’, producing a marked discontinuity in our perception
unless they are introduced with great subtlety. It is clear that there are all
sorts of degrees of balance between a sonic art fundamentally rooted in the
relationships of fixed pitches (to which timbral characteristics are
subservient) and a sonic art based on the sophisticated control of timbral
possibilities (to which pitch characteristics are subservient). Although we
may achieve transformations of structural organisation away from one pole
and towards the other within a single sonic composition, it is important to
be aware where the perceptual focus lies. This seems to be a problem both
for conventional musicians (for example, in assessing timbrally articulated
works based on a relatively fixed harmonic field) and for text-sound-artists
anxious to abandon linguistic reference.
 Figure 14.5 A schematic analysis of vocal pitch-usage.

Paralanguage and rhythm

A detailed discussion of the nature and implications of rhythmic

organisation is outside the scope of this book. Just as with pitch and stress,
the language stream has a characteristic average tempo and rhythm. We
may guess that the typical tempo is in some way related to the speed of our
ability to decode the signals and thence to our typical metabolic rate (it
would appear that small birds have a higher metabolic rate, judging by the
speed of articulation of their utterances which for us may congeal into
timbral entities). Variations of tempo around this may be indicative of
physiological state or mental attitude, or of cultural factors such as in the
slow tempo of English Forest of Dean and American Alabama accents.
Each rhythm may be described as regularly irregular. In a normal
speech-stream all syllables fall within a small range of possible durations.
Within this range they vary according to conventional stress patterns,
idiosyncratic or special semantic use of sustain-stress, or, within limits,
randomly.
The tempo and rhythm of speech is undoubtedly one of the archetypal
models for our physiological-intellectual interpretation of the meaning of
rhythm, particularly irregular rhythm. It differs from the typical bipedal (or
more generally bilateral, as with, for example, the heart) physiological
rhythms of the body, in that it has no underlying regular pulse, but only an
average rate of flow. Rhythmic fluctuations beyond a certain point (just as
with extreme slow or fast tempi with bipedal-type rhythms) go beyond the
scope of conventional paralinguistic interpretations and may be
‘understood’ in terms of more universal (and possibly extremal) gestural
criteria (for example, agitation, terror, melancholy).
A sonic art perspective, however, brings a more precise conception of
timing into play. Through performance practice deriving from music, or
through computer control, we can stretch and alter the rhythmic
characteristics of the language-stream so that they become, for example,
unnaturally regular, unnaturally distributed in sustain-stress pattern,
unnaturally fast or slow, unnaturally grouped in regular or evolving patterns
and so on. In this way the perception of rhythm as an aspect of
paralanguage can be extended or destroyed.
Paralanguage and morphology

As we have described previously, it is possible to use paralinguistic

indicators to differentiate regional accent, status, idiolect and even the
attitude of the speaker. Our interpretations are usually based upon the
observation of a field of timbral types and morphologies and our ability to
relate these to a known (typical) field. From a sonic art point of view we
may be interested in working with any of these aspects of the language
stream. For example we may collect, through recording, materials spoken
by human voices with particular striking idiosyncratic characteristics. This
would be akin to Fellini's technique of selecting his actors from amongst the
general public in markets etc. for the particular idiosyncratic features of
their faces or general appearance. With computer synthesis we may, in fact,
construct a voice having a particular ‘voice set’ and control (e.g. gradually
transform) these qualities, such that gender, age, health-image or body-
image (all aspects of ‘universal landscape’) change. Alternatively, from a
narrower linguistic viewpoint we might change status, regional origin or
role (for example, sergeant major parade-ground voice-set, newspaper
vendor voice-set).
We might also control the (semi-involuntary) physiological indicators
known as ‘vocal characterisers’3 and ‘vocal qualities’4 on a general or
moment-to-moment basis; vocal qualities such as pitch range, control and
rhythmic smoothness (discussed previously), vocal hoarseness, general lip-
openness, over- and under-voicing, slight or heavy breathiness, forceful or
relaxed articulation control or thin or full resonance. As our landscape- and
utterance-based interpretation of these para-linguistic features relates
holistically to multi-dimensional fields and their articulation, the
independent control over each aspect of the language stream which might
be possible using computer technology combined with the electronic
paralanguage of reverberation, echo, phasing, double-tracking, social-
landscape distancing and yet-to-be-discovered techniques, opens up a truly
mind-bending plethora of aesthetic possibilities.

1 The original text is in Joyce (1960): 328–330.

2 The Editor begged to differ, having believed all his life it was a state of the weather (Ed.).
3 For example, laughing, giggling, snickering, crying, whimpering, sobbing, yelling, muffled
whispering, muttering, moaning, groaning, whining, breaking, belching, yawning.
4 From which we might deduce attitude; for example, anger, dominance, submission, intimacy,
secrecy.
 Chapter 15
THE GROUP

Personality, society

Once we begin to consider two or more utterances occurring in the same

acoustic space, we enter into the realm of human interaction, whether it be
viewed as ‘theatre’ (the projection of imaginary personas by actors) or the
overhearing of real or imaginary situations (as might be the case in a tape
composition). At this point landscape and utterance begin to blend. As we
are dealing with the landscape of human society, we begin to pass over into
the realm of theatre and a full consideration of all the implications of
theatrical arts is beyond the scope of this book. I will merely attempt to give
some flavour of the countless ramifications this connection has for sonic
art. For anyone wishing to pursue this aspect further I would highly
recommend the book Alternative Voices by Istvan Anhalt (1984). To some
extent an utterance, particularly one which we can relate, no matter how
distantly, to the language stream, projects the whole personality (or
assumed persona) of the utterer which cannot be entirely understood in
isolation from its social context. We may often infer from it a whole web of
social situations, assumptions, meanings, attitudes, observances etc. which
are not contained in the semantic content of the stream (if any) but implicit,
pre-supposed or pointed to by it. The utterance implies the ‘other’ or ‘them’
or even implicates the listener as ‘you’ or ‘them’. In sonic art we have the
possibility of restructuring any aspect of these pointers.
In the monodic utterance, we may observe breaking, self-correction,
incompletion, repetition, distortion of various kinds, filled pauses, retraced
false-starts, stuttering, suppression, sudden interruptions, non-verbal
expressives or the gradual disintegration of language itself. From our
apprehension of the events we may relate them to internal soliloquy,
glossolalic ecstasy, childhood pre- and early linguistic sound-play,
monomania etc.. Or from the particular combination of semantics and
paralanguage we may construct a social role or context; the monologue of
the scientific address, the advertising hook line, the oratory of politics or
demagoguery.

Individual interaction

Such contextual implications are focused and/or clarified in the

communicative interaction of utterances. It is possible, even without the aid
of semantic content to develop a subtle structure of social interaction, roles,
power play, sexual encounters, group cohesion or anarchy. Ligeti's work
Nouvelles Aventures, for three solo voices and small instrumental ensemble,
does exactly this. Ligeti says of the piece —
All human affects, ritualised through forms of social intercourse, such as agreement and
dissension, dominance and submission, sincerity and lie, arrogance, disobedience, the
subtlest nuances of irony hidden behind a seeming consent, similarly high esteem that is
concealed behind seeming disdain—all these, and still many more, can be exactly expressed
through an a-semantic emotive artificial speech.
Such a text should not be expected to define precisely any conceptual relationships,
but ought to portray directly human emotions and modes of comportement in a way that
despite the conceptual meaninglessness of the text, scenic instances [Momente] and actions
could be perceived as being meaningful.
An imaginary language, which makes human feelings and comportement
communicable, must also be suitable for a work for the theatre, presuming that one regards
the theatre as a pedestal for individual and social modes of comportement, as an essence
[Konzentrat] of communication (and also that of isolation).
The point of departure for the composition was a conceptualisation of relationships
between affective modes of comportements, not an abstract structural plan. [...]
It seems significant to me that there is no ‘deeper meaning’ hidden beyond the
performed events. Despite the seeming absurdity and enigmatic character, the protagonists
and the emotional and social situations are directly intelligible, and transparent. We do not
find out what the story is about—and in a deeper sense there is, of course, no story, yet we
 quite precisely find out how the persons behaved and in what relationship towards each other
they stand. (Ligeti quoted in Anhalt 1984: 91–2)1

Clearly, percepts such as agreement, disagreement, dominance,

submissiveness, solidarity, anarchy are aspects of the social landscape.
Furthermore, attributes such as sincerity, lying, persistence or hesitation are
more easily perceivable as such in a context in which utterances are
exchanged. We may even make very fine distinctions in this context. For
example, the performing together of actions which turn out to be too
difficult (see the hocket in the Ligeti work) imply a desire for solidarity. All
these features of the social landscape may be implied merely through the
subtle control of paralinguistic signs and the relative similarity/dissimilarity,
coordination/non-coordination, loudness etc. of the individual utterances.
Back-tracking slightly we may regard this as a somewhat more
refined articulation of what can already be observed in the communication
of higher animals. There is an (understandable) fear of anthropomorphism
in zoological studies which tends to veil the nature of communications in a
kind of neutralist vocabulary. Translating from the various studies that have
been done, we can discover a wide range of pre-semantic messages; Figure
15.1 illustrates some of these (many of which can, of course, be combined).
With semantics human beings have expanded many of these areas into
informal (courtship) and formal (political contest) rituals.
Within such landscapes of communication, we may observe or
establish particular roles for the individual utterers (‘initiator of actions’,
‘reinforcer of cohesion’, ‘disruptive influence’ etc.) such roles may change,
extend or develop. As all these complex interactions are, however, taking
place in the field of sonic articulation, we can imagine extending this
concept of role behaviour to the articulation of complex abstract sonic
objects, thus developing an articulate sphere of musical discourse based on
the logic of the landscape of utterance interaction but in which no utterers
are present. From here we may go on to imagine landscapes in which both
utterers and non-utterers appear to adopt roles relative to one another and
communicate (in a sense the orchestra in many operas may be thought of as
having a role, being a character within the ongoing drama). We may also
imagine an interface between the natural morphology of sound-objects and
the landscape of social utterance interaction (with or without utterers, or
even recognisable sources).
 A particular aspect of social interaction is, in fact, the failure or non-
existence of communication. a sense of non-communicating co-existence
can be enhanced in the virtual space of loudspeakers by the use of location
cues (spatial position, reverberation etc.) and other devices to suggest the
co-existence of two separate ‘worlds’. Again, these concepts can be
generalised to non-utterer, non-concrete situations; a sense of separate and
non-interacting sonic streams may be established and then, for example,
negated by the subsequent interaction of the materials. Such social
‘readings’ are particularly likely to affect our aesthetic perception where we
use utterances in the sonic landscape.

Chorusing, co-ordination and mass

Amongst the various types of animal interaction discussed above, there is

one particularly interesting group which I have labelled ‘US’. These are
signals made in a collective but uncoordinated manner by various members
of the group which seem to have the function of binding the group together.
They are not individual expressions of relationship or mutuality between
individual creatures, nor are they signals to the group about situation or
circumstances, external threat or internal disobedience. Instead, they
perform a quasi-ritualistic function. They express solidarity and may at the
same time express a general group mood (e.g. excitement).
 Figure 15.1 Some pre-semantic messages of higher animals rendered in ‘human’.

This solidarity chorusing behaviour can be observed in many species,

from crickets and frogs to the group howl of wolves. A similar comment
can be made about some of the one-to-one communications between
individuals of equal status. As the call of any particular individual in these
situations is, we might say, syntactically-semantically neutral (it does not
refer to any specific or immediate cause of interest or anxiety in the
environment or within the individual's ‘psyche’) it may become the element
of a group syntax or semantics. In this way group calls which, as far as we
can tell, have the gross function of affirming the solidarity (or even mere
similarity) of the group may develop a degree of internal structure and co-
ordination.
Thus, choruses of the spring peeper frog (hyla crucifer) have a
characteristic structure with each chorus arising in the same manner time
after time.
Typically [...] three frogs sing as a group and a large chorus seems to be nothing more than a
number of these trios calling from the same breeding site. Furthermore, each of these trios
develops in the same manner. The call is initiated by a single individual sounding the note of
A for a varying number of times. After a brief rest, if he does not have an answer, he gives a
trill. This trill apparently acts as a stimulus since it usually results in another individual
starting to call on the note G#. When this happens the two individuals continue giving their
respective notes—A, G#, A, G# etc.—for an indefinite number of times. If a third individual
does not start calling, they stop their alternating calls, rest, and one of them—usually (and
perhaps invariably) the one that is calling G#—gives the trill. At the sound of the trill, the
third individual of the trio starts giving his call which is B. Thereafter, the three continue to
call, each giving its respective note in the order indicated, A, G#, B, for an indefinite number
of times.
(Goin, (orig. 1949) quoted in Bogert 1960:209)

It is interesting to note that this ‘group syntax’ has already evolved in such
a lowly creature. Furthermore, the normal functionalist approach to the
explanation of all animal communications (for example, in terms of sexual
bonding, status etc.) are difficult to apply when three individuals are
involved! The bou-bou shrikes of East Africa —
[...] can sing in duet with such a rapid reaction time that unless an observer is actually
standing between the two birds it is impossible to recognise that more than one bird is
singing. [...] a pair of birds can elaborate a whole repertoire of duet patterns by which they
can recognise one another in dense undergrowth and be distinguished from other pairs in the
neighbourhood. In this species either sex can start or finish and either bird can sing the
whole pattern alone in the absence of the partner. When the partner returns, the pair can
either sing in perfect unison or sing antiphonally again. Trio singing has also been observed
[...]
(Hooker 1968: 333–334) (See Figure 15.22).
Figure 15.2 Examples of duet patterns of various African bou-bou shrikes where X and Y represent
the two birds in the pair (after Thorpe and North (1965)).

Here, then, whatever the function or ‘meaning’ of the song, its syntax can
be articulated by two or three creatures acting ‘in concert’. In the music-
making of groups of humans, this mutual solidarity function is implicit. It
may be so distanced (within the structure of a larger society) that it is not
immediately perceived as a function of the musical activity by the
participants. Alternatively, the musical act may function specifically in that
role (ritual function of music in group ceremonies). Because, however,
many more levels of distancing are involved in human social
communication and interaction, we may represent within the convention of
group music-making the concept of disorder and strife as well as various
sophistically-differentiated conceptions of social cohesion.
Renaissance imitative vocal polyphony, for example, presents a
particular archetype of the relationship between the individual and the
group in the way in which the similar, but different, vocal lines are
harmonically co-ordinated. The sense of balance and equality within
harmony is quite distinct in its symbolic representation of the group from
Bach's Kyrie in the Mass in B minor. Here a sense of co-ordination of
utterance and planned development towards a final resolution is articulated
over a framework of highly affective dissonance. The social metaphor is
quite different.
And both of these differ quite markedly from choral works in which
the voices are rhythmically (and perhaps harmonically) co-ordinated in their
utterance, so we perceive only the group, or in which such groups are set off
in antiphonal relationship to each other. The organisation of the group may
point to specific group roles or functions within society, such as the
simulated collective meditation by a group of individuals in Stockhausen's
Stimmung or the quasi-religious action by a large group suggested in the
Introitus of Ligeti's Requiem.
At the other extreme, particular types of rhythmic disco-ordination
may present the group as a ‘sea of humanity’, a multitude or mob being
manipulated or out of control. Even this social image has subtle
ramifications. In his book, Anhalt (1984) gives some interesting insights
into this. Thus, in the third movement of Lutoslawski's Trois Poèmes
d'Henri Michaux, the composer uses a ‘fan’ effect where the chorus moves
from rhythmic synchronisation to a swarm effect gradually. Anhalt
describes this as “an allusion to the individual will, which seems to prevail
over that of the collective [...]” (Anhalt 1984: 137). In the second
movement, however, we experience ‘raw force’, ‘aggression’, ‘mob
behaviour’, the shouts of a crowd, either semi-concerted or synchronised
like the synchrony and asynchrony of the crowd cries at a great fight.
Alternatively:
The ‘Kyrie’ of Ligeti's Requiem is a powerful showing of a mass of human beings, swirling
and twisting in so many vocal currents, adding up to a turbulent sea of voices in which the
identity of an individual is painfully and irrevocably submerged on account of the number of
concurrently used similar melodic designs and overlapping registers. The canonic structures
here have a ‘blind leading the blind’ character, conveying the cumulative affect of a hopeless
predicament for the whole mass; [...]
(Anhalt 1984:2003)

With electro-acoustic projection of vocal sounds (microphones to

loudspeakers), it is possible to establish imaginary images of the group.
Thus, in Vox-I, one hears first of all a single undifferentiated timbre-stream;
through gradual processes of division one begins to separate off the
‘environment’ from the ‘utterance’ and then, within the utterance, two and
then four beings gradually emerge. Through various combinations of stream
mixing and spatial motion these four beings may recombine or dissociate.
Towards the end of the piece the four beings sing in unison (as opposed to
being combined into a single-stream utterance) and the landscape itself
takes on the articulation of speech.
Moving on to more sophisticated levels of social distancing and
human group behaviour, the situation becomes increasingly complex and
ramified and may include within itself subsidiary forms of group behaviour,
such as music-making in religious observances, rallies or art events. Social
landscape and utterance become intertwined in a multi-level sonic reality.
We may begin to think of parody, incompetent performance, stylistic
transformation and countless other levels of social signification. In
universal sonic art freed of narrow preconceptions concerning the
boundaries of artistic discourse, we may focus or defocus, manipulate,
confuse or extend any or all of these levels of perception.

1 Anhalt is here quoting from two articles originally in German. The English translation is
presumably his (Ed.).
2 After Thorpe and North, 1965 quoted in Hooker 1968: 334.
3 ‘Affect’ sic in last sentence of quote (Ed.).
Coda
 Chapter 16
BEYOND THE INSTRUMENT: SOUND-
MODELS

The universal instrument

As stated at the beginning of this book, the theories and speculations it

contains have been prompted by the development of the computer and its
application to the field of music. As in all other fields, the computer can
change our entire perspective on the way we do things because it is not a
machine designed for a specific task but a tool which may be fashioned to
fulfil any task which we can clearly specify. It is a metamachine. In
particular, it offers the possibility of being a universal sonic instrument, a
device with which we can model and produce any conceivable sound-object
or organisation of sounds.
The limitations on its potential are partly due to our lack of acoustic
understanding but primarily to our perceptual and aesthetic preconceptions
about the nature of sonic art and its instruments. At its most elementary
level these limitations could be seen in the small, digital packages available
for music-making in the early 1980s. Even the fairly versatile Fairlight CMI
had built into it the assumptions of instrumental fixity (‘an instrument is a
source of fixed timbre’), instrumental streaming (music is made up of
streams of fixed timbre) and intervallic uniformity in scale structures. On
smaller machines the software tended to make so many assumptions about
what constituted a ‘musical’ sound-object that it was often easier to
generate interesting sounds through ‘misusing’ the technology (for
example, through aliasing effects).
There is, however, a deeper level at which most of the early 1980s
digital technology was crucially limited. Its definition of an ‘instrument’ did
not correspond with our perceptual reality. As listeners in the real world we
do not tend to respond to each individual sound-object as an intrinsic entity
and build up a Fourier analytic picture of it. We relate sound-objects to
models and we may group many different sound-objects as examples of a
particular model. Thus, for example, the spectrum (and its evolution) of a
bass piano note is noticeably different from that of a high piano note—as
sound-objects they are remarkably different yet we relate both of them to
the sound-model ‘piano’ quite directly. The notion of sound-model
corresponds very closely to the notion of intrinsic morphology put forward
in earlier chapters.
The problem with both typical additive synthesis and FM synthesis
applications is that they assume the sound-model is at the same level as the
sound-object. Hence, once a particular sound-object has been built up by
these techniques it can be reproduced at all levels of pitch and intensity to
generate the percept of an instrument. In this sense, the technology is
deeply influenced by the ideology of music springing from Western
notation where timbre has been regarded as a secondary and fixed quality
over which pitch and duration are articulated. Although, of course, we must
have a physical understanding and representation of any sound in Fourier
analytic terms to be able to achieve synthesis, this does not correspond
directly to our perceptual processes.

From instruments to sound-models

The distinction between object and model may be understood more easily
when we move into the field of sound-sources possessing a repertoire, such
as the human voice. In this case the definition of one particular additive
synthesis or FM spectral type is obviously totally useless (except, of course,
if we wish to specify a different such type for every single vocal event in
the stream). What we can, however, specify are various invariants which
occur in the sound-source, for example, a description of the particular
placing and spread of the formants of a particular voice and also typical
articulation structures (the vocal attack-time, transition phenomena between
formants and so on) and the general spectral typology (e.g. the shape of the
individual glottal impulse).
All this, in fact, applies to standard musical instruments. As a simple
example, it is already well-known that the spectral richness of a piano tone
decreases with increasing frequency. This is a particularly simple law, but
we might also specify the ways in which spectral envelope, pitch, jitter and
amplitude envelope and fluctuations are correlated during various bowing
actions.
This specification of (near-) invariants over a field of possible sounds
is what I mean by a sound-model. It is a more general notion than the
typical limited view suggested by the concept of ‘musical instrument’. For
example, we might model the rules which govern the relationship between
spectral change, pitch change and loudness change in a metal sheet which is
being flexed. Together with other invariants this will effectively specify the
intrinsic morphology of the sound-model and the set of rules will govern
the behaviour of the sound-model when we articulate it through some input
device (which might be part of a program or a direct physiological-
intellectual input; see below).
With the computer as sound-source, however, we are not confined to
basing our sound-models on existing physical objects or systems. We may
build a model of a technologically (or even physically) impossible object.
We might specify the characteristics of the voice of an imaginary creature.
Once, however, the sound-model is specified, we are free to change the
invariants of its behaviour. We may transform it into an entirely different
sound-model.
The crucial difference between building sound-models and building
sound-objects is that the former preserves a clear and perceptually relevant
distinction between intrinsic and imposed morphology. If we articulate the
object within the rules specifying its invariants, we perceive an imposed
articulation of the sound-model. If, however, we proceed with some process
which changes those invariants, then we actually perceive the sound-model
itself to change. The intrinsic morphology changes, the perceived source
becomes ‘something else’. This is crucial to our understanding of the
perception of typical analogue synthesiser sounds. Largely because
instrument-definitions on such synthesisers are based on sound-objects and
not on sound-models, then no matter how we transform the sound-material,
we tend to perceive it as coming from a synthesiser. It was not just a lack of
detailing in the modelling of individual spectra in the voltage control
synthesiser that made its sound-world characteristically ‘synthesiser’ but
the more general lack of structuring in relation to perceptual sound-models.
Some of these concerns may be illustrated with reference to the digital
synthesis language Chant. Here the language makes a broad specification of
the (semi-) invariants of the sound-model ‘human voice’, for example, the
field of formant bands and their characteristics, typical values of vibrato,
vibrato variations, jitter and so on. Composition with this language may be
multilevelled. At a gross level we may specify merely pitch, loudness,
vowel type (a, e etc.), variations in type of vibrato, and so on. At a deeper
level we may specify particular modifications of the formant bands (for
example, to characterise a particular idiolect) with which we then compose.
At yet another level we may wish to impose transformations on the formant
bandwidths and the attack structures of individual events. This begins to
interfere with the invariants and rules which govern the sound-model
‘human voice’ and by doing so we can generate sounds of bells, drums and
so on. Defining classes of sound-objects at the level of sound-models,
therefore, has a direct relationship with our perceptual categorisation of
sound-events. Change the invariants and rules and one changes the
perceived model.
At the time of writing (1983), Chant does not model most of the
consonants. The modelling of such structures of articulation will be a key
development in the evolution of the digital computer as a powerful tool for
sonic art. Moreover, it should lead on to developing a modelling system for
natural processes themselves. Perhaps some rigour may be brought to this
through insights and mathematical techniques from differential topology.
This would give a sound theoretical basis to the concept of natural
morphology discussed in this book and allow us to have handles on the
evolution of a sound process that corresponds to the critical parameters of
the flow.
Given such powerful modelling systems, we may bring an imposed
morphology to bear upon the sound-models through some kind of real-time
or programmed input. The imposed fluctuations may then be made to
articulate what remains a ‘solid’ (relatively stable mass) object (by
changing the overall pitch level, loudness, loudness envelope, vibrato,
vibrato width, vibrato steadiness, tremolo width and steadiness, jitter, etc.
within certain limits) to ‘liquefy’ that object or a stream of objects (by
articulating the spectral envelope and pitch contour, possibly in relation to
one another, the typical event duration and density, the spread of pitch and
so on), or merely to interact with its ‘gaseous flow’. In the latter case I am
not thinking of simple vocoder-type processes but some way of mapping
bodily or vocal gestures into the flow properties of a sound (such as speed,
density, turbulence etc.).

Imitation, transformation

In computer modelling there is a trade-off between flexibility and speed. If

we are prepared to work within the constraints of a particular sound-model,
we can generate results fairly quickly. Once, however, we decide to stretch
or modify the basic models, we need to input much more information to the
system and its musical productivity slows down. Another approach to the
problem might be to begin with an existing real-world sound-object (such
as the sound of our own voice) which we are able to experiment with and
refine in real time before making a recording. Then, using powerful tools of
analysis (particularly linear predictive coding or phase vocoder techniques),
we may extract the perceptually distinct features of the recorded sound-
object. Thus the package of analysis and resynthesis programs available at
Stanford and IRCAM from the late 1970s (of which Andy Moorer was one
of the main authors) made it possible not only to achieve this analysis but
then to manipulate the parameters individually and to cross-synthesise. For
example, once the pitch-information of a voice has been separated from the
formant information it is possible to reconstitute the voice using different
pitch information (or different formant information). Alternatively the pitch
and/or formant information may be stretched or compressed in time. In fact,
the dilating of time structure of complex sonic events such as occur in the
speech stream or in ‘fluid’ sound-objects, gives us access to the detailed
behaviour of sonic processes; a kind of sonic magnification. Using the
programming power of the computer we may then reconstitute these objects
with entirely different time-flow characteristics.
Finally, we may transfer the formant characteristics of one source
onto another, the technique of cross-synthesis. For example, imposing the
formant characteristics of a speech stream upon a bubbling fluid stream
creates the perceptual model of a voice containing a stream. Alternatively,
imposing the formant characteristics of the fluid stream upon the voice
generates the sound-model of a stream in which a voice is contained. We
might now perhaps take a voice generated by Chant which changes into a
drum and then cross-synthesise this with the sound of a crowd. And, of
course, the sources for cross-synthesis may themselves be imaginary sound-
models.

Operational fields

At a larger-scale level the computer may allow us to analyse the detailed

structure of large-scale events such as the behaviour of flocks of birds
during an ‘alarum’, the detailed structure of semi-synchronous crowd
chants, the textural structure of poured pebbles and so on. With or without
this analysis, we can control in great detail the parameters of texture, from
the standard harmonic and motivic field to timbral field, pitch-motion-type
field etc. and subtle changes of all of these.
More significantly, however, we may group particular parameters and
types of articulation into fields governed by rules. These operational fields
may themselves be articulated through other inputs (physiological-
intellectual performance behaviour or higher level rules) given by the
composer. This grouping must be a very generalised facility so that the
sonic artist can choose his field of focus upon the sound-materials. This
might be in terms of harmonic fields, timbral fields, articulation types,
paralinguistic fields, idiolectic features, spatial cues, linguistic
grammaticality, natural morphological properties, sonic role-plays etc..
Only at this stage will the computer become a truly generalised tool for
sonic art.

Human interfacing

The other crucial feature in the application of digital technology to sonic art
is the development of sophisticated hardware and software tools to permit
human physiological-intellectual performance behaviour to be transmitted
as imposed morphology to the sound-models existing within the computer.
This connects directly with the whole area of the evolution of musical
instruments in their more conventional sense. We may anticipate in the not-
too-distant future the development of a whole generation of digital sound-
producing devices which are, to a greater or lesser extent, analogous with
existing musical instruments (in fact some are already here). The keyboard
synthesiser has been with us for a long time but now we are beginning to
see the emergence of blown and bowed synthesisers which present
themselves to the performer as analogues of conventional mechanical
instruments but in which the sound-production is entirely electronic. The
immediate advantage of this development is that we may, in fact, select the
timbre that the instrument produces by varying the program. More
significantly, perhaps, we may alter the way in which our various
articulations of the string, air column etc. appear as articulations of the
sound.
The concept of the transfer of parameters was already well-developed
in the field of live electronics using analogue synthesisers. Thus, the arm
motion speed, breath flow or, more typically, the resultant loudness
variation, pitch variation and so on, could be monitored by various
electronic devices (such as envelope followers, pitch-to-voltage converters
etc.) generating a voltage proportional in some way to the magnitude of the
input. The resultant voltages could then however be used to control any
desired feature of the resulting sound. Amplitude might control pitch and
pitch control amplitude. Both might control the parameters of a second
instrument or more complex features of an evolving electronic sound-
stream.
The analytic power of the computer at least potentially gives us the
ability to monitor in several simultaneous dimensions the subtle details of
performance behaviour. A sufficiently intelligent and fast machine should
be able to sense parameters of breath flow, formant structure, glottal, tongue
and lip vibrations, noise turbulence type and so on, separating these out so
that the information from each can be applied to the control of different
parameters of a sound-event. This sound-event may, of course, have nothing
whatever in common with the characteristics of behaviour which generate
the control information. Through practice, just as with the conventional
instrument, we can imagine the performer developing a sophisticated co-
ordination between his or her performance skills and the sonic output.
The design of sophisticated inputs has been one of the major
weaknesses of the digital instrument revolution but with the generality of
the computer there seems no reason why a whole range of multi-
dimensionally sensitive input devices should not be developed. These might
involve keyboards which were sensitive not only to which key was pressed
but also to finger velocity and/or pressure and to lateral motion of the
fingers (as in the analogue Buchla synthesisers) and these might be made
much more interesting than the more common approaches have been, by
adding devices allowing continuous contraction or expansion of the average
interval size, the warping of intervallic uniformity (perhaps in a pre-
programmed fashion), spectral change with register, linking attack velocity
to, for example, timbral stability and duration rather than to loudness and so
on. Bowed interfaces would be sensitive not only to pressure but also to
speed of bowing, the width of bow to touch the string, the sul tasto-sul
ponticello dimension, the temporal fluctuation of these things and so on.
Even quadrapan and pedal devices might be redesigned. We might, for
example, imagine a console with two quadrapan units which were also
capable of moving in the up-down direction and two foot pedals which
could be moved not only up and down but also from side to side and
backwards-forwards. We might even use totally novel ways of inputting
physiological-intellectual information, such as the monitoring of the many
dimensions of facial gesture.
What, however, is clear above all else is that the internal architecture
of sounds becomes both analytically and conceptually accessible and hence
available for more or less precisely defined composition and, as our ability
to monitor the subtleties of human intellectual-physiological gesture and
transfer them onto sound-materials increases, our notion of what ‘music’ is
must become much more generalised. It must embrace and systematically
investigate areas that have traditionally been regarded as the legitimate
property of psycho-acousticians, phoneticians, poets and sound-poets, of
nature recordists and audio-zoologists, of naturalistic and ‘effects’-based
film-sound engineering and much more. Musicians will concern themselves
with the affective and systematic ordering of timbre structure, sonic gesture,
sound-landscape, the subtleties of psycho-linguistic and psycho-social cues
and many other dimensions of the sound-universe, alongside the more
traditional parameters of pitch and duration. The era of a new and more
universal sonic art is only just beginning.

Postscript
It is clear that we are about to see a radical change in the nature of our
civilisation. The impact of the computer, the universal metamachine, could,
in a short time, destroy the whole basis for the work-ethic upon which most
of our present-day materialist culture is built. We can expect a rough ride
into this new world from the guardians of social orders which are no longer
relevant but when we finally arrive we may at last find the arts playing a
central role in the lives of the community in general, provided, that is, we
do not manage to commit geno-suicide in the meantime.
The possibilities opened up for musical (and all other types of art)
exploration are truly staggering. It is as if a magical dream has come true.
We have the potential to make real any sound-event we can imagine. What
will prevent us from getting to grips with this new situation is primarily our
aesthetic preconceptions and lack of sensitivity. The effect of the former is
obvious and has always been with us; the lack of sensitivity, however, may
prove to be the most debilitating. The question is simply, if one can do
absolutely anything, what precisely is worth doing? If it is not to be judged
in terms of the pre-existing criteria of available musics, we must have
enough personal musical integrity to admit that there is a distinction
between the arbitrary manipulation of materials according to some
preconceived plan and the construction or performance of valid sonic
experiences. I hope this book might open up some new pathways without
leading us into the sterile wasteland of formalism.
 BIBLIOGRAPHY

Anhalt, I.
(1984) Alternative Voices: Essays on contemporary vocal and choral
composition (Toronto, University of Toronto Press).

Ball, H.
(1974) Flight Out of Time: A Dada Diary (Elderfield, J., editor; Raimes, A.,
translator) (New York, Viking Press).

Barthes, R.
(1977) ‘The Grain of the Voice’. In Image-Music-Text (Heath, S., translator)
pp. 179–189 (London, Fontana/Collins).

Bastian, J.
(1968) ‘Psychological perspectives’. In Animal Communication (Sebeok, T.
A., editor) pp. 572–591 (Bloomington, Indiana University Press).

Bateson, G.
(1968) ‘Redundancy and Coding’. In Animal Communication (Sebeok, T.
A., editor) pp. 614–626 (Bloomington, Indiana University Press).

Berio, L.
(1959) ‘Poesia e musica—un’ esperienza”. Incontri Musicali 3 (1959): pp.
98ff also Contrechamps 1 (1983) pp. 24–35 (French) and (extract in
English) sleeve note to LP recording of Omaggio a Joyce (Vox Turnabout
TV34177).
(1967) Visage: sleeve note to LP recording Vox Turnabout 34046S

Bogert, C. M.
(1960) ‘The Influence Of Sound On The Behavior Of Amphibians and
Reptiles’. In Animal Sounds and Communication (Lanyon, W. E. and
Tavolga, W. N., editors) pp. 137-320 (Washington, American Institute of
Biological Sciences).

Boulez, P.
(1971) Boulez on Music Today (London, Faber and Faber).

Busnel, R.-G.
(1968) ‘Acoustic Communication’. In Animal Communication (Sebeok, T.
A., editor) pp. 127–153 (Bloomington, Indiana University Press).

Carroll, L.
(1994) Through the Looking Glass (London, Penguin Books).

Chowning, J.
(1971) ‘The Simulation of Moving Sound Sources’. JAES 19 pp. 2–6
(reprinted in CMJ 1(3) pp. 48–52).

Curtay, J.-P.
(1974) La poésie lettriste (Paris, Seghers).
(1981) Body Music 1 (Cassette with sleeve notes) (London, Audio Arts
Editions).
(1983) Lettrism, abstract poetry, mouth symbols and more ... (Unpublished
manuscript).

Darwin, C.
(1965) The Expression of the Emotions in Man and Animals (University of
Chicago Press).

Emmerson, S.
(1976) ‘Luciano Berio talks to Simon Emmerson’. Music and Musicians
(London) (May 1976) pp. 24–26.
(1982) Analysis and the Composition of Electro-Acoustic Music (London,
City University (PhD thesis): University Microfilms International).

Erickson, R.
(1975) Sound Structure in Music (Berkeley, University of California Press).

Fónagy, I. and Magdics, K.

(1963) ‘Emotional Patterns in Intonation and Music’. In Festgabe für Otto
von Essen (Berlin, Akademie Verlag) pp. 293–326.

Goody, J. and Watt, I.

(1963) ‘The Consequences of Literacy’. In Comparitive Studies in Society
and History 5.

Hall, D.
(1992) Klee (London: Phaidon Press).

Hayes, B.
(1983) ‘A progress report on the fine art of turning literature into drivel’.
Scientific American 249(5) pp. 16–24.

Helmholtz, H.
(1954) On the Sensations of Tone as a Physiological Basis for the Theory of
Music (New York, Dover).

Hiller, L.
(1981) ‘Composing with Computers a Progress Report’. CMJ 5(4) pp. 7–
21.

Hopkins, G. M.
(1959) The Journals and Papers of Gerard Manley Hopkins (House, H. and
Storey, G., editors.) (Oxford University Press).

Hooker, B. I.
(1968) ‘Birds’. In Animal Communication (Sebeok, T. A., editor) pp. 311–
337. (Bloomington, Indiana University Press).

Isou, I.
(1947) Introduction a une nouvelle poésie et une nouvelle musique (Paris,
Gallimard).

Joyce, J.
(1939) Finnegans Wake (London, Faber and Faber).
(1960) Ulysses (London, Bodley Head).

Kaufmann, W.
(1967) Musical Notations of the Orient (Bloomington, Indiana University
Press).

Kingdon, R.
(1958) The Groundwork of English Intonation (London, Longman).

Kostelanetz, R.
(1980) Text-Sound Texts (New York, William Morrow and Company).

Langer, S.K.
(1953) Feeling and Form (London, Routledge and Kegan Paul).

Levi-Strauss, C.
(1970) The Raw and the Cooked (London, Cape).

McAdams, S.
(1982) ‘Spectral Fusion and the Creation of Auditory Images’. In Music,
Mind and Brain (Clynes, M., editor) (New York, Plenum Press) pp. 279–
298.

Marler, P.
(1965) ‘Communication in Monkeys and Apes’. In Primate Behavior: Field
Studies of Monkeys and Apes (DeVore, L, editor.) (New York, Holt,
Reinhart and Winston) pp. 544–584.

Motherwell, R.
(1989) The Dada Painters and Poets: An Anthology (Harvard University
Press).

Nicholson, E. M. and Koch, L.

(1946) Songs of Wild Birds (London).

Pauli, H.
(1971) Für wen komponieren Sie eigentlich? (Frankfurt a.M., Fischer).

Robson, E.
(1981) Phonetic Music with Electronic Music (Parker Ford, Pa., Primary
Press).

Rodet, X., Potard, Y. and Barrière. J.-B.

(1984) ‘The CHANT Project: From the Synthesis of the Singing Voice to
Synthesis in General’. CMJ 8(3) pp. 15–31.

Rowell, T. E.
(1962) ‘Agonistic Noises of the Rhesus Monkey (Macaca mulatta)’.
Symposium of the Zoological Society of London 8 pp. 91–6.

Rowell, T. E. and Hinde, R. A.

(1962) ‘Vocal Communication by the Rhesus Monkey (Macaca mulatta)’.
Proceedings of the Zoological Society of London 138 pp. 279–294.

San Diego Extended Vocal Techniques Group

(1974) Index to a Recorded Lexicon of Extended Vocal Techniques (San
Diego, Center for Music Experiment and Related Research).

Schaeffer, P.
(1966) Traité de Objets Musicaux (Paris, Du Seuil).

Schaeffer, P., Reibel, G. and Ferreyra, B.

(1983) Solfège de l'objet sonore (Paris, INA/GRM).
(Three cassettes with accompanying text (original LP version 1966)).

Schafer, R. M.
(1969) The New Soundscape (BMI, Canada/Universal Edition).
(1977) The Tuning of the World (New York, Knopf).

Schouten, J. F., Ritsma, R. J. and Cardozo, B. L.

(1962) ‘Pitch of the Residue’. JASA 34(8/2) pp. 1418–1424.
Schwitters, K.
(1993) Poems, Performance Pieces, Proses, Plays, Poetics (Rothenberg, J.
and Joris, P. editors/translators) (Philadelphia, Temple University Press).

Shepherd, J., Virden, P., Vulliamy, G. and Wishart, T.

(1977) Whose Music? A Sociology of Musical Languages (London,
Latimer).

Stuckenschmidt, H. H.
(1959) Arnold Schoenberg (London, Calder).

Thom, R.
(1975) Sructural Stability and Morphogenesis (Reading, Mass., W. A.
Benjamin).

Thompson, D'A. W.
(1961) On Growth and Form (Cambridge University Press).

Thorpe, W. H. and North, M. E. W.

(1965) ‘Origin and significance of the power of vocal imitation: with
special reference to the antiphonal singing of birds’. Nature 208 pp. 219–
222.

Weber, M.
(1958) The Rational and Social Foundations of Music (Martindale, Riedel
and Neuwirth, translators) (Carbondale, Southern Illinois University Press).

Wessel, D.
(1979) ‘Timbre Space as a Musical Control Structure’. CMJ 3(2) pp. 45–52.

Wishart, T.
(1979) Book of Lost Voices (York, Wishart (private publication)).

Xenakis, I.
(1971) Formalised Music (Bloomington, Indiana UP).
[Revised edition: Stuyvesant, Pendragon Press (1992)].
 MUSIC EXAMPLES

This book is about listening. Trevor Wishart insists that only the ear can
validate or criticise music composition. His original lectures and plan for
this book included recordings of a vast array of musical examples from
many sources. Copyright problems have made the assemblage of a
complete accompanying recording prohibitively difficult.

What has been included on the accompanying CD are the otherwise

unobtainable extended vocal materials recorded by Trevor Wishart himself
and extracts of other of his works to illustrate the arguments of the text. In
addition regenerated acoustic and psychoacoustic examples have been
included. The CD track/index is given in the left-hand margin.

Within the current constraints of copyright law the author and editor invite
the reader to construct an ideal series of music examples from the following
list of commercial recordings which will be referred to in the text by the
example numbers indicated. This book makes most sense if the reader has
assembled these music examples and listens to them at the relevant point in
the text. All entries are CDs unless marked as LP or Cassette. Where the
author has made reference to specific sounds, transformations etc. timings
are given with respect to the actual recording cited, whereas where
reference is to a general style or approach no specific timings are given. For
well known works of which several recordings are easily available none is
specifically cited. (Where no source is cited the example may be found on
the accompanying CD.)

The author repeated music examples in his original lectures. He also

grouped sounds into a single Example when they were intended to be
directly compared. These aspects have been preserved in the text and the
sequence below.

The following are reproduced with permission: Example 3.5 (Jean-Claude

Risset); Example 3.20 (Roger Reynolds); Example 4.1 (David Wessel);
Examples 5.1 and 7.8 (Jean-Baptiste Barrière).

Chapter 1

Ex. 1.1: Denis Smalley: Pentes [8.10–10.52].

[Ode Record Co. Ltd. (New Zealand): CD MANU 1433]
[1.01] Ex. 1.2: Trevor Wishart: Musical Box from Menagerie [0.00–
1.20].
[YES Records (York): (LP) YES 8]
Ex. 1.3: Iannis Xenakis: Concret PH II.
[Nonesuch (NY): (LP) H-71246]
Ex. 1.4: Michael McNabb: Dreamsong [8.08–9.11].
[Mobile Fidelity Sound Lab: MFCD 818]
Ex. 1.5: Goon Show: Napoleon’s Piano.
[BBC Radio Collection (London): (Cassette) ZBBC 1016
(ISBN (0563) 225440)]
Ex. 1.6: Bernard Parmegiani: Etude élastique from De Natura
Sonorum [0.50–1.36].
[INA/GRM (Paris): INA C3001]
Ex. 1.7: Luciano Berio: Omaggio a Joyce [0.00–1.00].
[BVHaast (Amsterdam): CD 9109]
Ex. 1.8: Richard Coldman: Fret buzz [1.25–2.40].
[Incus Records (UK): (LP) Incus 31]

Chapter 2
Ex. 2.1: Japanese Joruri singing (gidayu style) (e.g. Takemoto
Tsunatayu: Kiyari Ondo).
[King Record Co. Ltd. (Japan): KICH 2008]
Ex. 2.2: North Indian singing (e.g. Sulochana Brahaspati: Khyal
(Raga Bilaskhani Todi)).
[Nimbus (UK): NI 5305)
Ex. 2.3: Jazz singing (e.g. Billy Holliday: Lady Sings the Blues).
[Verve (Polygram): 823 246–2]
Ex. 2.4: Josef Haydn: Dona nobis pacem from Missa in Tempore
Belli (‘Paukenmesse’)
Ex. 2.5: Japanese shakuhachi playing (e.g. Kohachiro Miyati:
Shika no Tone).
[Elektra Nonesuch: 7559–72076-2]
Ex. 2.6: A classical chamber work (using wind instruments) (e.g.
Antoine Reicha: Wind Quintet in D major op. 91 no. 3).
[Hyperion Records Ltd. (UK): CDA66268]
Ex. 2.7: Traditional jazz (e.g. Louis Armstrong and His Hot Five:
West End Blues).
[BBC Enterprizes (UK): BBC CD 597]
Ex. 2.8: Anton Webern: Symphonie op. 21 (1st movement).
[Sony Classical: SM3K 45845]
Ex. 2.9: Krzysztof Penderecki: Polymorphia (2.10–3.34 + 4.50–
5.36).
[Polskie Nagrania (Poland): PNCD017(A+B)]
Ex. 2.10: Iannis Xenakis: Pithoprakta (7.40–9.00).
[Le Chant du Monde (France): LDC 278368]
Ex. 2.11: Karlheinz Stockhausen: Carré (0.00–1.26).
[Stockhausen Verlag (Germany): Stockhausen 5]
Ex. 2.12: Pierre Boulez: Don (from Pli selon Pli) (3.05–5.00).
[Erato (France): 2292-45376-2]
Ex. 2.13: Spontaneous Music Ensemble (John Stevens, Trevor
Watts): Face to Face 5 (3.15–4.25).
[Emanem Records (UK): (LP) EMANEM 303]
[2.01] Ex. 2.14: Trevor Wishart: Anna's Magic Garden (2.30–3.31).
[Overhear Music (Keele, UK): Ohm 00l]

Chapter 3

The music examples from chapter 3 may be found on the accompanying

CD; except Example 3.5 they have been resynthesised based on models
generated originally by the Groupe de Recherches Musicales.

[3.01] Ex. 3.1: Melody with individual elements less than five
milliseconds long.
[3.02] Ex. 3.2: Melody with individual elements ten milliseconds long.
[4.01] Ex. 3.3: A low note on the piano.
[4.02] Ex. 3.4: As 3.3 but filtering out all frequencies below the second
harmonic.
[5.01] Ex. 3.5: An example of Shepard tones.
[6.01] Ex. 3.6: White noise: (a) normal (b) double speed (no pitch shift).
[6.02] Ex. 3.7: Complex sound: (a) normal (b) double speed (only a
third shift).
[7.01] Ex. 3.8: As 3.6.
[7.02] Ex. 3.9: Melody of filtered noise.
[8.01] Ex. 3.10: (a) piano with ‘flattened envelope’ (b) flute at same
pitch.
[8.02] Ex. 3.11: (a) flute with imposed ‘piano envelope’ (b) piano at same
pitch.
[9.01] Ex. 3.12: Sound with relatively constant envelope: (a) complete (b)
with start cut.
[9.02] Ex. 3.13: Piano note: (a) normal (b) with attack cut.
[10.01] Ex. 3.14: Bell sound: (a) normal (b) with attack cut.
[10.02] Ex. 3.15: Cymbal sound: (a) normal (b) with attack cut.
[11.01] Ex. 3.16: Vibraphone sound: (a) normal (b) with attack cut (3
versions).
[12.01] Ex. 3.17: Flute note: (a) with attack cut (b) normal.
[12.02] Ex. 3.18: Trumpet note: (a) with attack cut (b) normal.
[13.01] Ex. 3.19: The influence of onset synchrony on coherence.
[14.01] Ex. 3.20: Splitting an aural image into two (Roger Reynolds).
[15.01] Ex. 3.21: Different sound objects from a single source (metal sheet
and a taut string).
[16.01] Ex. 3.22: Sound of definite mass: resistance to filtering and its
transposition.
[17.01] Ex. 3.23: Grain illustrated with electronic impulses.
[17.02] Ex. 3.24: Grain illustrated for a bassoon note.
[18.01] Ex. 3.25: Speed up of melody into grain: (a) descending scale (b)
irregular melodic pattern.
[19.01] Ex. 3.26: Speed up of string of speech sounds approaches speech
multiplex.

Chapter 4

[20.01] Ex. 4.1: Two sequences based on David Wessel's researches into
timbre space.

Chapter 5

[21.01] Ex. 5.1: Chant example: transformation from bell to male voice.
[22.01] Ex. 5.2: Trevor Wishart: Red Bird transformation ‘(Li)-sss-(ten)’
to birdsong (1.26–1.42).
[October Music: Oct 001]
[23.01] Ex. 5.3: A typical vocally-produced multiplex.
[24.01] Ex. 5.4: As 5.3 but the field characteristic of the multiplex
changes with time.

Chapter 6

[25.01] Ex. 6.1: Trevor Wishart: Anna’s Magic Garden (2.50–3.31).

[Overhear Music (UK): Ohm001]
Ex. 6.2: Bernard Parmegiani: Etude élastique from De Natura
Sonorum [0.45–1.36].
 [INA/GRM (Paris): INA C3001]
Ex. 6.3: Luciano Berio: Visage (0.45–2.15).
[BVHaast (Amsterdam): CD 9109]
[26.01] Ex. 6.4: Trevor Wishart: Anticredos (unspatialised extract from
Wishart studio demo version) [unpublished].

Chapter 7

Ex. 7.1: Witold Lutoslawski: Trois Poèmes d'Henri Michaux (II:

Le grand combat) (2.20–3.50).
[Polskie Nagrania (Poland): PNCD041]
Ex. 7.2: Clément Janequin: Le Chant des Oyseaulx.
[Harmonia Mundi (France): HMC 901099]
Ex. 7.3: Ludwig van Beethoven: Symphony No.6 (Pastorale) (II:
Scene at the brook (entry of Nightingale (bar 129) to the
end)).
Ex. 7.4: Karlheinz Stockhausen: Trans (opening 2.00).
[Stockhausen Verlag (Germany): Stockhausen 19]
Ex. 7.5: Bernard Parmegiani: Matières induites (3.30-) + Ondes
croisées (–0.40) from De Natura Sonorum.
[INA/GRM (Paris): INA C3001]
Ex. 7.6: Goon Show: Napoleon's Piano.
[BBC Radio Collection (London): (Cassette) ZBBC 1016
(ISBN (0563) 225440)]
[27.01] Ex. 7.7: Trevor Wishart: Automusic (extract).
[Unpublished1]
[28.01] Ex. 7.8: Chant example: transformation from bell to male voice.
[29.01] Ex. 7.9: Trevor Wishart: Red Bird book/door slam (23.00–24.50).
[October Music: Oct 001]
[30.01] Ex. 7.10: Trevor Wishart: Red Bird lisss/birds (1.26–1.42).
[October Music: Oct 001]
Ex. 7.11: Alvin Lucier: I am sitting in a room.
[Sacramento: Composer/Performer Edition: (LP) ‘Source
 Record Number Three’ (in: Source: music of the avant
garde (Double issue 7/8 (1970)]
Ex. 7.12: Luc Ferrari: Presque Rien No. 1.
[Deutsche Grammophon: (LP) 2561 041]
Ex. 7.13: Larry Wendt: From Frogs.
[‘Poésie Sonore Internationale’ (Jean-Michel Place,
France) (Cassette) 10007]
Ex. 7.14: Michael McNabb: Dreamsong (0.00–1.20).
[Mobile Fidelity Sound Lab: MFCD 818]
[31.01] Ex. 7.15: Trevor Wishart: Musical Box from Menagerie (0.00–
1.20).
[YES Records (York): (LP) YES 8]
[32.01] Ex. 7.16: Trevor Wishart: Still Life from Menagerie (1.00–3.02).
[YES Records (York): (LP) YES 8]
Ex. 7.17: Karlheinz Stockhausen: Gesang der Jünglinge (5.10–
6.30).
[Stockhausen Verlag (Germany): Stockhausen 3]
Ex. 7.18: Luciano Berio: Visage (6.54–7.32).
[BVHaast (Amsterdam): CD 9109]

Chapter 8

Examples 8.1–8.12, from Trevor Wishart's Red Bird are all found on the
accompanying CD. The whole work is found on October Music (UK): Oct
001.

[33.01] Ex. 8.1: Sound-image ‘bellows/water-pump’ (20.26–22.20).

[34.01] Ex. 8.2: ‘Listen to reason’ transformation (4.01–5.31).
[35.01] Ex. 8.3: Garden landscape (11.00–12.12).
[36.01] Ex. 8.4: Reason landscape (37.42–39.09).
[37.01] Ex. 8.5: Animal-like sound images (a) emerge from the
background (15.02–16.30), (b) take on articulation of
words (16.57–17.57) and (c) gesturally interact with the
bird sounds (18.06–18.44).
[38.01] Ex. 8.6: Bird cadenza (33.15–34.27).
[39.01] Ex. 8.7: Word machine (20.25–20.42).
[40.01] Ex. 8.8: Phonemes as flocking birds (25.52–26.36).
[41.01] Ex. 8.9: ‘Rea’ transformation to bark (5.01–5.23).
[42.01] Ex. 8.10: ‘-ble’ bursts out into bubbling (4.06–4.14).
[43.01] Ex. 8.11: Ambiguous machinery/animal landscape: (a) a ‘glimpse’
of the animal sounds out of context (38.18–38.25) and
(b) placed in context (37.47–38.34).
[44.01] Ex. 8.12: ‘Rea’ transformation to clock: (a) to texture (4.30–4.44
and 6.07–6.31), (b) from texture to short reverberated
attack (27.13–28.38 and 29.04–29.10) and (c) from
attack to clock tick (30.49–30.53 and 31.45–32.45).

Chapter 10

Ex. 10.1: Jean Sibelius: Symphony No.4 (III: bars B+6 to C).

Chapter 11

Ex. 11.1: Arnold Schoenberg: Erwartung (bars 146–157).

[45.01] Ex. 11.2: Comparison of the simulated signals of the Great
Northern Diver (Loon), the wolf, the whale and the red
squirrel.

Chapter 12

The sound examples from Chapter 12 were all performed by Trevor Wishart
and are to be found on the accompanying CD.

The glottis and windpipe

[46.01] Ex. 12.1: Glottal vibrations (vibrating larynx): iterated impulses

(unpitched).
[46.02] Ex. 12.2: Glottal vibrations (vibrating larynx): normal rate of
vibration (pitched).
[46.03] Ex. 12.3: Windpipe vibration caused by forcible exhalation from
lungs: below the larynx.
[46.04] Ex. 12.4: Windpipe vibration caused by forcible exhalation from
lungs: above the larynx.
[47.01] Ex. 12.5: As 12.3.
[47.02] Ex. 12.6: The same (as 12.3) combined with the sound of the
larynx.
[47.03] Ex. 12.7: As 12.4.
[47.04] Ex. 12.8: The same (as 12.4) combined with the sound of the
larynx.
[47.05] Ex. 12.9: Octave division (normal pitch followed by division).

The tongue

[48.01] Ex. Tongue vibrations: tip at front of mouth.

12.10:
[48.02] Ex. Tongue vibrations: tip onto soft palate.
12.11:
[48.03] Ex. Tongue vibrations: tip strongly retroflexed.
12.12:
[49.01] Ex. Rolled ‘r’ with arched tongue middle mouth.
12.13:
[49.02] Ex. Rolled ‘r’ with arched tongue back of mouth.
12.14:
[49.03] Ex. Rolled ‘r’ with arched tongue against uvula (snoring).
12.15:
[49.04] Ex. Rolled ‘r’ with arched tongue brought into sung pitch
12.16: range.

The lips and cheeks

[50.01] Ex. Lip vibration: normal.

12.17:
[50.02] Ex. Lip vibration: folded towards teeth.
 12.18:
[50.03] Ex. Lip vibration: strongly pouted outwards.
12.19:
[50.04] Ex. Use of hands to assist production of long lip notes.
12.20:
[50.05] Ex. Lip oscillation variations.
12.21:
[50.06] Ex. Two independent lip oscillations.
12.22:
[51.01] Ex. Cheek vibration with pitch variation controlled by hands.
12.23:
[51.02] Ex. Sub-audio cheek vibration.
12.24:
[51.03] Ex. Independent cheek vibrations (with resultant beats).
12.25:

Filters

[52.01] Ex. Sung vowels.

12.26:
[52.02] Ex. Different routes from a to ü through the vowel space.
12.27:
[52.03] Ex. Effect of lip rounding on formants (vowels).
12.28:
[52.04] Ex. Different mouth (formant) positions for I: (u)le as
12.29: compared with (e)le.
[53.01] Ex. Filtered inhaled glottal clicks.
12.30:
[53.02] Ex. Vocal analogue of Shepard tone (rising and falling).
12.31:
[53.03] Ex. Filtered lip vibrations.
12.32:
[53.04] Ex. Use of hands to filter cheek vibrations.
12.33:
[53.05] Ex. Hand funnel filtering (noise sound).
12.34:
[54.01] Ex. Formant filtering to accentuate harmonics of a glottal
12.35: tone.
[54.02] Ex. Formant filtering to accentuate harmonics of a glottal
12.36: subharmonic.
[54.03] Ex. Formant filtering to accentuate harmonics of a high
12.37: pitched tongue vibration.

Filtered noise and whistles

[55.01] Ex. Changing formant structure of continuous noise

12.38: consonants (mouth cavity).
[55.02] Ex. Changing formant structure of continuous noise
12.39: consonants (mouth cavity/tongue combination).
[55.03] Ex. X/f hybrid continuous consonant.
12.40:
[56.01] Ex. The lip whistle: three registers.
12.41:
[56.02] Ex. Tongue whistle (tongue arch rear position).
12.42:
[56.03] Ex. Tongue whistle (tip of tongue in sh position).
12.43:
[56.04] Ex. Tongue whistle (tip of tongue in s position).
12.44:
[56.05] Ex. Tongue/lips double whistle.
12.45:

Double and treble production

[57.01] Ex. Different types of roar or bark (3 variants).

12.46:
[57.02] Ex. Tongue vibration (rrr) combined with glottal sounds (3
12.47: variants).
[57.03] Ex. Tongue vibration (RRR) combined with glottal sounds (2
 12.48: variants).
[57.04] Ex. Tongue vibration (retroflexed RRR) combined with
12.49: glottal sounds (2 variants).
[57.05] Ex. Tongue vibration (uvula U) combined with glottal sounds
12.50: (3 variants).
[58.01] Ex. Combinations of lip and glottis vibrations.
12.51:
[58.02] Ex. Combinations of cheek and glottis vibrations.
12.52:
[58.03] Ex. Combinations of noise sounds and glottis vibrations.
12.53:
[58.04] Ex. Combination with intermodulation of (a) ordinary
12.54: whistling with singing compared with (b) high s-whistle
with low glottal sounds.
[58.05] Ex. Two simultaneous tongue vibrations.
12.55:
[58.06] Ex. Combination of lip and tongue vibrations (with
12.56: intermodulation).
[59.01] Ex. Cheek and tongue vibrations.
12.57:
[59.02] Ex. Noise and tongue vibrations.
12.58:
[59.03] Ex. Lips and arched tongue vibrations.
12.59:
[59.04] Ex. Cheeks and arched tongue vibrations.
12.60:
[60.01] Ex. Noise and arched tongue vibrations.
12.61:
[60.02] Ex. Whistle and arched tongue vibrations.
12.62:
[60.03] Ex. Noise and uvular vibration of tongue.
12.63:
[60.04] Ex. Noise and lip vibrations.
12.64:
[61.01] Ex. Whistle and lip vibrations (‘trimphone’).
12.65:
[61.02] Ex. Simultaneous noise, tongue, lip, glottis vibrations
12.66: independently articulated.
[61.03] Ex. Glottis and tongue vibration through pouted lips
12.67: (‘trumpet’).

Air stream and other effects

[62.01] Ex. ‘Wamp’, ‘half-wamp’ and ‘quarter-wamp’.

12.68:
[62.02] Ex. Half-lunged multiphonic.
12.69:
[62.03] Ex. Unlunged whistle.
12.70:
[63.01] Ex. Manual-diaphragm flutter (continuous sound).
12.71:
[63.02] Ex. Manual-diaphragm drumming (short staccato sounds).
12.72:
[63.03] Ex. Shake-head flutter (pitched and noise versions).
12.73:
[63.04] Ex. Drum-glottis flutter.
12.74:
[63.05] Ex. Shake-body flutter.
12.75:
[64.01] Ex. Drum-cheeks flutter.
12.76:
[64.02] Ex. Strum-lips flutter.
12.77:
[64.03] Ex. Strum-nose flutter.
12.78:
[64.04] Ex. Diaphragm flutter with tongue vibration.
12.79:
[64.05] Ex. Diaphragm flutter with lip vibration.
 12.80:
[65.01] Ex. Ululation.
12.81:
[65.02] Ex. Ululation across break in voice.
12.82:
[65.03] Ex. Ululation with high depth (‘laugh’).
12.83:

Water effects

[66.01] Ex. Arched tongue vibration with saliva/water.

12.84:
[66.02] Ex. X/water.
12.85:
[66.03] Ex. Filtered X/water.
12.86:
[66.04] Ex. Filtered half-lunged X/water.
12.87:
[66.05] Ex. K-plosive + X/water (‘children's gunshot’).
12.88:
[66.06] Ex. Filtered half-lunged X/water: plosive staccato
12.89: production.
[67.01] Ex. Filtered half-lunged X/water: plosive staccato production
12.90: + tongue-tip vibration.
[67.02] Ex. Unvoiced/water sounds with pitch content.
12.91:
[67.03] Ex. Inhaled lip vibration (normal then water behind lips).
12.92:
[67.04] Ex. Inhaled air stream with water around sides of tongue.
12.93:

Transformations (exhaled sustainable sounds)

[68.01] Ex. Tongue oscillation transformations.

12.94:
[68.02] Ex. As 94 adding and subtracting water.
12.95:
[69.01] Ex. Pitched production: transformations between lunged and
12.96: half-lunged.
[69.02] Ex. Pitched production: transformations between under- and
12.97: over-breathed.
[70.01] Ex. Pitched production: transformations between normal and
12.98: fluttered.
[70.02] Ex. Transformation from tongue to lip vibration.
12.99:

Inhaled sounds

[71.01] Ex. Inhaled lip vibrations: tones, pulses, multiphonics.

12.100:
[71.02] Ex. As 100 + control with heals of hands.
12.101:
[71.03] Ex. Inhaled tongue vibration (retroflex position).
12.102:
[71.04] Ex. Inhaled tongue vibration (uvula position).
12.103:
[72.01] Ex. Inhaled pure tones.
12.104:
[72.02] Ex. Inhaled click trains.
12.105:
[72.03] Ex. Inhaled sub-harmonics.
12.106:
[72.04] Ex. Inhaled complex multiphonics.
12.107:
[72.05] Ex. Inhaled unstable complex vibration.
12.108:
[72.06] Ex. Inhaled unstable complex vibration with filtering.
12.109:
Pulses

[73.01] Ex. Epiglottis pulses.

12.110:
[73.02] Ex. Arched tongue pulses (rear, centre and front).
12.111:
[73.03] Ex. Tip of tongue pulses (from retroflexed to further
12.112: forward).
[73.04] Ex. Tongue/teeth and tongue/top lip pulses.
12.113:
[73.05] Ex. Teeth and lips pulses.
12.114:
[73.06] Ex. Lips pulses.
12.115:
[74.01] Ex. Pulses k and t: lunged, half lunged, unlunged.
12.116:
[74.02] Ex. Plosive half-lunged d.
12.117:
[74.03] Ex. Plosive half-lunged p.
12.118:
[74.04] Ex. Plosive unlunged p.
12.119:

Voiced pulses

[75.01] Ex. Pulse d (lunged) with secondary voicing and breath.

12.120:
[75.02] Ex. As 120: other types of voicing (windpipe, tongue
12.121: vibration, whistle).
[75.03] Ex. Pulse k (lunged): variation of airstream content.
12.122:
[75.04] Ex. Pulse k (lunged): variation of airstream type.
12.123:
[75.05] Ex. Pulse k (lunged): airstream glissando.
12.124:
[75.06] Ex. Plosive air stream effect (suppression of original pulse).
12.125:

Pulses with stops, buzzes, filtering

[76.01] Ex. Plosive d (unvoiced, voiced) compared with addition of

12.126: g-stop (d(g)!) (unvoiced, voiced).
[76.02] Ex. Plosive d with addition of p-stop (d(p)!).
12.127:
[76.03] Ex. Drum imitation (d(g)! with glottal vibration).
12.128:
[76.04] Ex. H-stream stopped with ch and k.
12.129:
[76.05] Ex. S-stream stopped with p.
12.130:
[76.06] Ex. Slap tongue (unlunged t), slap teeth pulses.
12.131:
[77.01] Ex. Buzzed lips: lunged.
12.132:
[77.02] Ex. Buzzed lips: half lunged.
12.133:
[77.03] Ex. Buzzed lips: unlunged.
12.134:
[77.04] Ex. As 134 with air exhalation.
12.135:
[77.05] Ex. Manually initiated lip pulses.
12.136:
[78.01] Ex. Pulse (unhinged k) with oral filtering.
12.137:
[78.02] Ex. Pulse (unlunged p) resonance control through pouting.
12.138:
[78.03] Ex. Plosive click with filter glissando.
12.139:
Simultaneous and alternated pulses

[79.01] Ex. Simultaneous pulses g + k.

12.140:
[79.02] Ex. Simultaneous pulses t + k.
12.141:
[79.03] Ex. Pulse alternation tktk...: lunged.
12.142:
[79.04] Ex. Pulse alternation tktk...: half-lunged.
12.143:
[79.05] Ex. Pulse alternation tktk...: unlunged.
12.144:
[79.06] Ex. Pulse alternation gthgth...: unlunged.
12.145:
[80.01] Ex. Pulse alternation ptpt...: lunged.
12.146:
[80.02] Ex. Pulse alternation ptpt...: half-lunged.
12.147
[80.03] Ex. Pulse alternation ptpt...: unlunged.
12.148
[80.04] Ex. Manual iteration lip pulses: lunged.
12.149:
[80.05] Ex. Manual iteration lip pulses: unlunged.
12.150

Pulses (clicks and combinations)

[81.01] Ex. Pulse during inhalation at epiglottis.

12.151:
[81.02] Ex. Unlunged click: lateral tongue movement.
12.152:
[81.03] Ex. Unlunged click: vertical tongue movement (kl).
12.153:
[81.04] Ex. Unlunged click: plosive vertical tongue movement.
12.154:
[81.05] Ex. Unlunged t- or tut-click.
12.155:
[81.06] Ex. Unlunged th-click.
12.156:
[82.01] Ex. Kiss (short) with various degrees of lip pouting.
12.157:
[82.02] Ex. Plosive + kiss combination.
12.158:
[82.03] Ex. Lateral + kiss combination.
12.159:

Transitionals and percussives

[83.01] Ex. Consonant transitions between harmonics of a given

12.160: glottal pitch: m, n, ng, ‘deep’ ng.
[84.01] Ex. Iterations of transitionals: mnmn... and n-ng-n-ng...
12.161:
[85.01] Ex. Hand clap with mouth as variable resonator.
12.162:
[85.02] Ex. Tapping teeth with mouth as variable resonator.
12.163:
[85.03] Ex. Popping finger out of closed lips.
12.164:
[85.04] Ex. Manual expulsion of air from cheeks through lips.
12.165:
[85.05] Ex. ‘Water drop’.
12.166:

Multiplexes and complex articulations

[86.01] Ex. Multiplex combinations: pk∫lgr + X+ (lunged and half-

12.167: lunged) + lip-flabber.
[86.02] Ex. As 167 + tongue movement.
12.168:
[86.03] Ex. As 168 + glottal glissando.
 12.169:

Chapter 13

[87.01] Ex. 13.1: Birdsong trills slowed down.

1 Original commissioned broadcast BBC2 Television (Sounds Different: ‘Music Outside’) 1980.
 MUSIC REFERENCES

The following works are referred to in the text in addition to those specified
by the author as music examples for audition (see above). For
electroacoustic works recording details are given otherwise the publisher of
the score.

Davies, Hugh (with Hans-Karsten Raeke)

Klangbilder [(CD) Klangwerkstatt Edition (Mannheim) SM 500 135 D]

Ferneyhough, Brian
Second String Quartet [Peters Edition]
Time and Motion Study III [Peters Edition]

Ferrari, Luc
Music Promenade [(LP) Wergo 60046]

IRCAM
IRCAM un portrait [(LP) Centre Georges Pompidou (Paris) IRCAM 001]

Ligeti, György
Nouvelles Aventures [Peters Edition]
Requiem [Peters Edition]
La Barbara, Joan
Voice is the original instrument [(LP) Wizard Records (New York) RVW
2266]

Reich, Steve
Come Out [(CD) Elektra Nonesuch 979 169–2)

Schaeffer, Pierre; Reibel, Guy and Ferreyra, Beatriz

Solfège de l'objet sonore [(3 Cassettes) INA/GRM (Paris) 4001–3 sc]

Schoenberg, Arnold
Erwartung [Universal Edition]
Pierrot Lunaire [Universal Edition]

Smalley, Denis
Orouboros [Unpublished]

Stockhausen, Karlheinz
Mikrophonie I [Universal Edition]
Hymnen [(CD) Stockhausen Verlag 10]
Telemusik [(CD) Stockhausen Verlag 9]
Stimmung [Universal Edition]

Wishart, Trevor
Anticredos [(CD) October Music (UK) Oct 001]]
Tuba Mirum [Wishart (York)]
Vox I [(CD) Virgin Classics (UK) VC 7 91108-2]
 INDEX
Allen, Jane xii
Anhalt, Istvan xi, 310–1, 315–6, 321
Armstrong, Louis 340
Auden, W.H. 175
Augustin (Saint) 15

Bach, J. S. 15, 30, 321

Ball, Hugo 287, 309
Barrière, Jean-Baptiste 96
Barthes, Roland 258
Bastian, Jarvis 239
Bateson, Gregory 239, 260
Baudelaire, Charles 290
Beethoven, Ludwig van 131, 133–4, 136, 343
Berio, Luciano 5–6, 113, 161, 260, 298, 301, 305–6, 340, 342–3
Bogert, C. M. 319
Boulez, Pierre 3, 6–7, 31, 34–5, 77, 80, 129, 341
Brahaspati, Sulochana 340
Breton, André 290
Brubeck, Dave 30
Buchla, Don 331
Burwell, Paul 36
Busnel, René-Guy 242–3

Cage, John 5
Cardozo, B. L. 54
Carroll, Lewis 299, 309
Chowning, John 194
Cobbing, Bob xiv
Coldman, Richard 5, 340
Coleman, Peter xii
Confucius 14
Cott, Jonathan 194
Curtay, Jean-Paul xi, xiii–xiv, 184–5, 263, 280, 283–4, 289, 294–5, 307, 311

Dante 179
Darwin, Charles 249–50
Davies, Hugh xiv, 36, 353
Descartes, René 31

Eliot, T. S. 169
Emmerson, Simon xii, 113, 136
Endrich, Tom xi
Erickson, Robert 67–8, 94
Ernst, Max 137

Fellini, Federico 314

Ferneyhough, Brian 42–3, 105, 107, 353
Ferrari, Luc 129, 136, 159, 343, 353
Ferreyra, Beatriz 4, 8, 70, 353
Flaubert, Gustave 179
Fónagy, I. 310
Fourier, J. B. 48, 50–1, 53, 55–6, 58–9, 325
Franklin, Aretha 258

Galilei, Galileo 85
Garland, Judy 259
Gauss 59–60
Goody, J. 20, 22
Goin 319
Gregory (Pope) 15

Hausmann, Raoul 293

Haydn, Josef 27–8, 340
Hayes, B. 179
Heisenberg, Werner 54
Helmholtz, Hermann 29, 31, 48, 50–1, 58, 62
Helms, Hans G. 309
Hiller, Lejaren 6, 25
Hinde, R. A. 254
Holliday, Billy 340
Holmes, Philip xii
Hooker, Barbara I. 320
Hopkins, Gerard Manley 184, 306
Hugo, Victor 290

Isou, Isadore 289–90

Jackson, Mahalia 258

Jacobson, Roman 299
Janequin, Clément 131–2, 343
Joplin, Janis 259
Joyce, James 299, 306–7
Jung, Carl 176

Kaufmann, Walter 18
Keane, David xii
Kepler, Johannes 47–8
Kerouac, Jack 308
Kingdom, R. 310
Klee, Paul 134
Koch, L. 292
Kostelanetz, Richard xiv, 6, 308–9

Laban, Rudolf 114

La Barbara, Joan 263, 353
Lancino, Thierry 65
Langer, Susanne 17
Lao Tse 14
Lévi-Strauss, Claude 163–4, 167
Ligeti, György 316, 321, 353
Lucier, Alvin 158–9, 343
Lutoslawski, Witold 130, 305, 321, 343

Machover, Todd 80
Magdics, K. 310
Mallarmé, Stéphane 290
Markov 70
Marler, Peter R. 253, 255
Marx, Karl 14
Mayes, Martin 36
McAdams, Steven 64–5
McNabb, Michael 5–6, 65, 159, 163, 285–6, 340, 343
Miyati, Kohachiro 340
Moorer, James Andy 328
Motherwell, R. 293

Newton, Isaac 38, 85

Nicholson, E. M. 292
North, M. E. W. 320

Orton, Richard xii

Palestrina, Giovanni da 15
Parker, Charlie 30
Parmegiani, Bernard 5, 111, 136–7, 155, 340, 342–3
Pauli, Hansjörg 129
Penderecki, Krzysztof 32, 341
Piaf, Edith 259
Planck, Max 54
Plato 13–14, 16, 35, 39, 47–8
Plutarch 45
Potard, Yves 96
Pythagoras 45, 47–8, 67, 71–5, 78, 129, 188

Rameau, Jean Phillipe 8

Raeke, Hans-Karsten 353
Reibel, Guy 4, 8, 70, 353
Reich, Steve 305, 353
Reicha, Antoine 340
Respighi, Ottorino 131
Reynolds, Roger 65
Rimbaud, Arthur 290
Ritsma, R. J. 54
Robertson, Robert ix
Robson, E. 306
Rodel, Xavier 96
Rowell, T. E. 254

Sabatier, Roland xiii, 291

Sachs, Andrew 138
Schaeffer, Pierre 4, 8, 32, 66–7, 70, 94, 129, 134, 353
Schafer, Murray 5, 175–6, 180, 292
Schoenberg, Arnold 15, 30, 39–40, 258, 311, 344, 353
Schouten, J. F. 54
Schwitters, Kurt 305–7, 309
Shepard, Roger 191
Shepherd, John 3, 11
Sibelius, Jean 192, 344
Smalley, Denis 4, 192, 340, 354
Steele, Jan 38
Stevens, John 341
Stockhausen, Karlheinz xiii, 34, 37, 104, 116, 133, 135, 144–5, 148, 160, 163, 165, 194–5, 252, 321,
341, 343, 354
Stuckenschmidt, Hans H. 40
Subotnick, Morton 105–6

Tarkieltaub, Jacqueline 307

Tchaikovsky, Peter 130
Thorn, René xiii, 90–1
Thompson, D'Arcy xiii, 88–9
Thoreau, Henry 5
Thorpe, W. H. 320
Tsunatayu, Takemoto 340
Tzara, Tristan 290

Valéry, Paul 290

Verlaine, Paul 290
Virden, Phil 3, 11
Virgil 179
Vulliamy, Graham 3, 11

Wagner, Richard 30, 165

Watt, I. 20, 22
Watts, Trevor 341
Weber, Max 12
Webern, Anton xiv, 32, 40, 302, 341
Wendt, Larry 159, 343
Wessel, David 78, 80–1, 342
Wishart, Jackie xii
Wishart, Trevor ix, x, 3, 11, 32, 36, 263, 284, 339–44, 354

Xenakis, Iannis 5, 32–3, 67, 69–70, 115, 258, 340–1