Thinking In and About Music

Thinking In and
About Music
Analytical Reflections on Milton Babbitt’s
Music and Thought

Z AC HA RY B E R N ST E I N

1
 3
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.

© Oxford University Press 2021

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

Library of Congress Cataloging-​in-​Publication Data

Names: Bernstein, Zachary, 1987– author.
Title: Thinking in and about music : analytical reflections on
Milton Babbitt’s music and thought / by Zachary Bernstein.
Description: New York : Oxford University Press, 2021. |
Includes bibliographical references and index.
Identifiers: LCCN 2020042599 (print) | LCCN 2020042600 (ebook) |
ISBN 9780190949235 (hardback) | ISBN 9780190949259 (epub) |
ISBN 9780190949242 | ISBN 9780190949266
Subjects: LCSH: Babbitt, Milton, 1916–2011—Criticism and interpretation. |
Music—20th century—History and criticism. |
Schenker, Heinrich, 1868–1935—Influence.
Classification: LCC ML410.B066 B47 2021 (print) |
LCC ML410.B066 (ebook) | DDC 780.92 [B]—dc23
LC record available at https://lccn.loc.gov/2020042599
LC ebook record available at https://lccn.loc.gov/2020042600

DOI: 10.1093/​oso/​9780190949235.001.0001

1 3 5 7 9 8 6 4 2
Printed by Integrated Books International, United States of America
to Jeffrey Bernstein and Andrew Sessler,

in love and gratitude

 Preface

After all, the way you think about music is the way you hear music.
You can’t possibly separate knowing how and knowing that or
thinking about and thinking that and thinking in.
—​Milton Babbitt, interview with Charles Amirkhanian

Milton Babbitt was, at once, one of the twentieth century’s foremost

composers and a founder of American music theory. These two aspects of
his creative life, “thinking in” and “thinking about” music, nourished each
other. Theory and analysis inspired fresh compositional ideas, and composi-
tional concerns focused theoretical and analytical inquiry. But the relation-
ship between Babbitt’s writings and his music is neither simple nor direct. It
is the purpose of this volume to look into Babbitt’s theoretical and composi-
tional work side by side—​to evaluate ways in which the ideas developed in
his writings illuminate his music and ways in which his music goes beyond
those ideas.
I first arrived in New York City as an impressionable teenager—​in fact,
at almost exactly the same age Babbitt had come to New York some seventy
years earlier. It was 2005. Babbitt’s ninetieth birthday was that coming spring,
and throughout the year there were many concerts celebrating his work. It
was by attending quite a number of these that I first encountered Babbitt’s
music, and I consider myself extraordinarily fortunate that my introduction
to Babbitt’s music was through direct experience of live performances, unen-
cumbered by ideology, by misplaced suppositions about his elitism, or by any
sense of what I was supposed to be listening for. Although I have since ac-
quired an interest in Babbitt’s intellectual background and find his writing a
continually enriching source of speculation and puzzlement, it is my original
fascination with the experience of his music that has sustained this project.
I eventually came to recognize that Babbitt’s writings suggested a compel-
ling story about his music, but that his music itself was richer, less tidily for-
malistic, and simply more interesting than his writings suggest. His writings
establish the principles behind his serial arrays, but as Joseph Dubiel has
x Preface

long argued, an array is not a composition; analysts must not conflate the
two. There is more in Babbitt’s music than a succession of series forms and
aggregates. I found the relationship between Babbitt’s systems and his music,
to put it as Babbitt might have, to be problematical—​it posed interesting
questions I wanted to explore. It was through this realization that I came to
feel that the simultaneous consideration of Babbitt’s writings and his music
was a worthwhile standpoint for an analyst to adopt.
The book opens with two chapters primarily concerned with Babbitt’s
prose and the compositional procedures that reflect principles he theorizes.
Chapter 1 excavates the influence of Heinrich Schenker on Babbitt. I discuss
how Babbitt’s compositional arrays build upon his interest in Schenkerian
hierarchy. Schenkerian organicism, in turn, provides expectations—​
hierarchical development, the surface of the music reflecting its presumed
source, and so forth—​that prove fruitful analytically when approaching
his music: one can interpret his music on the basis of expectations met
or subverted. Chapter 2 focuses on Babbitt’s philosophical and cognitive
interests, which lead via other routes to a similar set of principles. In both
chapters, although historical and intellectual context is given as necessary,
Babbitt’s writings are examined with the greedy eye of an analyst—​a reader
concerned above all with enriching and clarifying his own encounters with
Babbitt’s music.1 Within these chapters, I review the most important elem-
ents of Babbitt’s compositional procedures: trichordal arrays, all-​partition
arrays, the time-​point system, and cross-​references. I do not provide a
full-​scale exposition of his compositional techniques—​there already exists
an excellent book, Andrew Mead’s An Introduction to the Music of Milton
Babbitt, that does just that. But enough information is provided in these
chapters that a reader unacquainted with prior scholarship on Babbitt
will be introduced to the main features of his approach. (The book’s glos-
sary should also aid the uninitiated.) Moreover, the overview of his com-
positional procedures is illuminated by the newly available Milton Babbitt
Collection at the Library of Congress, a stunning resource likely to keep
Babbitt scholars busy for years.
Chapter 3 expands on prior discussions of Babbitt’s compositional
procedures with a close examination of a pivotal period in his compositional

1 Accordingly, I do not address Babbitt’s views on academia or society, much as these are important

aspects of his thought. Readers interested in these subjects will find valuable treatments in Brody
1993, Harker 2008, and Girard 2007 and 2010.
 Preface xi

development, the years around 1960. The various transformations of Babbitt’s

technique in those years are revealing about certain lifelong concerns of his,
particularly regarding perception.
The remaining chapters shift the focus to Babbitt’s music itself, although
I continually return to his writings and compositional procedures when
they prove relevant or enlightening. In general, Chapters 4 to 7 gradually
transition from an analytical stance focused on serialism to one that views
serial analysis from a greater distance. While Babbitt’s writings and com-
positional techniques offer insight to analysts, there is uncertainty about
the implementation of those insights and, more importantly, much going
on in his music that is not directly attributable to serialism and in some
cases is even in conflict with serial principles. Four topics almost wholly
unaccounted for in Babbitt’s writings guide these chapters: rhetoric, ges-
ture, temporality, and text setting, each of which is shown to supplement
and in some cases supplant serial considerations. Chapter 4 examines
Composition for Four Instruments (1948), a prominent early work, noting
both the challenges in identifying its series and the value in doing so. In
particular, it is shown that identifying the work’s series helps isolate aspects
of the work that are not explainable by reference to the series; these aspects
are shown to have rhetorical significance. Chapter 5 discusses text-​music
relations in five of Babbitt’s early texted pieces, highlighting the techniques
he developed for the projection of poetic form and meaning. In several of
these works, his typical serial procedures are altered or even abandoned
for the sake of text setting. Chapter 6 focuses on the issue of syntactical
completeness and its link to temporality. While Babbitt’s techniques
create expectations for when his arrays, and thus his compositions, will be
completed, his compositions often deviate from these expectations. This
inspires a reflection on just what the temporal experience of Babbitt’s music
is like. Local rhetorical devices are shown to be more responsible for the
perception of closure than serial hierarchy. Finally, Chapter 7 focuses on
the gestural characteristics of Babbitt’s music. Babbitt’s music is shown to
inspire embodied sensation, more in spite of than because of serial hier-
archy. At times, he breaks from serial expectations in order to heighten his
music’s gestural impact.
The analytical results of Chapters 4–​7 challenge the theoretical vision
outlined in Chapters 1–​2. How should Babbitt’s writing be taken as a guide to
his music given the many ways in which his music goes beyond his theories?
A brief Afterword meditates on the problematic picture that remains.
 Acknowledgments

Research can feel like a solitary exercise, but no author writes alone. This
book is the result of a lifetime of good fortune and the contributions of more
people than I can count.
Several hundred students at the Eastman School of Music provided un-
witting feedback on the book’s arguments in the form of class discussion.
(There is no more helpful audience than a skeptical classroom!) The students
in my 2016 PhD seminar—​Alyssa Barna, David Hier, Catrina Kim, Samuel
Reenan, and Tobias Tschiedl—​proved especially insightful. Eastman has also
repeatedly provided crucial material support. Professional Development
Committee Grants enabled me to air out the book’s ideas at numerous con-
ferences and supported indexing expenses, and a Spring 2019 Academic
Leave saw the book to its conclusion.
The ideas presented here first began to take shape during my time at the
CUNY Graduate Center, and I remain indebted to the mentorship I received
there. Joseph Straus’s encouragement has buoyed my work since its earliest
stages, and I can hardly imagine how my career would have unfolded without
the opportunities he provided for me. William Rothstein continues to rep-
resent an ideal of clear writing and clear thought. And Jeff Nichols has pro-
vided a constant reminder that one should seek truth rather than simplicity.
My time at CUNY was also enriched by a brilliant circle of classmates. Ellen
Bakulina, Steven Beck, Daniel Colson, Edward Klorman, Drew Nobile, and
Andrew Wilson deserve particular mention. Loretta Terrigno was a vital part
of this circle, too, but more on her in a moment.
The contents of this book were informed by many conversations and
exchanges over a very long period of time. I’d like to thank in particular
Matthew BaileyShea, Joseph Dubiel, Edward Klorman, Scott Gleason,
Harold and Sharon Krebs, Alison Maggart, Joshua Banks Mailman, Andrew
Mead, Robert Morris, Stephen Peles, Claudio Spies, Philip Stoecker, and
above all Daniel Colson, who was an invaluable companion in the early
stages of my thinking on Babbitt. As I was writing, Matt, Scott, Ed, and Phil
lent their careful eyes to several draft chapters. You can blame any remaining
errors on them.
xiv Acknowledgments

Betty Ann Duggan has generously enabled me to access and reproduce her
father’s sketch materials. Librarians, as ever, are the heroes behind the scenes;
I’m particularly indebted to those at the Library of Congress, Sibley Music
Library, New York Public Library, Juilliard, and the CUNY Graduate Center.
A Subvention Award from the Society for Music Theory offset the cost of
copyright permissions. Susan Monahan provided terrific help with indexing.
And the keen-​eyed staff at Oxford University Press and Newgen Knowledge
Works, including Sean Decker, Norman Hirschy, Joseph Matson, Ayshwarya
Ramakrishnan, and Suzanne Ryan could not have been more helpful, effi-
cient, or pleasant to work with.
My family has given me unquestioning (if occasionally bemused) support
as I pursued this most unlikely of career paths. My wife, Loretta, is a constant
sounding board, reality check, and source of comfort. She’s more patient than
I deserve. Andrew Sessler, my grandfather, showed me the value of relentless
curiosity. My father, Jeffrey Bernstein, gave me his warmth and loving con-
fidence. While neither lived to see this book completed, it would have been
unthinkable without their guidance and example. I dedicate it to them.
 Permissions

The following permissions to reproduce text and score excerpts are gratefully
acknowledged.

Portions of Chapter 1 and much of Chapter 6

First appeared in Music Theory Spectrum, vol. 38, no. 2. Used by permission.

Much of Chapter 3
Copyright 2018. Perspectives of New Music. Used by permission. First
appeared in Perspectives of New Music, vol. 56, no. 1, 2018.

About Time
By Milton Babbitt
Copyright © 1982 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

“As Long as It Isn’t Love,” from Three Theatrical Songs

By Milton Babbitt
Copyright © 1981 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

Canonical Form
By Milton Babbitt
Copyright © 1983 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

Composition for Four Instruments

By Milton Babbitt
Copyright © 1949 by Merion Music, Inc. Copyright renewed. A wholly
owned subsidiary of Theodore Presser Company. All Rights Reserved. Used
with permission.
xvi Permissions

Composition for Tenor and Six Instruments

By Milton Babbitt
Copyright © 1960 (Renewed) by Associated Music Publishers, Inc (BMI)
All Rights Reserved. International Copyright Secured. Reprinted by
permission.

Du
By Milton Babbitt
I am grateful to the European American Music Distributors Company for
granting permissions for the reproduction of excerpts.

Emblems (Ars Emblematica)

By Milton Babbitt
Copyright © 1989 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

Glosses
By Milton Babbitt
Copyright © 1988 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

Homily
By Milton Babbitt
Copyright Sonic Arts Editions. Used by permission of Smith Publications,
54 Lent Road, Sharon, Vermont USA.

None but the Lonely Flute

By Milton Babbitt
Copyright © 1991 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

“Now Evening after Evening”

By Milton Babbitt
Copyright © 2002 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.
 Permissions xvii

Philomel
Words by John Hollander
Music by Milton Babbitt
Copyright © 1964 (Renewed) by Associated Music Publishers, Inc (BMI)
International Copyright Secured. All Rights Reserved. Reprinted by
permission.

Post-​Partitions
By Milton Babbitt
Copyright © 1966 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

Septet but Equal

By Milton Babbitt
Copyright © 1992 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

A Solo Requiem
By Milton Babbitt
Copyright © 1977 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

String Quartet No. 6

By Milton Babbitt
Copyright © 1993 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

Tutte le Corde
By Milton Babbitt
Copyright © 1994 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

Two Sonnets
By Milton Babbitt
Copyright © 1955 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.
xviii Permissions

Vision and Prayer

By Milton Babbitt
Text by Dylan Thomas
Copyright © 1961 (Renewed) by Associated Music Publishers, Inc (BMI)
International Copyright Secured. All Rights Reserved. Reprinted by
permission.

“Vision and Prayer”

By Dylan Thomas
First published in the Sewanee Review, vol. 53, no. 3, Summer 1945.
Reprinted with permission.

“A Waltzer in the House”

Copyright © 2003 by C.F. Peters Corporation. All Rights Reserved. Used by
permission.

“The Widow’s Lament in Springtime”

By Milton Babbitt
I am grateful to European American Music Distributors Company for
granting permissions for the reproduction of excerpts.

“The Widow’s Lament in Springtime”

By William Carlos Williams (The Collected Poems: Volume I, 1909–​39)
Copyright © 1938 by New Directions Publishing Corps
Reprinted here by kind permission of New Directions Publishing Corps and
Carcanet Press Limited, Manchester, UK
 About the Companion Website

www.oup.com/us/thinkinginandaboutmusic
Oxford University Press has created a website to accompany Thinking In and
About Music. The website contains links to playlists with recordings of the
book’s examples and pdfs of the book’s figures that can be enlarged at the
reader’s convenience.
 1
On Milton Babbitt, Schenkerian

Introduction

The late writings of Milton Babbitt sing with nostalgia. Gifted with long life
and a steel-​trap memory, thrilled to have come of age in the vertiginous tur-
bulence of the 1930s, proud of his acquaintance with a long list of musical
giants, and well aware that the story of his development was central to the
story of the development of American music and musical thought, Babbitt
wrote and spoke again and again in his later years about his own musical
and intellectual formation. In these reminiscences, the galvanizing event
that made the man was not the hearing of any particular composition, the
reading of any particular book, or the meeting of any particular inspiring
personage—​although music, books, and people certainly do figure in these
stories—​but his arrival in New York City in 1934. Drawn to New York by the
discovery of a volume—​Marion Bauer’s Twentieth Century Music (1933)—​
that offered tantalizing glimpses of recent European music still so little heard
in the United States, and that suggested to the young musician, flush with
“curiosity and appetite for contemporary music” (Babbitt [1991b] 2003,
439), that New York was the place to encounter this music, Babbitt discov-
ered not only music, but an intellectual environment he scarcely could have
predicted.1
The rise of Nazi Germany had already loosed a trickle of what would soon
become a flood of refugees from Europe’s cultural and intellectual classes.
New York City would be the arrival point, and often permanent home, of
a vast number of European musicians and scholars of every stripe. Arnold
Schoenberg, among the first of these musical exiles, arrived in New York just
three months before Babbitt (Babbitt [1991b] 2003, 439). Although arriving
at different times and via different routes, the remnants of old intellectual

1 Babbitt recounted the story of his early years in New York dozens of times in his last several

decades. The fullest treatments are Babbitt (1991b) 2003 and (1999) 2003; see also Hilferty 2011. On
music theory in the New York City of Babbitt’s youth, see Girard 2007, 111–​84.

Thinking In and About Music. Zachary Bernstein, Oxford University Press. © Oxford University Press 2021.
DOI: 10.1093/​oso/​9780190949235.003.0001
2 Thinking In and About Music

circles were quick to find each other, and by the late 1930s had reconstituted
to a substantial degree (see Peles 2012, esp. 22). With broad talents and
broader interests, Babbitt moved easily among these groups, coming into
contact at an early age—​his late teens and early twenties—​with first-​rank
intellectuals from an exceptional variety of disciplines and subdisciplines. As
Stanley Cavell recollects, the young Babbitt “seemed to know every musician
in the world and every writer and painter in New York” (2010, 297; see also
Beardslee with Proctor 2017, 96). While he may have been on the “fringes”
of the circles formed by acolytes of Schoenberg and Heinrich Schenker, as
Aaron Robert Girard (2007, 170) has it, he moved among the fringes of many
groups. The heterogeneous thought with which he surrounded himself in
those early, formative years would mark him for life. Although his reflections
on that time express full awareness of the tragedies of the 1930s and ’40s—​
the tragedies that brought European intellectuals to New York and that kept
so many of them in poverty and humiliating anonymity—​they crackle with
delight for his having been a witness to the “jagged edges of abruption”2 of a
transforming world.
In Babbitt’s telling, his primary influences from those years—​or at least
those whose influence would have the longest effects—​ were Viennese,
by choice if not by birth. In a geometric twist on the Vienna Circle of log-
ical empiricists, Babbitt came to refer to these influences as his “Vienna
Triangle.” It was a shape whose vertices included the twelve-​tone composer
Arnold Schoenberg, the tonal theorist Heinrich Schenker, and the logical-​
empiricist philosopher Rudolf Carnap.3 It is, in many respects, a curious
pantheon: Schoenberg, Schenker, and Carnap were all born in the late nine-
teenth century and, at some point in their lives, all lived in Vienna; but they
shared very little else.4 Nonetheless, the influence of all three of these fig-
ures seeps out of nearly every page of Babbitt’s writings and compositions.
Rather than adopting any of their philosophies in full, Babbitt’s thought can

2
This lovely phrase appears in Babbitt (1960) 2003, 55.
3
Babbitt (1999) 2003, (1991b) 2003, and 1987, 17. The membership of the triangle changes some-
what in different tellings: two of the vertices are always Schenker and Schoenberg; the third is some-
times the Vienna Circle as a whole and sometimes just Carnap, Babbitt’s favorite philosopher from
that school.
4 Babbitt was fond of noting the proximity of Schoenberg, Schenker, and Carnap (e.g., Babbitt

[1976a] 2003, 339; [1999] 2003, 480), typically to lament that they seem not to have learned from one
another, but it should be noted that their time in Vienna scarcely overlapped. Schoenberg accepted
a position in Berlin in 1926, the same year Carnap moved to Vienna. The philosophers who consti-
tuted the Vienna Circle began meeting in 1924, but the publications and conferences that defined
the group in public consciousness only began in 1928, after Schoenberg had left Vienna for good and
only seven years from the end of Schenker’s life.
 On Milton Babbitt, Schenkerian 3

be characterized as an idiosyncratic selection of various aspects of each of

these thinkers’ work, chosen and refined in service of Babbitt’s own ends.
The name of Schenker is ubiquitous in Babbitt’s writings. Nonetheless,
in many ways the influences of Schoenberg and the tradition of logical em-
piricism of which Carnap is a part are more obviously pertinent to Babbitt’s
work than the influence of Schenker, and the scholarly literature reflects
this. Babbitt is typically introduced as one who “extended” Schoenberg’s
twelve-​tone technique (Morris [2000] 2010, 5), perhaps even as a “disciple”
of Schoenberg himself (Harker 2008, 347).5 While explicit discussions of the
philosophical sources of Babbitt’s thought have been somewhat less frequent,
at least until recently,6 Babbitt’s Carnapian concern for a “scientific” (Babbitt
[1961a] 2003, 78) understanding of meaning and discourse has been much
remarked on (e.g., Guck 1994 and Parkhurst 2013). These are, undoubt-
edly, profoundly important aspects of Babbitt’s thought. While Babbitt’s
Schoenbergian inheritance will not be lingered on in this volume—​it has al-
ready been addressed thoroughly by Andrew Mead (1994, esp. 5–​25)—​the
influence of Carnap will be taken up at length in Chapter 2.
But if Schenker is, at first sight, harder to account for in Babbitt’s Vienna
Triangle than Schoenberg or Carnap, he is no less central. Babbitt (1991a,
129) himself admits as much: “No influence has caused a greater transforma-
tion of our thinking about music than Schenker’s, be it in reaction to or in
reaction with.” Schenker’s influence can be seen in Babbitt’s views on a whole
range of issues, including his understanding of the tonal and twelve-​tone sys-
tems, his analyses of the work of a number of distinct composers, his view of
music perception, and even his use of metaphor in analytical prose. Most im-
portantly, Babbitt’s compositional technique is based to a significant extent on
his reading of Schenker. Given the depth of this influence, I believe that to un-
derstand Babbitt one must come to terms with the ways in which he is, indeed,
a Schenkerian, although a rather peculiar one.
Furthermore, Babbitt’s absorption of Schenker’s hierarchical organicism
defines an ideological and aesthetic frame within which his other influences
may be understood. His philosophical influences, for instance, may mani-
fest most visibly in Babbitt’s concerns about music-​theoretical discourse,7

5 Although he was certainly an intellectual “disciple” of Schoenberg, as Harker says, Babbitt was

not one of Schoenberg’s students. They met only a few times (see Babbitt 1995, 63–​65).
6 See Andreatta 2012, Brackett 2003, Derkert 2008, Gleason 2013, and Peles 2012.
7 Although Babbitt’s critique of music-​theoretical methodology and language are often explicitly

linked to his reading in philosophy (e.g., Babbitt [1961a] 2003; [1972] 2003), as noted in Cook 1995,
4 Thinking In and About Music

but also lead him to appropriate the logic of axiomatic deductive systems to-
ward the understanding of Schenker and, thus, toward the creation of new,
Schenker-​inspired musical hierarchies. Although Babbitt draws the twelve-​
tone system from Schoenberg, both his descriptions of it and his composi-
tional handling of it are deeply indebted to Schenker—​significant aspects of
his compositional practice can be roughly characterized as the animation of
Schoenbergian technology by Schenkerian ideology. Therefore, given the
centrality of Schenker in Babbitt’s thought, an examination focused on that
particular influence is revealing about other facets of his thought as well.
Babbitt was, in so many respects, an innovator. An unrepentant radical. His
music is jarring, rhythmically uneven, off-​balance, dissonant, chockablock
with new techniques. An avatar of high modernism, Babbitt was convinced
that composition exemplified a form of research—​that novel compositions
could extend human knowledge (see Babbitt [1962] 2003, 109; 1987, 182–​83).
What to make, then, of the inspiration he found in Schenkerian organicism—​
a conservative, even reactionary, aesthetic? Perhaps there is a tension, but
it is a tension characteristic of modernism, even modernity. As Michael
Levenson (2011, 2) has it, “Modernity remains haunted both by a search for
novelty and by the recollection of precursors.” Or in Babbitt’s ([1979a] 2003,
371) words, “in music, at least, what . . . characterizes a revolutionary period
is that it reexamines its past.” Like his musical heroes Schoenberg, Schenker,
and Brahms (on the latter, see Shields 2012, 353–​54)—​or, for that matter, his
literary idols James Joyce and William Faulkner (see Babbitt 1979b, 44; also
Peles 1998, 498–​99; Maggart 2017a, 152–​60)—​innovation for Babbitt was
never a matter of destroying the past or even progressing beyond it.8 The
achievements of composers and theorists past were to be reexamined and
mined for their ability to contribute to his own personal aesthetic ends.
Musical analysis reveals that some of these ends are concordant with or-
ganicism and some are not. As the later chapters of this book explore various

91, there are strong resonances between Babbitt’s complaints and Schenker’s polemics against herme-
neutic critics. Yet while Babbitt may have found some of his acid tongue in Schenker, one need not
impute direct influence in this regard. Among other things, Babbitt also found much of Schenker’s
methodology lacking, complaining, for instance, about Schenker’s invocation of the overtone system
and the metaphysical significance he attributed to the number five (Babbitt [1961a] 2003, 80–​81).
More likely, both Babbitt and Schenker are responding to what Stephen Peles (2012, 22) describes
as the “broader Austrian project of Sprachkritik”—​the critique of language that preoccupied fin-​de-​
siécle Vienna, including many of Babbitt’s influences.

8 In Babbitt’s words, “I never think of my music . . . as representing progress” (quoted in Bortz

2005, 94).
 On Milton Babbitt, Schenkerian 5

facets of his compositions, complications with the organicist model will

arise—​pieces or situations that, in one way or another, contest the expec-
tations engendered by hierarchical organicism. The picture that emerges
will suggest a more limited and nuanced view of the role of organicism in
Babbitt’s music.

Goethe, Schenker, and Hierarchical Organicism

The notion that a piece of art might be profitably described as having char-
acteristics of a living organism has deep roots in Romantic and even pre-​
Romantic aesthetics. Depending on how one defines the term “organicism,”
one can find it stretching back deep into the eighteenth century or even ear-
lier.9 Although it reached full bloom in German Romanticism, it was not an
exclusively German phenomenon: M. H. Abrams (1953, 156–​225), in a classic
history of literary organicism, focuses largely on England. But the touchstone
organicist for many of the musicians who drew inspiration from the con-
cept, including Babbitt’s primary musical influences, was Goethe.10 Goethe’s
scientific writings, and particularly his writings on the development and
typology of plants—​although intended as natural philosophy—​furnished
an aesthetic paradigm in which a piece of music could be understood as
organic—​as representing or embodying, in an idealized form, characteris-
tics of a living being. “The linear progression,” Schenker (1979, 44) writes
of the essential contrapuntal prolongation of his mature theory, “shows the
eternal shape of life—​birth to death. The linear progression begins, lives its

9 The term “organicism” also commonly denotes an outlook on metaphysics—​the belief that the

universe, or constituent substances of it, has properties of an organism, such as properties of mind.
This belief long predates organicist aesthetics. As Babbitt did not share this metaphysical outlook, it
will not be discussed here, but many organicist music theorists, including Schenker, also subscribed
to organicist metaphysics. Likewise, organicist political theory, while central to the general history
of organicism (and, again, to Schenker’s own organicist views), is not a part of Babbitt’s outlook.
On Schenker’s metaphysics, see Pastille 1995a; on his politics, see Clarke 2007, Ewell 2020, and
Schachter 2001.
10 On Schoenberg and Goethe, see Neff 1993. On Webern and Goethe, see Cox 2004, Moseley

2017, and Webern 1963. Schenker’s organicism and its antecedents have received much discussion
in recent years, including in Cherlin 1988, Duerksen 2008, Hubbs 1991, Keiler 1989, Korsyn 1993,
Morgan 2014, Parkhurst 2017, Pastille 1984, 1990, Snarrenberg 1994, and Solie 1980. The complex
debates regarding what organicism meant for Schenker, what aspects of his thought it affects, how
it intersects with his many other influences, how or whether the concept developed for him over
the course of his career, and from whom he got the idea are well beyond the scope of this study. But
as Peles 2001, 187 points out, it would be a mistake to reduce Schenker to a “cartoonish . . . cultural
throwback to a Goethean Naturphilosophie,” even if Goethe is unquestionably one of his influences.
As with Babbitt, Schenker’s organicism was infused with more contemporary concerns.
6 Thinking In and About Music

own existence in the passing tones, ceases when it has reached its goal—​all as
organic as life itself.”
One of the essential insights of Goethean organicism concerns the rela-
tion of parts and wholes. Goethe (1962, 366) describes, “From first to last,
the plant is nothing but leaf, which is so inseparable from the future germ
that one cannot think of one without the other.” The original conception—​
the leaf—​grows outward, “repeating, recreating /​In infinite variety” as “each
leaf elaborates the last,” until “at length attaining preordained fulfillment”
(Goethe 2009, 2). As the leaf develops into infinite variety, it takes on the
characteristics of the many distinct organs of a mature plant. But the orig-
inal conception remains: “the various plant parts developed in sequence are
intrinsically identical despite their manifold differences in outer form” (56).
A significant aesthetic consequence for musicians and music theorists who
subscribe to organicist views is therefore a conception of the interrelatedness
of the whole composition and its parts: a search for an explanation of what
Goethe calls “the harmony of the organic whole,” or its realization in music
(quoted and translated in Pastille 1990, 32).
The plot here involves not only a finished whole composed of self-​similar
parts but also a story about how the whole developed. As self-​evident as this
temporal aspect may be in a theory of plant formation, when translated into
the realm of music, it means that an organicist account of a piece is to a signif-
icant degree an ontology: a fictionalized reconstruction of how a piece came
to be, conceptually.11 Furthermore, there are ascriptions of agency here: the
initial impulse is not a passive element that is “repeated and recreated” by an
exterior force; it is actively recreating itself. It contains characteristics that
motivate its own development.
Musical organicism involves the identification of an initial impulse that can
stand in for Goethe’s leaf. For many authors—​including a number significant
to Babbitt, such as Schoenberg (see Neff 1993, 414–​18), Anton Webern,12
Robert Handke, and Reinhard Oppel13—​the organic source is a theme

11 Monahan 2013, 364n.74 points out the agency ascribed to a “fictional composer” in much

Schenkerian analytical prose. Babbitt’s analyses also frequently rely on this device. For a vivid ex-
ample, see Babbitt 1987, 137–​43.
12 Webern 1963 cites Goethe repeatedly, linking his organicist theories not only to general the-

matic or motivic conceptions but also to the twelve-​tone series specifically; see esp. 40–​41 and 53.
13 Babbitt reports that his first encounter with analytical prose was an accidental stumble across

Handke 1909, discovered left on a table at the New York Public Library in early 1934. This led him
shortly to Oppel 1921. These articles “change[d]‌the course of [his] life”: “They made it possible for me
to realize that thinking about music can transmute into thinking in music” (Babbitt 1991a, 126, em-
phasis original). And the consequences for his own music were profound: “We became convinced that
the coherence of [new, atonal] music was very sensitive to, and dependent on, its initial conditions. And
 On Milton Babbitt, Schenkerian 7

or a motive.14 The theme or motive is generally introduced early in a piece, and

the remainder of the piece can thus be understood as the theme or motive’s
unfurling. This version of organicism can also be found in Schenker’s earlier
writings. In Harmony, for instance, he proposes an equation: “In Nature: pro-
creative urge → repetition → individual kind; in music, analogously: procre-
ative urge → repetition → individual motif ” (1954, 6–​7). Although motivic
association was generally not granted the same prominence in Schenker’s
later writings as it received in Harmony, it would continue to be critically im-
portant to his analytical technique for the rest of his life. As will be discussed,
Babbitt was not immune to motivic organicism: he uses the framework to dis-
cuss motivically based music such as that of Béla Bartók and Edgard Varèse.
But the dominant organicist paradigm of Schenker’s last decade—​the
theory of Free Composition, the theory that influenced Babbitt’s compo-
sitional technique—​was not the temporal development of an unfolding
motive but the hierarchical development of a background through suc-
cessive elaborations. A piece of music is said to begin with an initial tone
and its Naturklang—​the “chord of nature” defined by the tone and its first
five harmonic partials. The Naturklang unfolds into one of the three forms
of the Ursatz and from there, through a long sequence of nested voice-​
leading transformations, the piece gradually comes into view. In this ver-
sion of organicism, the temporal and agential metaphors remain. “Earlier”
levels appear earlier in the presumed hierarchical process—​closer to the
Naturklang and the Ursatz—​while “later” levels appear relatively shortly
before the realization (Ausführung) of the completed composition.15 Later

what in traditional terms is a more explicit statement of initial conditions than a fugue subject? This
view is exactly what Handke and Oppel were adumbrating in very restricting and restrictive terms”
(126). The first sentence of Oppel 1921 demonstrates the thematic organicism of these articles: “Das
Lebensfähigkeit einer Fuge ist von der Gestalt ihres Themas abhängig” (10) (“A fugue’s ability to live
depends on the shape of its subject,” translation my own). Handke’s focus on the “linear principle” (“Das
Linearprinzip”) is especially interesting in light of Babbitt’s interest in Schenker, to whose work he was
introduced just a few months later. Handke uses this (vaguely defined) principle to describe not only
thematic development but also harmonic succession and the organization of subject entries.

14 This thematic conception of organicism proved remarkably durable. Reti 1951 is a prominent,

extended example.
15 Many modern Schenkerians replace the temporal metaphors with spatial metaphors: levels

closer to the Ursatz are “higher” or “deeper,” farther from the “surface” (see Snarrenberg 1994, 45–​
49). As Arndt 2016, 92–​93 notes, even the term “hierarchy,” with its connotations of a fixed and static
structure, is a term native to American Schenkerian practice. Babbitt uses temporal metaphors in
his discussion of Schenkerian hierarchy and other, analogous hierarchies more frequently than most
American commentators, although he does use the term “surface” to refer to the completed composi-
tion and the more immediately available relationships contained within it. This usage of “surface” will
be retained in the present volume.
8 Thinking In and About Music

levels are described as motivated by earlier levels: as Carl Schachter (1981,

122) describes, “Schenker often writes about the levels as if they were ani-
mated by kinetic impulses that travel from one level to another; thus the fore-
ground contains the ‘goals’ to which the earlier levels lead” (see also Schenker
1979, 5). The result, in Schenker’s (1996, 18) formulation, is that “Originating
in the background, marching forward through the middleground, the strata
of composing-​out multiply all the way to the diminutions of the lower order
in the foreground.”
In addition to the hierarchical reorientation, a critical difference between
Schenker’s understanding of organicism and that of the aforementioned the-
matic organicists is the source of the initiating organic impulse. Schenker’s
hierarchy begins with a natural given—​which is to say, with the universal
and, therefore, the generic. All pieces share the Naturklang, and most share
one of two common Ursätze. For a thematic organicist, the organic impulse
is composed and is therefore a distinguishing aspect of the piece of which
it is a part. Roger Sessions ([1938] 1979, 258), writing during a time in
which he was mentoring Babbitt, criticized Schenker’s concept of a generic
Ursatz, finding that “the interest of these or any other works begins precisely
at the point where their individual qualities begin to appear and to grow in
an inevitable manner. It is only at this point that organic life may be said to
begin.” Babbitt also criticized an aspect of this model: he vigorously rejected
Schenker’s appeal to natural phenomena such as the overtone series (see
Babbitt [1961a] 2003, 80–​81). But Babbitt, unlike Sessions, could support a
generic hierarchical source, such as an Ursatz or a twelve-​tone series, that can
be construed as the earliest layer of any number of different compositions.
Before diving into the organicist claims in Babbitt’s own writings, a point
about the function of such rhetoric should be clarified. For Babbitt, as for most
Americans who have adopted organicist rhetoric, organicist claims about
temporal development, agency, holistic unity, and coherence are metaphors.
Using the form for the categorization of conceptual metaphors introduced in
Lakoff and Johnson (1980), some of his more common tropes can be charac-
terized as elaboration is growth (simpler, prototypical stages of a work
develop into more complex ones through a dynamic process), association
reflects genealogy (similar events arise from a common source), and the
commonplace states are locations (e.g., there are “points of structural
origin”). Metaphors such as these, all typical of the organicist tradition, are
widespread in Babbitt’s prose and speech.
 On Milton Babbitt, Schenkerian 9

For Schenker, though, such claims are conceivably not metaphorical, and
this constitutes a significant difference between the thinkers. Bryan Parkhurst
(2017, 73) argues that Schenker’s organicism is “methodological rather than
rhetorical” (emphasis original). Much nineteenth-​and early-​ twentieth-​
century continental European organicism can be understood similarly, and
the transference of organicist concepts from an ideological environment
that fostered their interpretation as methodological to a twentieth-​century
Anglo-​American culture in which they could, at most, be applied metaphor-
ically, is a critical juncture in the history of musical thought.16
But to label Babbitt’s organic rhetoric metaphorical is not to thereby di-
minish it. Indeed, it raises perhaps even more urgently the questions of
what organicist metaphors accomplish for Babbitt: what he intended by
them, what he learned from them, and the ways in which they guided him.
Furthermore, because Babbitt’s Schenkerian metaphors are of a piece with
distinctly non-​rhetorical borrowings from Schenker—​such as the construc-
tion of serial arrays on the model of Schenker’s vision of tonality—​taking
Babbitt’s metaphors seriously provides a way toward understanding these
other facets of Schenker’s influence.
Principally, Babbitt’s organicism can be understood as working toward
two complementary aims, one explanatory and one aesthetic. Regarding the
explanatory aspect, Babbitt’s invocation of organicism resembles Kendall
Walton’s ([1993] 2015) conception of “prop oriented make-​believe.” Prop
oriented make-​believe is an imaginative exercise undertaken to help illu-
minate aspects of “props”—​any real-​world items—​under consideration.
One of Walton’s examples involves using the familiar visualization of Italy as
a boot to facilitate locating the city of Crotone, which appears in the boot’s
arch (176–​78). The act of imagining Italy as a boot, or the description of it
as such, is undertaken not because one has any particular interest in visu-
alizing footwear. Rather, the metaphor—​the “make-​believe” of imagining
Italy as a boot—​is a descriptive aid. The metaphor’s value lies in its heuristic
usefulness.
Similarly, Babbitt’s use of organicist metaphor serves, in part, a heuristic
purpose. Organicism provides an explanatory shorthand for the part-​part
and part-​whole relationships he found necessary for musical composition

16 For assorted perspectives on this transference, see Derkert 2008, Rothstein 1986, and

Snarrenberg 1994.
10 Thinking In and About Music

and comprehension. But there is a further, aesthetic component to this im-

aginative exercise. While musical value judgments rarely appear in Babbitt’s
prose, it is clear from the examples he chooses to analyze, the features he draws
attention to within them, and above all the compositional consequences he
derived from his theorizing and analysis that he also sought to promote or-
ganicism as an aesthetic ideal. An organicist analysis, thus, is a sort of test.
Although the act of listening remains the final arbiter (Babbitt 1998a, 25),
organicist analysis guides investigation into a piece’s comprehensibility and
aesthetic richness. Accordingly, Babbitt’s youthful encounter with Schenker’s
writings and students gave him more than a suggestive method for analyzing
the tonal masterworks. It framed his sense of what music had achieved and
what possibilities were yet to be explored.17

Babbitt and Schenker

Babbitt appears to have first encountered Schenker’s work in 1935, at the

age of nineteen, in his first composition lesson with Roger Sessions. They
examined the analysis of Beethoven’s Sonata in F Minor, Op. 2, No. 1 in Der
Tonwille (Babbitt [1985b] 2003, 391).18 As the decade wore on, the emigra-
tion of many of Schenker’s pupils and disciples—​particularly Ernst Oster
and Oswald Jonas—​gave Babbitt further contact with the Schenkerian tradi-
tion.19 By the time of Babbitt’s first published musical essays, in the late 1940s
and ’50s, the influence of Schenkerian thought is fully in evidence. Babbitt’s
review of Felix Salzer’s Structural Hearing (1952) provides his most explicit
early appraisal of Schenker’s achievement. The following passage is indicative
of his thinking at the time:

17 A comparison might be drawn to Schenker’s own aesthetic judgments, which similarly, and

much more explicitly, conflate organicism and value. But one should resist pushing this analogy too
far. As discussed in Pastille 1995b, Schenker’s value judgments are inextricable from his theory of
genius, which Babbitt did not share. Moreover, while Babbitt’s writings suggest an organicist system
of valuation, as a teacher Babbitt was ideologically welcoming. Consider the music of students of his
as diverse as Stephen Sondheim, Stanley Jordan, Donald Martino, Paul Lanksy, Laura Karpman, and
Tobias Picker (see Hilferty 2011 for interviews with several of them).
18 Although this seems to have been Babbitt’s first direct encounter with Schenker’s writings, his

interest had already been piqued by Citkowitz (1933) 1985, Sessions (1935) 1979, and Weisse (1935)
1985; see Babbitt (1985) 2003, 391 and (1999) 2003, 476. Schenker’s analysis of Beethoven’s Op. 2, No.
1 appears in translation in Schenker 2004.
19 Oster and Jonas both emigrated shortly after the Anschluss. On Babbitt’s contact with them, see

particularly Babbitt (1999) 2003, 478–​80.

 On Milton Babbitt, Schenkerian 11

Schenker’s analysis originated in aural experience, and the Urlinie is, at least
indirectly, of empirical origins. On the other hand, it is (and this is merely
an additional merit) completely acceptable as an axiomatic statement
(not necessarily the axiomatic statement) of the dynamic nature of struc-
tural tonality. Stated in such terms, it becomes the assertion that the tri-
adic principle must be realized linearly as well as vertically; that the points
of structural origin and eventuation must be stabilized by a form of, or a
representation of, the sole element of both structural and functional sta-
bility: the tonic triad. It asserts that melodic motion is, triadically, purely
diatonic (of necessity, since any other triadic motion is, at least relatively,
triad-​defining, and thus establishes multiple levels of linear motion, rather
than a single, directed motion); that a work of music ends organically, not
merely temporally. (Babbitt [1952] 2003, 23)

This passage has been cited as representing the vast gulf between Babbitt and
Schenker’s philosophical orientations, and indeed it does that.20 The defini-
tion of the Urlinie as an axiom in a logical system, rather than the natural or
even spiritual ideal Schenker intended it to be, is a stark difference indeed.
It is reflective of Babbitt’s general theoretical project at the time, which was
in large part concerned with the rational reconstruction of another musical
system; namely, the twelve-​tone system, as will be discussed shortly. It is a
signal that Babbitt intends to bring the explanatory power of philosophical
logic to Schenkerian theory. But this statement also represents more than
that: it reveals a deeper continuity underlying the new rhetoric.
Notable among the extended list of “assertions” of the Urlinie-​axiom is the
claim that “a work of music ends organically, not merely temporally.” This
claim is distinct from the rhetoric of axiomatic logic, but it is not an out-
lier in Babbitt’s discourse. Direct references to “organisms,” “kernels,” and
other traditional organicist substantives appear only occasionally in Babbitt’s
writings, but the general concepts underlying those characterizations—​most
notably, an attribution of agency to initial musical assumptions that has de-
termining force over later developments—​remain. What has changed is
the source of the initial assumptions. By de-​naturalizing the source of the
Urlinie—​by making it a chosen axiom rather than a natural given—​Babbitt
opens the way for this model to be transferred to systems besides tonality.

20 See Muhkerji 2014, 160; Schuijer 2008, 252; and Snarrenberg 1994, 50; also Berry 2016, 174–​77

for broader context.

12 Thinking In and About Music

Stripped down, the basic model that Babbitt borrows from Schenker is
that a complete, unified, and coherent piece of music can be characterized
as arising from the progressive hierarchical development of an initial “as-
sumption.” In a tonal piece, that initial assumption is, following Schenker,
the Urlinie, and the means by which the Urlinie develops are, for the most
part, the various transformations of Schenkerian hierarchy—​the most sig-
nificant emendation Babbitt wished to make to a Schenkerian vision of to-
nality was an increased attention to motivic and pitch-​class association (see
Babbitt 1987, 137–​43; [1999] 2003, 479–​80). But Babbitt applies this model
to a wide variety of music—​tonal, atonal, and twelve-​tone. Consider the fol-
lowing passages from Babbitt’s 1949 essay on Bartók’s string quartets. He
lauds “Bartók’s concern for the total composition, and the resultant evolu-
tion of the maximum structure from a minimum assumption” ([1949] 2003,
2). The initiating kernel in Bartók’s String Quartet No. 4 (1928) is not any
sort of tonal Urlinie: Babbitt calls it a “thematic assumption”; essentially, it is
a motive. And, “From this thematic assumption arises Bartók’s polyphony”
(3). The total composition is thus described as the result of “successive
elaborations” of the original thematic assumption. Echoing Schenker’s (1979,
130) claims that large-​scale formal patterns arise from tonal transformations
(“all [forms] have their origin in, and derive from, the background”), Babbitt
([1949] 2003, 7) continues, “Bartók’s formal conception emerges as the ul-
timate statement of relationships embodied in successive phases of musical
growth.”
During the course of this “evolution” or “growth”—​note the retention of
these characteristically Schenkerian metaphors—​various pitches may be-
come “central tones” within their surrounding context; this can result in the
sense of polytonality. But Babbitt rejects the term “polytonality,” finding it
“self-​contradictory” (3). Although he does not explain the contradiction, it
appears to be that, for him, and again reflecting a Schenkerian orientation,
the term “tonality” denotes a strict hierarchy built around a single tonic,
and the tonal system instantiates a theory of perception in which wholes
are reducible to singular referential sources. Therefore, polytonality is im-
possible: a piece could not be perceived as a unified, whole piece of music
and have multiple tonics. It is also, he finds, mistakenly applied to Bartók: the
sense of polytonality in his music is said to come not from the genuine coin-
cidence of multiple tonal hierarchies but as an outgrowth of motivic develop-
ment. This line of reasoning is echoed in Babbitt’s ([1952] 2003, 29) critique
of Salzer: Salzer’s prolongational readings of Bartók are “disappointing”
 On Milton Babbitt, Schenkerian 13

because he did not realize that “in Bartók’s music the motivic is structural
and serves to project the essential structural motion.”
This sort of motivic organicism underlies all of Babbitt’s analyses of
works that he terms “contextual”21—​that is, neither tonal nor serial. In
each, including analyses of Bartók, Varèse, Igor Stravinsky, middle-​period
Schoenberg, and others (see Babbitt [1964] 2003; [1966] 2003; 1987, 157–​
62), some initial assumption—​often a motive, but occasionally a sonority or
even simply an interval22—​is shown to develop, usually hierarchically. The
term “contextual” itself is interesting: Babbitt draws attention to the degree
to which pieces drawing from neither tonality nor serialism must define their
own assumptions and develop them in individual ways.23 But despite this
focus on each piece’s particularities, the basic model for Babbitt’s description
of contextual compositions is consistent across his discussions of highly di-
vergent composers and throughout his long career.
Unsurprisingly, given his compositional interests, it is in Babbitt’s writings
about the twelve-​tone system that we see his fullest description of any musical
system. A more comprehensive look at Babbitt’s compositional techniques
will be given subsequently; for now, I will simply observe a few ways in which
Babbitt conceived of the twelve-​tone system in Schenkerian terms. That he
viewed the twelve-​tone system as essentially analogous to Schenker’s vision
of the tonal system is evident in his earliest descriptions of it. Although much
of his writing on the twelve-​tone system is concerned with outlining var-
ious technical possibilities of mod-​12 pitch-​class space—​combinatoriality,
for instance—​a great deal of his description of the system and, even more,
the potential compositional applications of the system have evidently been
guided by Schenkerian thought.

21 The term “contextual” receives its most thorough discussion in Babbitt 1987 (see 9 and passim),

but it occurs throughout his writing.

22 See, for instance, Babbitt’s ([1964a] 2003, 150) discussion of the minor third in Stravinsky’s

Symphony of Psalms (1930).

23 The term “contextual” is apparently borrowed from the technique of literary analysis called

“contextual criticism,” as demonstrated in the work of Eliseo Vivas and Murray Krieger (see Krieger
1956): the denotation of the term “contextual” for Babbitt and the literary contextual critics is nearly
identical, signifying in both cases a focus on the internal characteristics of a work. Contextual criti-
cism was a technique practiced by the literary movement known as New Criticism, which is aligned
in certain broad respects with Babbitt’s analytical approach (as noted in Girard 2007, 318–​20; see
also Berry 2016, 177–​80). W. K. Wimsatt, Jr. and M. C. Beardsley (1946), prominent New Critics,
developed the concept of the “intentional fallacy,” which Babbitt frequently cited (e.g., in Babbitt
[1952] 2003, 24; [1970] 2003, 238). Wimsatt (1954) 1958, which includes two essays coauthored with
Beardsley, was in Babbitt’s personal library, the contents of which are recorded in the Milton Babbitt
Collection.
14 Thinking In and About Music

Consider his description of the three forms of the Urlinie as “the first
manifestations of the extension of the triadic principle, [which] serve as
the framework within which the unique aspects of the individual compo-
sition assume shape and significance during the unfolding from the Ursatz
background through the phases of the middleground to the foreground”
([1952] 2003, 23–​24). Compare that with his description of the choice of a
twelve-​tone series, the “unique compositional stage represented by the fact
of the set, the element with regard to which the generalized operations of
the system achieve meaning, and from which the progressive levels of the
composition, from detail to totality, can derive” ([1955] 2003, 44–​45).24 That
is, just as Schenker’s genealogy of a tonal composition originates with the
Ursatz and unfolds therefrom, so Babbitt’s genealogy of a twelve-​tone com-
position begins with the series. Later developments in a piece are shaped by
the possibilities inherent within the chosen series. Just as Schenker (1979,
3) proposes that “the origin of every life . . . becomes its destiny”—​and thus
the conceptual origin of music determines its ultimate development—​so for
Babbitt ([1961b] 2003), “set structure”—​the conceptual origin, in his telling,
of a twelve-​tone composition—​“is a compositional determinant.”
The word “determinant” deserves explication. In Babbitt’s music, the
path from the series to the composition is neither direct nor specified in any
particulars by the operations of the twelve-​tone system. The fact that two pieces
share a series does not imply that they have much more in common than, for
instance, two pieces with a 3-​line Urlinie, and Babbitt frequently reused series
in pieces that are otherwise completely distinct. Once, on discussing the idea
of a piece’s “form” “aris[ing] out of the specific implications of the set,” Babbitt
([1955] 2003, 47n.28) explained, “Naturally, this does not mean to say that
a given set uniquely implies a given composition, but rather that a given set
defines, in these terms, certain general possibilities which are uniquely asso-
ciated with this set.” The idea, in other words, is more one of construal than
cause—​of defining possibilities that enable associations. A piece can, and
should in this telling, be understood as determined by its series because of the
explanatory, and perhaps aesthetic, benefits of such an understanding, even
though this determination is not compositionally binding. In this respect,
Babbitt’s claims about determination are like other causal organicist claims.
In the “prop oriented make-​believe” of organicist metaphor, to speak of Y as

24 “Set” is Babbitt’s term for what most authors call a “series” or “row.”
 On Milton Babbitt, Schenkerian 15

“arising from X,” “originating in X,” “generated by X,” “derived from X,” “deter-
mined by X,” and so forth is not to make a factual claim about Y’s origins, but to
suggest that it is fruitful to imagine a causal relationship between X and Y. And
note: imagine, not hear. Organic determinism is a matter of construal, not per-
ception. A listener may construe a work as presenting causal relationships if
doing so has explanatory benefits. A composer—​perhaps through sensitivity
to “set structure”—​might create a relationship about which such a causal attri-
bution is plausible.25
The specific means by which a series might be taken as implying further
development will be detailed in a moment, and the many ways in which
Babbitt’s music simultaneously resists this sort of hierarchical explanation
will be discussed at length in later chapters. But Babbitt continued to view the
twelve-​tone system as an analogue to Schenker’s vision of tonality as late as
1997: “[the twelve-​tone system] is formulable at a fairly general and deep level
as the replacement of the analytical and synthetic notion of prolongational
parallelism (‘Schichten’) by that of transformational parallelism” (132).26

The Implications of the Series

In his telling, as discussed in the previous section, hierarchical develop-

ment in Babbitt’s music begins with the “unique compositional stage”
represented by the choice of the initial series. Babbitt’s series are composed
of either six or twelve distinct pitch classes; the six-​note series are invariably
all-​combinatorial hexachords and, when they appear, are generally linked
with their complements to form complete twelve-​tone series.27 Elaboration
outward from the initial series begins by seizing on the basic fact that it is
a twelve-​tone series; that is, it features the complete aggregate of all twelve
pitch classes: “The minimal configuration identified uniquely within the

25 Dubiel 1997, 40 provides a related interpretation of Babbitt’s usage of “determinant,” albeit with

different emphases.
26 This sentiment is reiterated in Babbitt 1998a, 29–​30. While Babbitt drew attention to similari-

ties between tonality and twelve-​tone music in the terms described here, he was also quick to note
differences between the systems, particularly early in his career; see Babbitt (1960) 2003, 55–​56.
27 Analytical practice, by Babbitt and others, has been inconsistent about the description of

hexachordal series. In some cases, authors consider them as part of a twelve-​tone series composed
of two successive iterations of the hexachords. In other cases, the hexachordal series themselves
are taken as the basic hierarchical source. This book will consistently use the latter approach, as it
observes the principle that the series of a piece should be identified with that piece’s deepest-​possible
hierarchical starting point.
16 Thinking In and About Music

twelve-​tone system is the collection of elements of the chromatic scale, and

it would thus seem desirable to consider this collection the basis for the
construction of the fundamental unit of progression” (Babbitt [1946] 1992,
75; see also Babbitt 1987, 68). Accordingly, the first several stages of hierar-
chical development involve building further aggregates out from the initial
series, eventually resulting in “arrays” that present four distinct techniques
of aggregate construction. Just as every hierarchical level in a Schenkerian
analysis is defined by one of the techniques of tonal transformation, every
hierarchical level in one of Babbitt’s arrays is defined by the completion of
an aggregate.
The contrapuntal lines that are combined in these arrays come from one
of two categories of sources. One strategy, seen in Figures 1.1, 1.2, and 1.3,
presents derived series constructed of segments of an underlying series.
The other, seen in Figure 1.4, presents lines that are simply members of a
series class transformed using the basic twelve-​tone operators (transposi-
tion, inversion, retrograde, and retrograde inversion). For the most part,
Babbitt used the first strategy from 1948 to 1960, the second from 1961
to 1979, and used both strategies, sometimes in combination, in his final
decades.28
Glosses (1988) is for children’s choir.29 Perhaps for that reason, it presents
an appealingly didactic example of the first strategy, the use of arrays gen-
erated from segments of a series. The piece begins with the second altos,
as shown in Figure 1.1, who introduce the work’s six-​note series. The piece
alternates between passages like the opening, which presents straightfor-
ward iterations of the series, and passages like that shown in Figure 1.2, based
on “trichordal arrays” derived from the series.
Trichordal arrays are generated from the segmental trichords of the se-
ries for the piece of which they are a part. This understanding of the origin
of trichordal arrays is propounded throughout Babbitt’s writings and the
broader analytical literature.30 It also reflects his compositional process: as
is apparent in Babbitt’s sketches, Babbitt typically begins composing a
trichordal-​array piece by sketching its series and working out the orderings

28 Mead 1994 justifies the division of Babbitt’s output into three compositional “periods” on the

basis of these tendencies. For more on Babbitt’s arrays, see particularly Mead 1994; Dubiel 1990b,
219–​35; and Babbitt (1974) 2003.
29 For more on Glosses, see Maggart 2017b.
30 See, e.g., Babbitt (1976b) 2003; Babbitt 1987, 26–​30 and 85–​97; and Mead 1994, 25–​30 and 54–​

123. Babbitt 1986a, 82 connects derivation to Schenkerian hierarchy, in that transformations appli-
cable to full series forms are applied individually to trichords.
Figure 1.1. Glosses, mm. 1–​3, with trichord orderings indicated

Figure 1.2. Glosses, mm. 28–​32, with trichord orderings indicated

Figure 1.3a. Trichordal array for Glosses, mm. 28–​45. Dashed lines indicate
partitioning. Bar lines indicate aggregates. Boxes indicate further aggregates
formed between pairs of lines.

Figure 1.3b. Secondary set in the Soprano 1 line across mm. 31–​59 of Glosses
18 Thinking In and About Music

of its segmental trichords before combining those trichords into arrays.

Since the series that is the source of the trichordal arrays is ordinarily not
explicitly stated during those arrays, it will be referred to as the arrays’ “un-
derlying series.” As can be seen in Figures 1.1, 1.2, and 1.3a, the beginning of
Glosses’ trichordal array is generated from the first two segmental trichords
of Glosses’ series.
The segmental trichords of the series define what I will call “trichord
orderings” that can be extracted from the series. Like miniature series, tri-
chord orderings can be transposed, inverted, and/​or retrograded, forming
the basis of derived series. The original (“S”) form is that presented in the
underlying series; it is characterized by its pitch-​class intervals. As Figure
1.3a demonstrates, the lines of trichordal arrays present derived series like
that of Webern’s Concerto for Nine Instruments, Op. 24 (1934)—​that is,
each derived series presents trichords in each of the four ordering possi-
bilities.31 Consider the Soprano 1 line: it begins with <−5, −2>, the second
trichord ordering introduced in the piece’s series, before presenting its
retrograde inversion (<−2, −5>), inversion (<+5, +2>), and retrograde
(<+2, +5>).
Trichordal arrays develop both the ordering possibilities inherent
within the underlying series and the harmonic content of the series. Glosses
demonstrates this principle, too, with didactic clarity. The C-​type hexachord
of the series,32 {F, G, A, B♭, C, D}, is also the first hexachord of the trichordal
array, heard in isolation in the altos. Glosses helpfully precedes the trichordal
array with the series such that the origin of the array can be readily heard.
Many trichordal-​array pieces do not state their series at the outset; indeed,
in some pieces, the series is never stated. Nonetheless, given the orderings
and content presented in the arrays, the underlying six-​or twelve-​note se-
ries can usually be inferred. As Glosses demonstrates, the sequence in which
trichord orderings are developed generally reflects the sequence in which
those trichord orderings appear in the series: Glosses’ array presents first
<+2, −5> and then <−5, −2>, the first two trichord orderings of the series in
order. Although not every piece with trichordal arrays is as straightforward
as Glosses, the principle holds broadly.

31 On this aspect of Webern’s Concerto, see Babbitt 1987, 23.

32 Throughout this book, I will use the classification of all-​combinatorial hexachords introduced
in Martino 1961. The A-​type hexachord is the set class (012345), the B-​type (023457), the C-​type
(024579), the D-​type (012678), the E-​type (014589), and the F-​type (02468t).
 On Milton Babbitt, Schenkerian 19

The array for Glosses, as is typical for a trichordal array, demonstrates

four techniques of aggregate construction. Each of the four lines presents a
derived series, each line is paired with another line such that the separate
hexachords of each create aggregates (boxed in Figure 1.3a), and all four lines
together are arranged to create surface aggregates (delineated by bar lines in
Figure 1.3a). The surface aggregates are each distinguished by partitioning
(indicated with dashed lines in Figure 1.3a): every eight aggregates, the array,
also in typical fashion, cycles through the eight partitions of four trichords
into one or two parts (or, equivalently, through the fifteen possible combin-
ations of four trichords arranged into complementary pairs and a tutti). As
the array continues, each line is followed by another line such that the last
hexachord of the first line and the first hexachord of the second line create
yet another aggregate, called a “secondary set,” as shown in Figure 1.3b.33
The result of this process, as Babbitt ([1974] 2003, 317) puts it, is that the
“emergent history of environments” of atomic local events is “secured by the
complex of imbricated and concatenated aggregates.” And the “set structure”
of the underlying series shown in Figure 1.1 has “determined” the array in
two distinct ways: the combinatorial possibilities of the hexachord facili-
tate the array’s aggregate construction, and the subset content of its first two
trichords determines the array’s linear content. The ability of the opening
two trichords to generate the piece’s C-​type hexachord is a crucial adjunct to
this process.34
Figure 1.4a presents the beginning of the array for My Ends Are My
Beginnings (1978).35 In some respects, this array is much like that for
Glosses: it too presents the same four techniques of aggregate construction,
also dependent on the combinatorial structure of its series, including contin-
uation via secondary sets (as shown in Figure 1.4b). But there are two crucial
differences. First, the lines of the array are no longer derived from the ini-
tial series; they simply are forms of that series, transformed using the classic
twelve-​tone operators. Second, the partitioning is varied. Every aggregate

33 As secondary sets cut across linear aggregates, they are effectively extra-​hierarchical. Babbitt’s

music contains many extra-​hierarchical connections like this: important features (such as aggregates)
or relationships (such as cross-​references) that bridge hierarchical boundaries. This reflects Babbitt’s
general interest in saturating his music with significant relationships. The Schenkerian analyses that
appear to have inspired Babbitt’s arrays also contain extra-​hierarchical connections of various kinds,
such as motivic relationships (see Cohn 1992) and prolongations that fill gaps between hierarchically
distinct events, such as branches of an interruption (see Goldenberg 2012).
34 A trichordal “generator” of a hexachord is a set class such that two of its members can be com-

bined to form that hexachord. See Babbitt (1955) 2003, 44.

35 The full array for My Ends Are My Beginnings is given in Mead 1994, 273.
20 Thinking In and About Music

Figure 1.4a. First four aggregates of the array for My Ends Are My Beginnings.
Bar lines indicate aggregates. Boxes indicate further aggregates formed between
pairs of lines.

Figure 1.4b. Secondary set in the beginning of the uppermost line of My Ends
Are My Beginnings

of Figure 1.3a includes four trichords, while the aggregates of Figure 1.4a
all present distinct partitions. The partitioning of each aggregate in Figure
1.4a is indicated beneath the example: “(4231),” for instance, indicates that
two lines contribute four-​note segments, one line contributes a three-​note
segment, and one line contributes a single note. Figure 1.4a presents the be-
ginning of an “all-​partition array,” an array that proceeds through every pos-
sible means of partitioning the aggregate for however many parts happen
to be in that array. Since My Ends Are My Beginnings has four lines, the full
array presents thirty-​four aggregates, ranging from the maximally even (34)
partition, in which every line contributes three notes, to the maximally un-
even (121) partition, in which a single line contributes the entire aggregate.
For arrays with more lines, there are more possible partitions. In addition
to four-​line, thirty-​four-​aggregate arrays, the most common all-​partition
arrays in Babbitt’s music include twelve-​line, seventy-​seven-​aggregate arrays;
six-​line, fifty-​eight-​aggregate arrays; and, somewhat differently, twelve-​line
 On Milton Babbitt, Schenkerian 21

arrays using only the fifty-​eight partitions whose segments are no longer
than six notes.
The array counterpoint shown in Figures 1.3a and 1.4a is “partially
ordered” (see Babbitt [1976b] 2003, 346). The ordering of each line—​
expressing a twelve-​tone series—​is fixed, but the relative disposition of the
various segments found within each aggregate is left completely free from the
perspective of array construction. That is, within the first aggregate of Figure
1.4a, B must precede B♭, which must precede E♭, but G, from another line,
could interject at any point and not violate the principles of the array. The
compositional criteria motivating the arrangement of the various segments
in relation to each other are properly extrinsic to array construction. The
most important of these criteria is the saturation of the surface of the music
with cross-​references to other aggregates of the array or to the piece’s series, a
technique to be discussed shortly.
The principle of the aggregate is carried yet further, in many pieces, with
the translation of pitch structures into rhythm. Most often, from 1960 on, this
is accomplished with the time-​point system, which translates pitch classes
into beat classes and, thus, pitch-​class intervals into time-​point intervals—​
that is, durations, considered in relation to a defined time-​point modulus.36
Babbitt’s explanations for the development of the time-​point system invoke,
characteristically, both organicist and cognitive concerns. The adoption of
the time-​point system as an analogue to the twelve-​tone system is justified
as an attempt “not only to fill some of the holes in my holism but to fur-
ther reduce the admittedly context dependent pitch structure by introducing
the reinforcing redundancy of interdimensional parallelism” (Babbitt 1998a,
32). “Reinforcing redundancy,” in this case, refers to the cognitive theory that
repetition facilitates perception and memory; this will be taken up at greater
length in Chapter 2.
In support of both holistic unity and “reinforcing redundancy,” Babbitt
often uses precisely the same array structure for both pitch and rhythm (Mead
1994, 48). Indeed, the sketches for numerous later pieces reveal that Babbitt
tended to use the same array chart to work out both time-​point and pitch-​
class structure. Figure 1.5 shows the first page of the array chart for None but
the Lonely Flute (1991). None but the Lonely Flute also uses an all-​partition

36 The time-​point system is introduced in Babbitt (1962) 2003. As discussed in Mead 1994, 38–​45

and 54–​123, Babbitt used a number of different rhythmic techniques early in his career. Although
Babbitt used the time-​point system in most pieces after 1960, its implementation varies dramatically,
particularly before 1982 (see Mead 1987 for a comprehensive treatment).
22 Thinking In and About Music

Figure 1.5. Array chart for None but the Lonely Flute, p. 1 (Milton Babbitt
Collection, Box 11, Folder 1). Used by kind permission of the Estate of Milton
Babbitt.

array, proceeding through the fifty-​eight partitions of the aggregate into six
or fewer parts.37 Those six lines are arranged into three combinatorial pairs.
When realized in the pitch domain, each line-​pair is projected in a discrete

37 If C = 0, the array for None but the Lonely Flute is T of that given for The Joy of More Sextets
5
(1986) in Mead 1994, 278–​79.
 On Milton Babbitt, Schenkerian 23

Figure 1.6. Sketch for None but the Lonely Flute, p. 1 (Milton Babbitt Collection,
Box 11, Folder 1). Used by kind permission of the Estate of Milton Babbitt.

register: the highest line-​pair appears between C6 and B6, the middle line-​
pair between C5 and B5, and the lowest line-​pair between C4 and B4. Figure
1.6 shows the first page of the sketch for the piece, demonstrating that both
pitch-​class and time-​point structure were composed from the same array
chart. The completed score for this passage is shown in Figure 1.7.
The uppermost staff in each system of Figure 1.6 indicates the pitches to
be used in the piece. The lower three staves in each system indicate time
points, with stem direction delineating between the two time-​point lines
contained within each staff. In this notation, pitch-​class intervals corre-
spond to time-​point intervals. Semitones in this piece, as in most of Babbitt’s
late music, correspond to the duration of a sixteenth note, and there-
fore octaves correspond to moduli of twelve sixteenth notes. Each of the
six time-​point lines is articulated by a single dynamic value, spanning the
six dynamics between 𝆏𝆏 and   𝆑𝆑, with higher lines on the page matched to
24 Thinking In and About Music

Figure 1.7. None but the Lonely Flute, mm. 1–​9, with time-​point and pitch-​class
aggregate rhythm marked

louder dynamics. Therefore, as the first time point indicated is F♯ (stemmed

down in the uppermost time-​point line and therefore 𝆑), and the second
time point is G (stemmed up in the uppermost line and therefore 𝆑𝆑), and
the interval between F♯ and G is one semitone, the first sixteenth-​note value
of the piece is 𝆑 and the second is   𝆑𝆑 (although as will be discussed sub-
sequently, the opening sixteenth-​note value is subdivided for the sake of a
cross-​reference). The third time point is C (stemmed down in the lowest line
and therefore 𝆏𝆏). As this C is five semitones from the second time point,
G, the duration between the second and third time points is five sixteenth
notes. (See Figure 1.7 for verification.) The specific pitch classes with which
the time-​point lines are sketched are arbitrary from the perspective of the
 On Milton Babbitt, Schenkerian 25

time-​point system: there is no particular significance to the F♯ beginning the

piece’s time-​point array. Rather, these pitch classes are merely artifacts of the
original array chart shown in Figure 1.5: F♯ and G are the first notes of the
top two lines of the array, and C is the first note of the bottom line.
Despite the fact that pitch and rhythmic arrays express equivalent mate-
rial, they remain independent, both conceptually and in practical applica-
tion. As Babbitt explains, “the rhythmic system is closed, and as its structure
is independent of pitch clarification, it can be applied as independently as the
pitch system.” He continues by noting that the two dimensions are not simply
“coordinat[ed],” they interact “polyphonic[ally]” in “structured rhythmic
counterpoint” (Babbitt [1962] 2003, 131; see also Mead 1994, 48). As can be
seen in Figure 1.7, the time points in mm. 1–​9 of None but the Lonely Flute
complete the first two aggregates of the array while the pitches complete the
first three aggregates. The rhythmic independence between the dimensions,
the fact that they unfold separately and often at different rates, means that
even though both dimensions are drawn from the same array, they generally
do not end up completing the same amount of material. This independence
is thus a primary contributor to the problem of completeness that will be
discussed in Chapter 6.
Having extended the principle of the aggregate inherent in the series into
four distinct levels—​the individual series, the secondary set, the combinato-
rial pair, and the array—​and having established a parallel array system in the
rhythmic dimension, Babbitt seems to have gone as far as he was interested in
going with the aggregate itself. The remaining hierarchical levels are formed
not from aggregates, but from an abstraction of the principle of the aggregate.
Just as the initial twelve-​tone series presents one of every possible pitch class,
the remaining hierarchical levels also present exhaustive lists: lists not of pitch
classes but of various other kinds of musical things. Many of the techniques
with which the principle of exhaustive completion is developed have been
explored by other commentators, most notably Mead (1994). The most im-
portant of these techniques regards the partitioning of arrays: as discussed
previously, both trichordal arrays and all-​partition arrays present exhaustive
lists of partitions. Other common instances of this principle include exhaus-
tive lists of combinations of instruments or registers. Composition for Four
Instruments (1948), which proceeds through the fifteen possible subsets of its
ensemble, is exemplary (see Figure 4.6). But the principle of exhaustive com-
pletion of a list of possibilities extends broadly and manifests quite variously
26 Thinking In and About Music

in Babbitt’s output. Suffice it to say that this principle underlies practically

every aspect of nearly all of Babbitt’s compositions.38 At every level, exhaus-
tive lists are embedded within exhaustive lists. Across six decades of com-
positional output, it is one of few constants. And given Babbitt’s aesthetic
commitments, its systematic benefits are clear. Exhaustive completion gives
Babbitt the means to hierarchically extend the implications of the series all
the way out to the organization of complete works.39

Analepsis and Prolepsis

The principle of exhaustive completion secures the hierarchy essential to

Babbitt’s Schenkerian vision. Nonetheless, the hierarchical logic of the arrays
only loosely determines the surface of Babbitt’s music—​as discussed previ-
ously, the “partial ordering” of Babbitt’s aggregates means that nearly every
partition can be realized in a countless number of different configurations
that all equally satisfy the requirements of array construction.40 The pri-
mary criterion guiding the realization of arrays on the surface of Babbitt’s
music is therefore extrinsic to his systematic hierarchy, but it is related.
His music is nearly saturated with references to hierarchically significant
aspects of the piece: the series itself or segments of the array. For instance,
simultaneities and, especially, simultaneous attacks—​in which each note is
struck together—​tend to realize segments of the series or array.41
The specific techniques guiding which aspects of the piece are chosen for
reference and how the references are to be projected are relatively flexible
until about 1982, after which Babbitt settles into something like a regular
practice; this late technique of cross-​references to the array will be discussed
shortly. In many of these pieces, including those using the late technique of
array cross-​reference, these references are arranged with a consistency that

38 Babbitt’s principle of exhaustive completion of a list of possibilities is commonly referred to as

the principle of “maximal diversity” (e.g., Mead 1994, 19–​20), or, in Babbitt’s (1987, 87) terms, the
“spirit of maximum variety.” Following Dubiel 1992, I prefer “exhaustive completion,” on account of
its precision. Babbitt’s music not only reflects an aesthetic preference for diversity or variety but a sys-
tematic extension of the exhaustive list inherent in a twelve-​tone series.
39 Babbitt (1986a, 82) acknowledges that he borrowed the notion of a “parallelism of process” uni-

fying the large and small scale from Schenker.

40 Dubiel 1997 and Mailman 2019 and 2020 make much of this flexibility.
41 Mead 1994 discusses many of the techniques Babbitt uses to relate the surface and the array.

The principle that simultaneities should reflect the series was stated as early as Babbitt’s dissertation
([1946] 1992, 95–​96).
 On Milton Babbitt, Schenkerian 27

might be called “systematic.” However, even in such pieces, non-​systematic

references to the series or the array abound, in addition to the consistent sys-
tematic techniques: the general principle seems to be that Babbitt wants his
music to reflect the series and array almost as much as possible (barring cer-
tain pronounced exceptions, such as those discussed in Chapter 7). It is to
these inter-​level associations—​this “richness of relation”—​that Babbitt refers
when he expresses his desire to make his music “as much as it can be.”42
As will be discussed, the specific concept of relating the surface and deeper
levels of structure likely stems from Schenker. But a possible source for the
more general idea that a piece should contain as many relationships as pos-
sible is Gottfried Wilhelm Leibniz, the seventeenth-​century German philos-
opher and mathematician.43 At the very least, there is much about Leibniz
that resonates with Babbitt’s thought, and Babbitt apparently recognized this.
Leibniz is the only pre-​twentieth-​century philosopher Babbitt ever cites ap-
provingly. Babbitt calls Leibniz one of his “best friends,” a “colleague,” and
even “a fine guy to be influenced by” (1991a, 132; [1991b] 2003, 442; 1987,
22). In particular, Leibniz is mentioned as an antecedent of Babbitt’s under-
standing of rhythm (Babbitt [1991b] 2003, 442; 1998a, 32). In apparent rec-
ognition of his debt, Babbitt’s Ars Combinatoria (1981) borrows its title from
a logical system detailed in Leibniz’s De Arte Combinatoria (1666). In none
of the cited comments does Babbitt expand on just what it was about Leibniz
he found appealing, and it is therefore difficult to trace specific lines of in-
fluence. I will simply note some striking correspondences between the two
thinkers that provide a possible source for aspects of Babbitt’s technique and
aesthetics.
For Leibniz, harmony increases in direct proportion to the number of
relationships expressing rational “order”—​that is, relationships which are
“cogitable” (which Laurence Carlin [2000, 105] defines as those which are
“thinkable or potentially observable”). Harmony is dependent on variety; it
is the rational “reduction” of variety to unity, a “similitude” encompassing
“dissimilar things” (Carlin 2000, 100–​101). Therefore, Leibniz claims,

Harmony is the perfection of cogitability, insofar as there are cogitable

things. Harmony is when many things are reduced to some unity. For

42 Babbitt (1984) 2003, 386–​87 and 1987, 64 and 183. This interpretation of Babbitt’s phrase is clar-

ified by context in Babbitt 1987, 64.

43 I am grateful to Joseph Dubiel for drawing my attention to the Babbitt-​Leibniz connection.
28 Thinking In and About Music

where there is no variety, there is no harmony. Conversely, where variety is

without order, without proportion, there is no harmony. Hence, it is evident
that the greater the variety and the unity in variety, this variety is harmo-
nious to a higher degree. (quoted and translated in Carlin 2000, 106)

Babbitt likewise sought the “spirit of maximum variety” (1987, 87)—​within

the rational constraints of serial relations. And while Babbitt did not speak
of “harmony” per se as an aesthetic goal of his, the view that a piece should
maximize the number of definable relationships between events is a clear im-
plication of his writings and compositional technique.
The desire that rational relationships should be maximized was, for both
authors, not merely an aesthetic or theological conceit. It was founded in
their views on cognition. Despite the fact that Babbitt’s understanding of cog-
nition was informed by centuries of research following Leibniz, their views
about the cognitive benefits of rational relations are aligned in certain broad
respects. For Leibniz, the desideratum of cogitability means that the various
relations in a harmonious object can be “taken at the same time” (quoted and
translated in Carlin 2000, 107). That is, as Carlin (2000, 107) writes, “har-
mony results from a given collection of entities (in this case relations) when
they are simultaneously considered by a mind.” Babbitt similarly sought a
hierarchy such that the various events of a musical work could be “construct-
ible . . . as a totality”—​conceived of as a unit in the listener’s “constructive
memory” (Babbitt 1998a, 25).
Although Leibniz did not discuss art or music in organicist language, he is
an archetypal metaphysical organicist in that he believed each substance—​
each “monad”—​to be imbued with “perceptions” and “desires” (1898, 418).
Each separate substance, then—​ in our perfect, harmonious universe—​
connects to all else, across time, through orderly relations:

For all is regulated in things, once for all, with as much order and mu-
tual connexion as possible, since supreme wisdom and goodness can act
only with perfect harmony. The present is big with the future, the future
might be read in the past, the distant is expressed in the near. We might
get to know the beauty of the universe in each soul, if we could unfold all
that is enfolded in it and that is perceptibly developed only through time.
(1898, 419)
 On Milton Babbitt, Schenkerian 29

A favorite analytical observation of Babbitt’s is to remark on how the surface

of one passage is organized to recall or predict another passage in the same
work—​“analepsis” and “prolepsis,” as he put it.44 And his own music is re-
plete with such cross-​references. Whether or not Leibniz inspired Babbitt’s
interest in forging relationships across timespans, their visions of an ideal,
trans-​temporal interconnectedness are concordant.
From a technical perspective, the concept that the surface should reflect
significant aspects of array hierarchy appears to have been developed as an
analogue to the late Schenker’s understanding of “concealed repetition,”
adumbrated in Free Composition and latent in many of Schenker’s analyses
of the 1920s and ’30s. In Free Composition, Schenker distinguishes between
“motives,” which are simply “repetition in the foreground,” and the “con-
cealed repetitions” in which a middleground pattern is transferred to the fore-
ground (1979, 99–​100; see also Burkhart 1978). These concealed repetitions
are characterized as the “prime carriers of synthesis” (Schenker 1979, 100),
and the analyses of Schenker’s last fifteen years demonstrate the profound
importance this concept had in his theory. “Motivic” repetition, pertaining
only to the foreground—​although “in its place, beneficial”—​does not have
the same significance. The music in which that technique predominates is “of
the lesser talents” (99).45
Babbitt reflects Schenker’s understanding of the surface’s role in projecting
significant aspects of hierarchy in both his music and his analyses. The
instances of analepsis and prolepsis he describes generally do not connect
surface aspects of both passages; rather, the point of these discussions is to
show that the surface of one passage calls to mind a hierarchically prior as-
pect of another passage. In an example he returned to repeatedly, the registral
configuration—​a “secondary aspect”—​of the opening of Schoenberg’s String
Quartet No. 4 (1936) is taken as predicting a hexachordal area to which the
piece later moves (Babbitt 1987, 64–​67). In Babbitt’s own music, analogously,
cross-​references connect the surface of one passage with deeper aspects of
another.
As mentioned previously, most of Babbitt’s later music uses a consistent
technique for organizing the surface into cross-​references to the array. This

44 Babbitt’s analytical interest in analepsis and prolepsis is noted in Straus 2012, 26.
45 As Burkhart 1978, 159–​62 points out, Schenker 2004—​the first Schenker analysis Babbitt
encountered—​presents vivid examples of motivic parallelism. Did Babbitt realize the compositional
implications of motivic parallelism as early as that first encounter?
30 Thinking In and About Music

technique is exemplified in None but the Lonely Flute and is visible in the
sketches for the piece. In Figure 1.6, the numbers written below the upper-
most staff in each system indicate cross-​references.46 Each time-​point interval
in Figure 1.6, isolated by vertical lines resembling bar lines, is subdivided
into a string of equal note values. The number of equal note values into which
each time-​point interval is subdivided corresponds to the length of the array
segment being referred to. Underscored numbers and numbers in paren-
theses indicate the overall length of the referenced segment: underscores are
used if the final pitch class of the segment is sounded in the reference, and
parentheses are used if it is not. A note’s placement within that subdivision
reflects its position within the referenced segment and is indicated by the
number written below it in Figure 1.6. Therefore, the underscored “2” under
C, the first note of the sketch, indicates that that note is part of a reference
to a two-​note segment of the array in which C is the second note. The first
note of that segment is not referred to here, a circumstance Babbitt signals
using a rest. Therefore, the opening time-​point interval of one sixteenth note
is subdivided into two thirty-​second notes, of which the first is silent and the
second is C. As can be seen in Figure 1.6, None but the Lonely Flute—​like al-
most all of Babbitt’s late music—​is saturated with cross-​references.47
Figure 1.8 details the places in the array to which the cross-​references
in the opening of None but the Lonely Flute refer.48 The first line references
a long sequence of components from the lowermost line of blocks V and
VI of the array, shown in Figure 1.9.49 Cross-​references are typically or-
dered in sequences like this, with references to a number of consecutive

46 See Dubiel 1997, 37–​39; 2008, 135–​39; Leong 2011; and Mead 1983 and 1987, 213–​18 for more

on the technique discussed here (or, in the case of Mead 1983, a closely analogous technique).
47 A related understanding of the twelve-​ tone series, surface phenomena, and cognition can
be found in Schoenberg’s conception of the Grundgestalt. That theory, despite some of its recent
applications, was conceived by Schoenberg in relation to his twelve-​tone music: the Grundgestalt is
the first presentation of the twelve-​tone series (the “basic set”), including its registral and rhythmic
aspects (see Rufer 1954, esp. vi–​viii). The motives included within the Grundgestalt are then de-
veloped separately. Thus, the surface of the music is dominated by fragments of the series, rather
than complete melodic statements of it, yet “the ‘Grundgestalt’ is coherent because of . . . permanent
reference to the basic set” (Schoenberg [1949] 1975, 91). Babbitt expressed scant admiration for
Schoenberg’s writings (see Babbitt [1950b] 2003), but he was well aware of them.
48 The notation for time points here uses traditional analytic notation, with the time-​point classes

corresponding to numbers 0–​e and the opening downbeat written as time point 0. This analysis is
equivalent to Babbitt’s sketch notation, shown in Figure 1.6, provided F♯ = 0. See Leong and McNutt
2005 and 2019 for more on None but the Lonely Flute.
49 A “block” is a section of an array in which each line completes a series form. Given the require-

ment to present all partitions, block boundaries in all-​partition arrays are often blurred: several lines
will have already started their next series form before others have completed theirs. Figure 1.4a,
which presents the first block of the array for My Ends Are My Beginnings, is a further example.
Figure 1.8. Cross-​references in None but the Lonely Flute, mm. 1–​5

Figure 1.9. Blocks V and VI of the array for None but the Lonely Flute.
Components referred to in mm. 1–​3 are shaded. Parentheses indicate pitch
classes inserted between notes of the series; brackets indicate serially expected
notes that are absent.
32 Thinking In and About Music

components from some line in the array. This results in extended series
segments being heard, beyond the individual cross-​references: mm. 1–​3
present, in order, the complete first hexachord of series form RI5. Time
points that are unarticulated by any attack, such as those in m. 3, corre-
spond to blank components in the array—​aggregates in which the line
being referenced does not participate. Measures 4–​5, at the bottom of
Figure 1.8, reference two sequences from the first block of the array. Their
referents can be seen in Figure 1.5.50
The cross-​reference technique exemplified in None but the Lonely Flute
is used in most of Babbitt’s late pieces. The only significant development in
Babbitt’s reference technique that would follow is that in many of his pieces
from 1993 on, individual instruments or registers consistently refer to spe-
cific instruments or registers for the duration of a section.51 Over the course
of these pieces, the patterns of reference change, typically in an exhaustive list
of combinations. In a number of pieces, such patterns of reference form the
primary exhaustive list that spans the piece.
Babbitt mentions this technique (if elliptically) in an intriguing quotation
that alludes to both the technique and its aesthetic motivation:

In my latest works, particularly those beginning with my Sixth String

Quartet, the “macrophrases”—​or even “-​phases”—​are most immediately
characterized by changes in the degree of interdependency among the
pitch-​class and temporal materials of the instrumental parts, with their
constantly changing range of reference both proleptically and analeptically,
by intimation and recollection. The total organicism, after which I always
have strived, does not immediately manifest itself in such clearly reflected
dimensions whose structures, though ultimately genidentical, are more
often, or almost always, “polyphonic,” rather than “homophonically” coor-
dinated. (Babbitt 1998a, 28–​29)52

50 Considered in isolation, individual cross-​references can be considered ambiguous. For instance,

there are three aggregates in None but the Lonely Flute with a two-​note component ending in C, all of
which might correspond to the first cross-​reference. But only the referents shown in Figure 1.8 allow
the references to index a consecutive sequence of components. In Bernstein 2017, I overemphasized
the potential ambiguity. I am grateful to Daniel Colson for discussion on these matters.
51 Pieces that use this technique include String Quartet No. 6 (1993), Triad (1994), Tutte le Corde

(1994), Piano Quartet (1995), Manifold Music (1995), Clarinet Quintet (1996), When Shall We Three
Meet Again? (1996), Danci (1996), Piano Concerto No. 2 (1998), The Old Order Changeth (1998), and
Concerto Piccolino (1999).
52 Babbitt goes on to mention that his reference technique is “inspired . . . by that source of plenty,

the Schoenberg Fourth String Quartet,” and explicitly mentions the proleptic registral disposition of
that piece’s first aggregate.
 On Milton Babbitt, Schenkerian 33

The first sentence discusses the differentiation between sections—​

“macrophrases”—​ by changes of reference: in different sections, the
instruments (or registers) refer to each other in different patterns (change
their “degree of interdependency”). These references create the analepsis
and prolepsis Babbitt valued. The second sentence of this quotation is more
refractory, but it appears that the “genidentical”53 structures in “clearly re-
flected dimensions” are the all-​partition arrays sounding separately in each
instrumental part or register in these pieces.54 These are indeed “coordi-
nated” polyphonically, in the sense that they usually unfold disjointly, rather
than with the same harmonic rhythm.55
But from the perspective of this discussion, the most striking aspect of
this quotation is Babbitt’s assertion that he is seeking “total organicism.” The
polyphonic disposition of the arrays means this organicism is somewhat
obscured where they are concerned: it “does not immediately manifest it-
self ” in the relationship between the arrays. More “immediately” available,
as the first sentence claims, is the “interdependency” secured by the cross-​
references. And thus, especially in complex late works like String Quartet
No. 6, do the principle of exhaustive completion and the principle of sur-
face reference to the array complement one another. The series develops
outward hierarchically in nested exhaustive lists such as all-​partition arrays.
The surface, in turn, is arranged to reflect this hierarchy. The result is “total
organicism.”

Conclusion

It is an odd sort of compliment to congratulate someone on winning a game of

his own devising. But Babbitt’s compositional re-​imagination of Schenkerian
organicism clearly is, from some perspectives, immensely powerful. If one
wants music to be “as much as it can be”—​in the terms he defines—​the com-
plementary mechanisms of hierarchical expansion and surface reflection re-
alize this ideal to a startling degree. But the range of possible meanings one
might attribute to music on the basis of Babbitt’s writings, the qualities that

53 Carnap (1967) 2003, 199 explains, “we call genidentical . . . two states of the same thing” (em-

phasis original).
54 Following Lake 1986, such combinations of multiple simultaneous arrays are referred to as

“superarrays”; see also Mead 1994, 204–​63.

55 See Mead 2009, 230–​32 for an overview of the superarray of String Quartet No. 6.
34 Thinking In and About Music

contribute to his sense of musical “richness”—​as astonishingly multivalent as

they are—​are still limited. It is a central contention of this study that Babbitt’s
music is not so limited, and that exploring the full range of ways in which
his music is “as much as it can be” will require looking beyond those that he
drew from organicism. This is especially true because the expectations one
develops as a result of the organicist basis of his compositional systems—​
that his music presents neat hierarchies of exhaustive lists that can be shown
to be derived from a single initiating twelve-​tone series, for instance—​are
often not met. Later chapters will explore these anti-​organicist tendencies in
Babbitt’s music.
By adopting an analytical stance that looks beyond Babbitt’s writings
and the compositional techniques described therein, I accrue a debt to a
generation of excellent scholarship. Many analysts of Babbitt’s music in the
past thirty years have distinguished their analytical approach from that
suggested by Babbitt’s writings in one way or another. Joseph Dubiel (1990b,
1991, 1992, 1997, 2008, 2012, forthcoming) has been a sustained and
trenchant expositor of the position that Babbitt’s compositional techniques
offer little, and even confounding, guidance to the listener, and that analysts
would do well to focus on other matters instead. As he observes, Babbitt’s
arrays “greatly underdetermine what finally happens in the music, and the
presence of such structures in the finished work can be shadowy—​even
more, can be inconsistent: occasionally unmistakably vivid, at other times,
as good as totally absent” (Dubiel 2008, 124). This is unquestionably true,
and Dubiel’s critique of analytical accounts that mistake the array for the
finished music has been influential on the composition and organization
of this volume. Analyses in Chapters 6 and 7, in different ways, address
Babbitt’s music using concepts that have little or nothing to do with array
construction. Mead (2004, esp. 273n.6) and Emily Adamowicz (2011) also
directly distance their analytic approach from Babbitt’s writings, and sev-
eral other analysts simply do not engage his writings or compositional
techniques, properly concluding that there is much else worth attending to
(see Bailey 2002; Howland 2010, 39–​95; Rahn 1976; and perhaps especially
Greitzer 2007, 105–​59).
But it is a second basic contention that the organicist potential of Babbitt’s
music should still be taken seriously. It is useful for an analyst to consider
Babbitt’s organicism, if not as a governing principle, then as an inter-
pretive guide. It will, at times, function as an ideal against which Babbitt’s
 On Milton Babbitt, Schenkerian 35

actual music might be fruitfully compared.56 At other times, certain facets

of Babbitt’s organicism seem not to be relevant, but others still might be.
Babbitt’s music demands a flexible listener, one who takes insights from or-
ganicist principles when it is analytically fruitful and puts them aside when it
is not. It is music passionately resistant to doctrine.

56 Parkhurst 2017, 76 suggests a similar strategy for reading Schenker’s absolutism.

 2
Construction, Cognition, and
the Role of the Surface

A Way into Emblems

If “the limits of music reside ultimately in the perceptual capacities of the

human receptor” (Babbitt [1961] 2003, 84), then the first two measures of
Emblems (Ars Emblematica) (1989), shown in Figure 2.1, must have at least
one toe over the line. A purely numerical count of the measures’ contents fails
to capture their complexity but shocks nonetheless: twenty-​five attacks, five
sequential dynamic levels, six distinct rhythmic divisions, fourteen pitches
spanning seven octaves. Nor are the measures that follow much simpler.
A series of fragmentary gestures separated by pauses are anchored by a set of
fixed pitches—​notably the high G, left ringing across mm. 1–​5, and the low B,
snuck into m. 1 but reasserted   𝆑 and doubled in m. 7. A “human receptor” (a
listener, shall we say) seeking to account for these events, one who attempts
to understand the events as purposefully ordered and combined into music,
surely must struggle, at least on first encounter.
But further listening begins to pull the overflow of auditory stimuli back
into the realm of music. The first chord has three pitches in it. The burst of
septuplet thirty-​second notes that follows begins with the same ordering,
under retrograde inversion. Trichords evidently provide a good starting
point, and thankfully they are short, audibly differentiated, and easy to grasp.
The listener’s mindset adjusts to focus on them. Figure 2.1 is annotated to de-
pict some trichords that might be captured by this adjustment. Repetitions—​
(015)s and (037)s in mm. 1–​2; (025)s in mm. 3–​4—​mix with novelties, such
as the (027) that jumps out of the middle of m. 1. Repeated trichords become
fixed in memory, solidified in importance; and eventually, the novelties be-
come incorporated. By the end of m. 9, ten trichord orderings, representing
six set classes, have begun to define the piece’s harmonic language.
The path to understanding these trichords as components of longer
sequences is difficult to describe completely, but even on the basis of this

Thinking In and About Music. Zachary Bernstein, Oxford University Press. © Oxford University Press 2021.
DOI: 10.1093/​oso/​9780190949235.003.0002
 Construction, Cognition, and the Role of the Surface 37

Figure 2.1. Emblems (Ars Emblematica), mm. 1–​9. Trichord orderings are
annotated in reference to the series shown in Figure 2.2.

opening passage, glimpses of a larger organization emerge. Continuity in

some dimension—​simultaneities, successions of even note values, strings
of notes under slurs—​inspire cognitive grouping; this in turn suggests that
grouped trichords are presented in purposive sequence. In the slurred sep-
tuplet of m. 1, the first trichord is a repetition of the ordering introduced in
the opening chord, but the (037) into which it leads is new. The (037)’s link
with the (015) that precedes it, via the slur and the shared rhythmic value,
suggests that this ordering of (037) is not only after the ordering of (015), but
that it meaningfully follows (015). And so, a listener may hypothesize <D,
38 Thinking In and About Music

Figure 2.2. Emblems (Ars Emblematica), mm. 82–​83

C♯, F♯, A> as a four-​note grouping. By tracking such successions as the piece
unfolds, a sense of larger patterns gradually emerges.
Hindsight helps, of course. Figure 2.2 presents the first moment in which
all of the trichord orderings are presented in clear, direct sequence, placed
within the compact form that enables their most efficient construal: as the
segmental trichords of a twelve-​tone series. The appearance of the series as a
successive barrage of steady dynamic and rhythmic values, framed by rests,
distinguishes it from the surrounding context and fixes it in memory as a
unified Gestalt. Although they pass quickly, by the time a listener has reached
m. 82, the ten trichord orderings have each been heard many dozens of times.
The series of mm. 82–​83 can thus be heard as an encapsulation of the mate-
rial developed in the preceding measures, the “unique accumulator” (Babbitt
[1976b] 2003, 351) of the piece’s varied harmonic contents. A second pass
through mm. 1–​9, or a review of them in memory, grants them a different
character than they had at first: the harmonic variety remains, but the dis-
tinct trichords are no longer wholly independent from one another. They are
contained and conditioned by the series from which they may be understood
as developing or toward which they may be taken to lead.
There remain many aspects of mm. 1–​9 yet unaccounted for: the regis-
tration of the pitches, the rhythm, the dynamics, and much else. But this ex-
ample demonstrates that underlying his music’s daunting surface, Babbitt’s
compositional techniques support its comprehension. Musical cognition
and perception were items of serious concern for Babbitt; he returned to
them again and again in his writings and interviews, particularly after 1960,
 Construction, Cognition, and the Role of the Surface 39

and thought deeply about them in relation to his own composition. Babbitt’s
compositional structures, in all their complexity, were not forged out of de-
sire for score-​based, out-​of-​time, intellectual apprehension. Not for him,
Augenmusik.1 His procedures were meant to be heard—​to be responsive to,
and even to facilitate, the listening experience. “Musical structure necessarily
eventuates in the mind, the memory of the beholder” (Babbitt [1984b] 2003,
386), as he writes. Structures not grasped by a listener’s mind and memory,
however interesting, are not relevant to, not “useful” for, musical composi-
tion (Babbitt 1987, 118–​19).
Yet despite his concern, Babbitt’s writings on perception and cognition are
elliptical at best. As Dubiel (2012, 9) correctly observes, “The one thing Milton
never was articulate about was how he hoped for his music to be heard.” But
he did leave hints. In his analyses, citations of philosophers and cognitive
scientists, statements about his compositional procedures, and comments
in interviews, a reader may begin to stitch together aspects of Babbitt’s psy-
chology of music. Most important is the evidence of his compositions. The
picture that emerges of a “Babbittian” approach to listening is sketchy and
incomplete, but it provides the faint outline of a guide for the perplexed: a
blueprint for the comprehension of his music in his terms.2

Rudolf Carnap and Phenomenalistic Construction

As mentioned in Chapter 1, Rudolf Carnap was, along with Schenker and

Schoenberg, a vertex of Babbitt’s “Vienna Triangle” of influences. Carnap, to
be clear, was a philosopher, not a cognitive scientist; his subject was the def-
inition, determination, and clarification of knowledge, not its acquisition.
Nonetheless, his influence on Babbitt—​and the influence of the broader phil-
osophical tradition he represented—​was extensive and included aspects re-
lated to cognition. Despite not engaging mental processes directly, Carnap’s
work is relevant to Babbitt’s understanding of cognition in a foundational

1 Dubiel 2012, 9 and Shields 2012, 351 make the same point.
2 Babbitt’s personal library contained hundreds of volumes on all the topics discussed in this
chapter (philosophy of science, epistemology, memory, cognitive psychology, information theory,
and perception), and hundreds more on related topics (e.g., logic, linguistics, philosophy of mind,
philosophy of language, aesthetics, philosophy of art, and formal literary criticism). He did not an-
notate his books, so in general I hesitate to draw conclusions from the presence of a book in his li-
brary. But his knowledge of contemporary research on these topics, particularly in English but also in
German, was evidently vast and went far beyond the relatively few sources he cites in print.
40 Thinking In and About Music

sense, in that Carnap seeks to clarify how immediate perceptions are or-
ganized into knowledge. Moreover, as cognitive science took on a more
prominent role in Babbitt’s writings after 1960, Babbitt’s Carnapian views
on constructional systems, concept formation, and language were not
discarded. Cognition was simply folded into a pre-​established philosoph-
ical outlook. Before this cognitive turn, Babbitt uses a Carnapian framework
to understand how musical systems and pieces could be conceptualized.
Afterward, his prose is augmented by remarks on how they could be men-
tally constructed in real time by a listening observer.
Rudolf Carnap (1891–​ 1970) was a German-​ born philosopher. Like
Schenker and Schoenberg, he spent formative years in Vienna, and like
Schoenberg and many of Schenker’s followers, he would eventually be
forced to immigrate to the United States. In Vienna, Carnap and a rotating
cast of philosophers, scientists, and other intellectuals—​notably Moritz
Schlick, who founded the group known as the Vienna Circle—​met in the
1920s and early 1930s in an attempt to reimagine the role of philosophical
inquiry in light of contemporary scientific developments. Their primary
aim was to place philosophy on an empirical footing using modern logic,
such as that developed by Gottlob Frege, Bertrand Russell, and Alfred North
Whitehead—​hence the moniker “logical empiricist” often attached to the
group. The meaning of concepts would be supported by sensory observa-
tion and logically entailed deductions. This resulted in a radical critique of
metaphysics and of metaphysical assumptions in philosophical and scien-
tific language. Such concepts were taken to be unsupported empirically and
thus meaningless, according to the strict definition of meaning the Vienna
Circle propounded. But in nearly every detail, the Vienna Circle would revise
its assumptions frequently and dramatically during its short existence, and
Carnap himself would repeatedly change his views.3
Babbitt encountered contemporary philosophy at a young age.4 He took
classes in symbolic logic and the philosophy of science at the University of
Pennsylvania, where he enrolled in 1931 with an intention to major in phi-
losophy (Babbitt 1995, 59); but as with his musical interests, the key event
in his early life would be his arrival in New York two and a half years later
(after a “brief, unfulfilling stint at the University of North Carolina” [Harker

3 Uebel 2019 provides an excellent and accessible English-​language overview of the Vienna Circle.

See also Ayer 1959.

4 On this period of Babbitt’s life, see particularly Babbitt (1999) 2003; Brody 1993; and Harker

2008, esp. 347–​50.

 Construction, Cognition, and the Role of the Surface 41

2008, 348]). At Washington Square College he encountered the lively philo-

sophical circle surrounding James Burnham and Philip Wheelwright, who
engaged a set of issues related to that of the Vienna Circle (Babbitt 1979b, 43).
Babbitt’s personal contact with members of the Vienna Circle itself would
not come until somewhat later. Babbitt met Carnap during the latter’s stay
at Princeton from 1952 to 1954, although they seem not to have had a close
relationship (see Babbitt [1999] 2003, 480–​81). The member of the Vienna
Circle with whom Babbitt likely had the most interaction was Carl Hempel, a
colleague of Babbitt’s at Princeton from 1955 on.5
Babbitt’s prose is steeped in the argumentation and phraseology of the
Vienna Circle, but it can be difficult to trace any specific position of his to
any particular philosophical predecessor. A strong candidate, however, for
a foundational influence on Babbitt’s outlook is Carnap’s 1928 The Logical
Structure of the World ([1967] 2003), commonly known as the Aufbau, re-
flecting its original German title (Der logische Aufbau der Welt). Babbitt did
not cite the Aufbau, but its prominence and publication date virtually guar-
antee that it would have loomed large to a German-​reading student of phil-
osophical logic in the early 1930s.6 Carnap’s views would develop over the
next four decades—​and receive significant challenges from many, including
Willard Van Orman Quine, Nelson Goodman, and other philosophers
Babbitt admired and cited7—​but the Aufbau outlines positions Babbitt
would maintain without substantive revision throughout his career.
In the Aufbau, Carnap sketches a phenomenalistic constructional system.
That is, he outlines a formal logical process for building up justified complex
statements from observations. His purpose is to demonstrate how know-
ledge can be founded on empirically observable and verifiable particulars.
The starting point is “autopsychological,” the subjective experiences of an
observer—​“my experiences” (Carnap 1967, 13). That which is “given” by the
external world to each individual’s sensory apparatus forms the standpoint

5 On a personal level, this contact may have been more influential for Babbitt than his acquaint-

ance with Carnap, even though by the time Hempel had moved to Princeton, many of Babbitt’s phil-
osophical views were already formed. Notably, the model of explanation sketched in Babbitt (1972)
2003 is indebted to Hempel 1965. The title of Babbitt 1981 is a nod to Hempel 1958.
6 The 1967 translation of the Aufbau by Rolf A. George was in Babbitt’s personal library, although

the original 1928 publication was not; but Babbitt seems to have kept few books from before World
War II. The many German-​language books in Babbitt’s library indicate the original language of publi-
cation would not have proved a barrier. (He also read Schenker in the original, as discussed in Babbitt
[1999] 2003, 478–​79.) Taruskin 2010, 155–​56 also connects Babbitt and the Aufbau.
7 See particularly Quine 1951 and Goodman 1977, both discussed in Babbitt (1972) 2003. In per-

sonal conversation with the author in 2007, Babbitt named Quine, Goodman, and Carnap, along
with Donald Davidson, as among his favorite philosophers.
42 Thinking In and About Music

from which the world is understood. In this way, the Aufbau presents a
phenomenalist system with a quasi-​solipsistic focus on the individual.8 An
observer’s sensory experience provides the basis for systematic construction,
rather than a physical reality separate from the observer; this distinguishes
the phenomenalistic perspective from “physicalism.” From the basis of ex-
periential data, logical analysis consists of a process of “rational reconstruc-
tion,” “an inferential procedure whose purpose it is to investigate whether or
not there is a certain logical dependency between certain constituents of the
experience” (310). An important part of this exercise is to prune discourse
of statements that do not have a determinate logical connection to the em-
pirical given: “Initially, such a formulation will not produce an increase in
knowledge, but only increased purity of knowledge” (308). Questions an-
swerable in relation to empirical observation and logical justification can, as
a result, be distinguished from meaningless “pseudoproblems,” such as those
of metaphysics.
Carnap founds his system on “relation description.” The various objects
within the system are defined in relation to one another (19–​20), rather than
in terms of any sort of inherent essence or set of properties. This approach
is reflected by the system’s sole logical primitive (the undefined basic con-
cept from which all other relations are defined): “recollection of similarity.”
From observed and recalled similarities, an observer can formally describe
objects by their relations to other objects and gradually make increasingly
complex statements, such as statements about classes of objects. Complex
statements can be constructed from simpler ones if they include the simpler
statements’ extensions (the objects they describe)—​as, for instance, the set
class (015) includes the set {C, E, F} because every instance of {C, E, F} is
included within (015). The process of constructing complex statements out
of simple statements, successively iterated as the system grows in range and
complexity, is termed “subsumption” (57).
Babbitt never presents a fully realized phenomenalist constructional
system, but his analytical and compositional work is, in general, concordant
with the approach sketched in the Aufbau.9 He shares Carnap’s methodolog-
ical preferences, Carnap’s desire to purify discourse of statements lacking

8 Gleason 2015 explores solipsism in the music theory of Babbitt and, particularly, his students.

Carnap calls his position “methodological solipsism” ([1967] 2003, 101ff.); as he describes, it is
distinct from proper solipsism because he assumes structural properties of reality will be held in
common by separate observers (107).
9 Some of Babbitt’s students would formalize phenomenalist systems, with Boretz (1969) 1995

being the most elaborate and sophisticated example. See Gleason 2013, esp. 104–​15.
 Construction, Cognition, and the Role of the Surface 43

empirical verification, and even some of Carnap’s language. Consider, for

instance, Babbitt’s preference for the term “rational reconstruction,” in
place of “analysis,” for the process of coming to understand a composition
or group of compositions (e.g., Babbitt [1960] 2003, 66; [1963] 2003, 142;
see also Gleason 2013, 123–​25; Andreatta 2012, 60). Just as Carnap’s rational
reconstructions are not attempts to discover the origin of objects but rather
the logical connections supporting the apprehension of those objects, for
Babbitt, rational reconstruction is a task for listeners seeking to understand
the experience of a piece of music. As he describes, “A rational reconstruc-
tion of a work or works, which is a theory of the work or works, is, thereby,
an explanation not, assuredly, of the ‘actual’ processes of construction, but
of how the work or works may be construed by a hearer, how the ‘given’ may
be ‘taken’ ” ([1972] 2003, 301). The listener comes to understand a work not
by recapitulating a composer’s creative process but by construing logical
connections available in the heard experience. Like Carnap’s system in the
Aufbau (and unlike some of Carnap’s later work), this is an avowedly phe-
nomenalist approach; it is the experiential “given” that is the basis of con-
strual, not the physical properties of sound, the written score, or any aspect
of cultural context.10 This approach is, again, not directly concerned with
cognition, in that it concerns a posteriori reflection rather than in-​time expe-
rience. But it still provides the beginnings of a model for how listeners make
sense of a work. A listener begins with direct observation and groups the
resulting data into ever broader units and classes using logical inference.
Carnap’s primitive relation in the Aufbau is recollection of similarity.
Babbitt never articulates a single basic relation for the apprehension of all
music—​nor would it have been in character to do so, given as he was to em-
phasize music’s “contextual” diversity—​but the observation of similarity be-
tween musical objects is at the heart of his comments on a wide range of pieces
and musical systems. A central aspect of his theory of the twelve-​tone system
is invariance. In theoretical writings (notably Babbitt [1960] 2003) and in
virtually all of his analytical remarks on twelve-​tone music, he draws atten-
tion to the patterns of fixed elements held in common by the operations of
the twelve-​tone system. His understanding of “contextual” compositions is
little different in this respect: the fixed items he traces are motives, intervals,
and sonorities, rather than the ordered interval sequences and aggregates of

10 In this respect, Babbitt treats music as what Carnap describes as a “psychological,” rather than as

a “cultural,” object.
44 Thinking In and About Music

twelve-​tone music, but recognition of similarities remains the foundation of

his analytical approach to these works. Even his tonal analyses tend to focus
on fixed motivic patterns. And as in the Aufbau, the constructional process
suggested by Babbitt’s writings is typically a hierarchical one, as uninter-
preted surface data are gradually subsumed by more and more complex and
inclusive classes and entities. Pitch classes may be subsumed into twelve-​tone
series. Series, then, are subsumed by aggregates, hexachordal areas, and on
upward, until, theoretically—​Babbitt’s published analyses never reach this
ultimate stage—​entire pieces may be construed as wholes.
Rational reconstruction is an analytic procedure, not a creative one.
Its interest to a composer might seem mysterious. But Babbitt’s composi-
tional technique can be readily understood as being designed to facilitate
rational reconstruction in its listeners. Indeed, to a degree his techniques
can be understood as facilitating precisely the phenomenalistic construc-
tional procedure of Carnap’s Aufbau. Similarities—​twelve-​tone invariants—​
are guaranteed by Babbitt’s methods of array construction and his surface
references. As Babbitt ([1960] 2003, 58–​66) point outs, the standard op-
erations of the twelve-​ tone system—​ transposition, retrogression, and
inversion—​leave interval patterns fixed. In a piece composed of series forms,
the varied forms can be related on that basis and can be further grouped into
classes on the basis of additional invariants, such as the invariant hexachords
of the series forms within hexachordal areas. Derived series create further
patterns of invariance. The saturation of the surface of Babbitt’s music with
cross-​references enables a logical “relation description,” in Carnap’s terms, of
the counterpoint—​a determinate significance for sonorities or successions in
the context of their referents. In a more extended sense, the nested exhaustive
lists at all levels of Babbitt’s music provide for a hierarchical “subsumption” of
more surface-​level events within broader categories. The common property
of exhaustive completion uniting those lists forms the basis for considering a
relation among them.
Nonetheless, Babbitt’s compositional reinvention of Carnapian construc-
tion differs from the Aufbau in an important respect. In applying Carnap’s
analytic vision to music, Babbitt introduces a temporal element foreign to
rational reconstruction. Rational reconstruction was conceived as a way of
organizing a given set of facts. But one conceptualizing a piece while listening
will necessarily only be given access to it as it unfolds in time. The arrow of
time in Babbitt’s writings thus serves an epistemological function: later
events have the power to clarify or cast ambiguity on earlier events. Often
 Construction, Cognition, and the Role of the Surface 45

in comments on his own works, Babbitt will sketch a narrative of “disam-

biguation,” with the identity of the underlying series—​frequently withheld
during the piece’s opening, as in Emblems—​being gradually revealed as the
logical nexus to which other elements relate.11 On the qualitative feel of such
an epistemological experience, Babbitt is silent, but one may infer the pre-
sumed cognitive process. As in the analysis with which this chapter begins
the listener should attend to the varied events of the piece, note resemblances
where possible, and attempt to relate events to each other using a logic that is
minimal in its assumptions—​a logic that is as efficient as possible—​yet max-
imally determinate in its scope. Context, in this mode of listening, grants
meaning to events, but the varied scope of context, continuously changing
as a piece unfolds in time, inevitably affects how events are interpreted.12
Babbitt’s compositional procedures provide relationships that attentive
listeners may use to form provisional rational reconstructions of a composi-
tion as time goes by. “The musical composition’s accruing progress in time”
is, he declares, his “eternal concern” (Babbitt [1976b] 2003, 346).13
Phenomenalistic construction thus provides a possible model for the com-
prehension of Babbitt’s music. As Babbitt is keen to emphasize, the demands
his music places on perception itself are “simple and demonstrably realistic”
([1961a] 2003, 82): no more than the recognition of pitches and, especially,
“intervals—​the fundamental building blocks of any music” (2001, 3). But
perception is only the first step. Once phenomena have been observed, there
remains the problem of constructing the phenomena, of mentally trans-
muting discrete events into an “all-​of-​a-​piece” of music. As he clarifies,

Perception is a matter of conceptualization. I still encounter people who

say, “But can you hear this?” Well obviously that’s not the right question.
Obviously you can hear these things—​you hear the notes. The question is
how you conceive them, how you conceptualize them, how you relate them.

11 See Straus 1986, 22; also Babbitt 1998b and the program note for Septet but Equal (1992) (Milton

Babbitt Collection, Box 15, Folder 2).

12 Lewin (1986) 2006 explores this issue in great detail.
13 Although the composer himself did not make this connection, the epistemological experience

Babbitt describes bears comparison with Schoenberg’s notion of the “musical idea.” The musical idea
is a narrative whereby the relation between events is hinted at, gradually muddled, and then, at a piv-
otal moment late in the piece, clarified, such that seemingly disparate events are shown to be logically
connected (see Schoenberg 1995). As Jack Boss (2014, esp. 33) points out, serial clarity is one vari-
able Schoenberg manipulates to realize this narrative in his twelve-​tone music. Similarly, in Babbitt’s
narratives of disambiguation, the revelation of the series is the moment at which the various elements
of the piece are shown to be relatable as subsets of a single, underlying entity.
46 Thinking In and About Music

It’s not a matter of hearing in any physiological or discriminatory sense.

(Babbitt 1987, 44)

The task of mentally constructing a piece of Babbitt’s music on his terms

requires no more than an ordinary perceptual apparatus, a good memory,
and a willingness to seek out the logical connections relating events. It is, in
the end, a simple matter. But this is not to say that it is easy. His desire to make
music as “much as it can be,” to saturate his music with cross-​references, rich
counterpoint, and nested hierarchies, with an overwhelming deluge of rela-
tional plenitude, puts a heavy burden on listeners trying to make sense of the
music as they listen.
In the face of such a task, listeners may justifiably worry if they are hearing
a piece correctly. Babbitt’s notorious comparisons of music composition and
scientific research (e.g., Babbitt [1958] 2003) and musical and scientific dis-
course (e.g., Babbitt [1961a] 2003) could be interpreted to suggest that he
desired auditory construal of his compositions to also be scientific in nature.
“Science” is a vexed term, as will be discussed further, but in a naïve inter-
pretation of it, Babbitt’s invocations of science might suggest that a listener
should strive for objective truth—​for the universally valid interpretation of
a work at which all competent listeners will arrive. But Babbitt’s statements
imply a view closer to Carnap’s “autopsychological” standpoint. While
there are objectively verifiable facts about compositions—​the first chord of
Emblems includes an F—​the task of constructing those facts into music is a
matter of individual construal. As he repeatedly stated, the proper question
is not whether a piece is twelve-​tone, in some immanent sense, but whether
a listener finds a twelve-​tone construal most fruitful (see Tommasini 1996).
In Babbitt’s music, the “richness of relation” listeners face surely elicits from
each individual an experience highly conditioned by their own taste and
predispositions. The listening experience is, of course, always dependent on
the listener, and this must be especially the case as listeners struggle to wring
sense out of a stimulus just this side of unmanageable.
For indeed, Babbitt’s music seems to approach the edge of human powers
for comprehension. This would, in time, become something he felt the need
to investigate. As he frequently recounts, the advent of electronic synthesis
forced the issue: where a cap on musical complexity had previously been
imposed by the physical limitations of live performers, technology meant that
music was subject only to those “far less well-​understood limits: the percep-
tual and conceptual capacities of the human auditor” (Babbitt [1962] 2003,
 Construction, Cognition, and the Role of the Surface 47

109). And so Babbitt set out to determine what those limits are. Because in
order to inform a viable compositional method, phenomenalistic construc-
tion must account for the realities of human listening. Simply put, musical
phenomena need to be perceived if they are to form the basis of construction.
Logical relations cannot only exist on paper: events need to be audibly relat-
able by human listeners. But what can a human listener perceive, recall, and
mentally organize? The emerging discipline of cognitive science provided
Babbitt with guidance.

George A. Miller, Information Processing, and Memory

The “perceptual and conceptual capacities of the human auditor” are not only
incompletely understood—​although half a century of research since Babbitt’s
first statements on the subject has certainly improved matters—​they are also
limited along a number of intersecting axes. There are straightforward per-
ceptual limitations: sounds too high or low to hear, too short to be heard as
an instance of a pitch, or too fast for every pitch to be distinguished. Timbre,
newly available as a dimension of virtually unrestrained compositional cre-
ativity, was quickly understood to affect perception of pitch and duration.
As Babbitt (1964c) describes, an early task for the electronic composer—​at
least one with his compositional predilections—​was to assess the conditions
under which pitch and duration could be accurately perceived. Babbitt him-
self engineered experiments to “test the extension into the electronic do-
main of those invariants which have provided the basis for the formational
and transformational principles of past and present musical systems” (175).
Such tests can establish the immediate sensory impact of electronically pro-
duced sonic events. But to hear those events as music requires more than
the immediate discrimination of those sounds. For music to “eventuate in
the mind, the memory of the beholder,” the sounds heard must be retained.
And so, memory becomes a crucial issue for Babbitt, and discussions of how
music, including his own, facilitates memory would be a prominent part of
his writings thenceforth.
A touchstone source for Babbitt, as for many others, was George A. Miller’s
classic 1956 essay, “The Magical Number Seven, Plus or Minus Two: Some
Limits on Our Capacity for Processing Information.” As the title of Miller’s
paper suggests, there are cognitive limits on absolute judgement—​that is, the
ability to differentiate between discrete values—​and immediate memory,
48 Thinking In and About Music

the ability to recall a sequence of stimuli. These combine to put a cap on the
amount of information that can be transmitted or received. But as Miller
(1956, 95) writes, there are several ways in which “we manage to break (or
at least stretch) this informational bottleneck.” The most significant of these
as far as Babbitt is concerned is hierarchization. Immediate memory is lim-
ited in its ability to retain sequences of undifferentiated bits of information,
but Miller documents that if those bits can be grouped into “chunks” that are
independently learned and retained, the limit can be overcome. To borrow
Miller’s example, it is difficult to remember fifteen disconnected phonemes,
but easy to remember five words of three phonemes each. The limit is on the
length of the sequence, not on the total amount of information contained
within each element of the sequence: “Since the memory span is a fixed
number of chunks, we can increase the number of bits of information that it
contains simply by building larger and larger chunks, each chunk containing
more information than before” (93).
The mental process of transforming bits of information into chunks is
called “recoding.” Miller lingers on verbal recoding, in which a verbal name
substitutes for a number of bits of information, thereby repackaging them
for easy recall. Musical examples of verbal recoding abound: “the chromatic
scale,” for instance, is a single chunk encompassing twelve notes; “the first
theme of Mozart’s G-​minor Symphony” encompasses many more. But per-
haps the most commonplace example of musical recoding and chunking
concerns repetition—​or “redundancy,” as Babbitt typically characterizes it,
borrowing the term from information theory.14 Repetition—​even under
variation, and even only the repetition of a single dimension, while other
dimensions vary—​can transmute a string of notes and rhythms into a
single, memorable unit.15 With recoding accomplished by the observation
of repetition, chunking can take place and repeated events can be recalled.
Cognition continues, then, at a higher level, relating and thus retaining ever
larger chunks. As Babbitt ([1972] 2003, 290–​91) has it, “The memorative ca-
pacity is itself dependent on recoding and unitizing, on inducing principles
of interevent dependency and regularity, and, as a result, structure breeds
structure.”

14 According to Shannon and Weaver (1949) 1963, 13, redundancy is an inverse measure of infor-

mation content. Roughly speaking, the more redundant a signal, the less its components are inde-
pendent from each other.
15 The opening of Schenker’s Harmony makes much the same point, although without reference to

cognitive science (1954, 4–​20).

 Construction, Cognition, and the Role of the Surface 49

“The Magical Number Seven” was written in the midst of the “cognitive
revolution” then sweeping American psychology and linguistics, a general
push toward the study of cognition and memory and away from behav-
iorism (see Mandler 2002 and Miller 2003). Babbitt cites several leading
participants in this revolution at various points, including Eugene Galanter,
Noam Chomsky, and Miller himself.16 Following this revolution, Babbitt
came to describe the promotion of cognitive chunking and hierarchiza-
tion as an essential desideratum of musical structure and the achievement
of chunking and hierarchization as the foundation of competent listening.
Cognitive science, in other words, has aesthetic consequences. It sets out
minimum requirements for a successful musical experience.
This view is put forth explicitly in Babbitt’s (1986b) interview with Ev
Grimes in Music Educators Journal. In response to a question about how
one teaches music, Babbitt replies, “The critical aspect of hearing music is
musical memory.” He then goes on, paraphrasing Miller, to describe how
memory can handle only “five to seven unrelatable chunks” of information.
Babbitt’s solution for this limitation is the same as Miller’s: “you can chunk
in larger and larger and larger elements, and as the chunking becomes larger
and larger, these five to seven may [become] a whole piece of music” (59).
Babbitt seems to feel this response answers Grimes’s question because he
recognizes that chunking is a teachable skill.17 In a later interview, he would
expand on the idea that a competent listener is one able to proficiently chunk
and that a successful piece of music is, at minimum, one able to be chunked:

For, although everyone possessed of physically normal hearing “hears” the

same thing and things, the listener declares his appropriateness by concep-
tualizing, “chunking,” acquiring cumulative knowledge at the first hearing
of a work as it proceeds, and at later hearings by differently applying the
knowledge of past performances. If the work is not constructible eventually

16 Galanter is cited in Babbitt (1972) 2003, 304. Chomsky is alluded to in Babbitt (1965) 2003, 199;

Babbitt’s comments in Westergaard 1968, 71; and Babbitt (1972) 2003, 292. He mentioned or alluded
to Miller 1956 a number of times.
17 As discussed subsequently, a similar statement by Babbitt on pedagogy and the development

of memorative capacity can be found in Westergaard 1968, 69–​70 and 72. Although Babbitt does
not expand on how a teacher instills competent chunking, a great deal of music education can in-
deed be described as training for recoding. When students learn common patterns such as scales and
arpeggios, they recode them for later retrieval. Heightened pitch and interval memory trainable by
aural skills instruction may help students recognize redundancies. Frequent, practical exposure to
music—​an activity Babbitt repeatedly lamented was vanishing in public schools (e.g, Babbitt [1991b]
2003, 453; [1994] 2003, 461)—​can similarly attune students to schemas that can be recoded and later
retrieved.
50 Thinking In and About Music

as a totality, whether the failure resides in the listener’s constructive memory

(the creator of musical structure) or in the work itself is the central question
of music analysis. (Babbitt 1998a, 24–​25)

In Babbitt’s statements on cognition, one may glimpse the responsibilities

of listeners and composers. Listeners must learn to differentiate and chunk
the stream of musical experience; otherwise, they quickly reach cognitive
limits on what they can absorb. How should they go about this? Babbitt does
not clarify, but given its significance to cognition in general, to his writings
about music, and especially to his own compositions, one critical factor must
be recognizing redundancies. Redundancies enable recoding, which enables
the successive hierarchical gathering of bits of data into chunks large enough
that whole pieces may be perceived and, eventually, recalled. In music like
Babbitt’s, in which literal repetition is virtually absent, this is no easy task—​it
may involve, for instance, noting that the first three notes of the septuplet in
m. 1 of Emblems replicate the intervallic pattern of the piece’s first sonority
projected in register despite the fact that the ic5 of the opening sonority is ex-
panded by an octave in the septuplet. Cross-​references, a frequent source of
redundancy in Babbitt’s later music, present even greater challenges.
In rationally reconstructing the listening experience, analysts are tasked
with demonstrating the cognitive comprehensibility, or lack thereof, of the
work under consideration. Many of Babbitt’s later analytical comments,
demonstrated above all in Babbitt (1987), can be read as concerned with
cognition. Consider his comment that the correspondence in Schoenberg’s
Fünf Orchesterstücke (1909) between a moment near the opening of the first
movement and the chord that begins the third movement is there to “make
your life simpler” (1987, 159). Babbitt presents this comment as a throwaway,
but as with many of Babbitt’s quips, this is no joke: a listener’s life is simpler
from a cognitive perspective because redundancy facilitates recoding.
Babbitt’s cognitive views task composers with responsibilities comple-
mentary to those assigned to listeners. They should write music receptible
to chunking as a result of redundancies or some other means. As will be
discussed in the next section, a number of specific techniques Babbitt
employs seem designed to facilitate cognitive hierarchization. But Babbitt’s
music, as he is willing to admit, is challenging—​and part of the challenge is
that despite all of his expressed concern about cognition, his music refuses to
promote any single or obvious groupings that enable ready chunking. Some
of the difficulty, as he acknowledges, is his interest in “contrapuntalization”
 Construction, Cognition, and the Role of the Surface 51

of the various musical dimensions, rather than the “conjoining” of them

that produces literal repetition (Babbitt 1987, 167). A listener must be able
to simultaneously track pitch repetitions unsupported by rhythmic repeti-
tion and rhythmic repetitions unsupported by pitch repetition. More of the
difficulty, surely—​although he does not acknowledge this—​is the music’s
insistent irregularity, which leads to ambiguity about where chunks begin
and end.18 A knowledgeable listener is likely to be just as flummoxed by an
ambiguous grouping as an inexperienced one; this is the nature of ambi-
guity. The challenge Babbitt’s music puts forth is thus in part the challenge
of maintaining an active, creative disposition while listening: to be willing to
experiment with different grouping patterns and not be put off by being oc-
casionally stumped.19
Consider again the trichordal analysis of Figure 2.1. The segmentation
was chosen to highlight trichords that correspond to series segments, but
there are many other ways a listener might segment this passage, some of
which could be even more conducive to redundant chunking than the anal-
ysis shown. (As ever in Babbitt’s music, linear array counterpoint, composed
entirely of equivalent series forms, is the most redundant grouping of all,
although the slow and irregular unfolding of those series forms, a particu-
larly refractory aspect of Emblems, likely presents additional memorative
challenges.) Trichordal segmentation is available in the music, but so is much
else, and a listener must be willing to hop around between different hearings
in order to find the most helpful segmentation at any given moment.
The cognitive turn in Babbitt’s writings after 1960 represents an impor-
tant rhetorical shift, but his interest in Carnapian construction appears not
to have been meaningfully revised. What cognitive science gives Babbitt is
a basis for believing that what is “given” by music can actually be “taken.”
Provided a listener is presented music they can perceive and recall, the task of
construction continues as before. Additionally, and importantly, his views on
cognitive science and Carnapian construction theory are mutually reinfor-
cing: redundancy and hierarchization play central roles in his takeaways
from both disciplines. In many of his statements, epistemology and cognitive
science bleed into each other. To wit:

18 See Lerdahl and Jackendoff 1983, 13–​ 17 for discussion of how rhythmic features facilitate
grouping.
19 A related point is made in Maggart 2017b, 74.
52 Thinking In and About Music

In that familiar epistemological experience in which knowledge of a work is

acquired as the work proceeds in time, the offered and perceived reach and
richness of relation, depth and scope of reference, create that sense of suc-
cessive subsumption and cumulative containment which memory requires
if the work eventually is to be entified as a unified totality, an all-​of-​a-​piece
of music. (Babbitt [1983] 2003, 386)

Babbitt’s Cognitively Informed Compositional Techniques

The twin, reinforcing concerns of phenomenalistic construction and cog-

nitive chunking, in concert with his organicist views, can be seen to moti-
vate virtually all of Babbitt’s compositional techniques. (The realization of
those techniques in specific pieces of music is a more complicated subject,
as the remainder of this book will explore.) As already mentioned, the in-
tervallic invariance relating twelve-​tone series forms and their subsump-
tion in the nested hierarchies of exhaustive lists discussed in Chapter 1
correspond neatly to Carnap’s phenomenalistic construction. They also
match Babbitt’s views on cognition. Invariants are redundancies; they fa-
cilitate chunking. Hierarchy too enables chunking, and therefore the sub-
sumption of local perceptions into broader contexts. The level-​invariant
principle of exhaustive completion can be interpreted as an attempt to se-
cure an extended kind of cognitive redundancy among lists as diverse as
pitch-​class successions and sequences of ensembles. In statements about
his music, a cognitive rationale for his procedures comes up repeatedly.
Babbitt articulated a cognitive motivation perhaps most clearly for three
of his techniques: derived series, cross-​references, and interdimensional
parallelism.
Babbitt’s work with trichords long precedes his public engagement with
cognitive science and the publication of Miller (1956), dating as far back as
the unpublished Music for Four Instruments (1938).20 But whether he real-
ized the cognitive benefits of the technique early on or merely came to jus-
tify the technique cognitively after the fact, he saw that trichordal derivation
solves a cognitive bind within twelve-​tone music. Out of context, at least, it

20 The opening of this work is reproduced in Perspectives of New Music 14/​2–​15/​1, 206–​7 and is

discussed in Harker 2008, 350–​52.

 Construction, Cognition, and the Role of the Surface 53

is difficult to remember twelve disconnected pitch classes—​twelve exceeds

Miller’s magical limit—​but the internal repetitions of a trichordally derived
series render it more readily memorable:

The assumption of a single note transposed eleven times requires

the recall of the order of eleven individual, independent (unless one
assumes the structure of the twelve-​tone set as a priori) phenomena to
remember the set to within total transposition, while the trichord as-
sumption requires only the recall of the two intervals of the trichord
and the order of application of three operations and their combinations;
the latter suggests a psychologically more realistic recoding in terms
of the “magic number seven, plus or minus two.” (Babbitt [1972] 2003,
283–​84)

In this passage, Babbitt is discussing the series of Webern’s Concerto for

Nine Instruments, which is itself trichordally derived. In Babbitt’s published
music, trichordal derivation is never applied so simply; it always appears as an
elaboration of an underlying series that is less redundant (see Figures 1.1–1.3).
The function of trichordal derivation in such works might be understood, at
least in part, as presenting the underlying series in cognitively manageable
units and repeatedly reiterating those units until they have been thoroughly
recoded.
In pieces such as String Quartet No. 2 (1954) or Composition for Tenor
and Six Instruments (1960), which present a sequence of arrays generated
from trichords and other subsets that eventually leads to the series (see
Chapter 3), trichordal derivation may be interpreted as having an instructive
function. It gradually teaches the series to listeners, enabling them to recode
its separate trichords, and often hexachords—​introduced and reiterated
as combinations of trichords—​before the series is presented as a whole. In
pieces such as Composition for Four Instruments and Danci that never state
their series in full, derivation may be understood similarly—​with the caveat
that other elements of the music may also need to be invoked in order for the
series to be conclusively disambiguated (see Chapter 4). And in still other
trichordal pieces—​such as Glosses, “The Widow’s Lament in Springtime”
(1950), Two Sonnets (1955), and Partitions (1957)—​which state their series
at the outset before developing its separate trichords (see Chapters 1, 3, and
5), derivation may be understood to have a confirming function. The series
54 Thinking In and About Music

may not be fully grasped on first hearing, but the trichords that follow, in
isolating and reiterating its separate trichord orderings, enable it to cohere
in memory.21
As discussed in Chapter 1, the surface of almost all of Babbitt’s pieces—​
the direct successions and simultaneities—​are virtually saturated with cross-​
references to the array or to the underlying series. Often these references
are composed to facilitate grouping: as in None but the Lonely Flute, a fixed
rhythmic value unites the members of a reference. Although specific strate-
gies for organizing and realizing cross-​references vary, the principle of sur-
face reflection of the series or array is among the most consistent aspects of
Babbitt’s compositional technique. Babbitt referred to these cross-​references
frequently in comments on his own music and in discussion of analogous
techniques in other composers, particularly Schoenberg. Often, as quoted in
Chapter 1, references are mentioned in relation to temporal experience: they
are “analeptic” or “proleptic,” serving “intimation” or “recollection.” The fre-
quency of such statements, and the fact that these statements often make no
mention of cognitive efficiency, suggests an aesthetic, as well as a cognitive,
purpose for these references. As Babbitt (1987, 23) says about Schoenberg,
in a statement that, like many of Babbitt’s analytical comments, seems to be
simultaneously about his own music, “It’s the idea of constantly intimating
where you are going—​constantly predicting so that when you arrive in cer-
tain places in the piece, you have the sense of having gotten someplace which
has already been predicted.” This sense of arrival seems to be part of his mo-
tivation, whether arrival of an array segment predicted by a cross-​reference
or a cross-​reference prepared by an array segment. And as Chapter 1 details,
organicist aesthetics, too, play a role.
But cognitive efficiency is also served by cross-​references in a number
of respects that concerned Babbitt. One, naturally, is redundancy. Cross-​
references are formed of series segments, and the result of saturating the sur-
face with them is that listeners are quickly familiarized with the contents of
the series. Within minutes, almost any one of Babbitt’s compositions presents
dozens, sometimes hundreds, of references, all of which reflect segments of
the same underlying series class. More distantly, cross-​references in most

21 In pieces that deploy trichordal and all-​partition arrays simultaneously, such as Paraphrases

(1979), The Head of the Bed (1982), or Groupwise (1984), the extracted trichords may be taken to have
instructive and confirming functions at different points, depending on whether the underlying series
has been presented in the all-​partition array. Babbitt’s notes on Paraphrases (apparently prepared for
a talk at Amherst College, perhaps coincident with a 1980 performance of Paraphrases there) claim
the trichords in that work “restat[e]‌to clarify” (Milton Babbitt Collection, Box 12, Folder 2).
 Construction, Cognition, and the Role of the Surface 55

pieces from the second half of Babbitt’s career, such as None but the Lonely
Flute, are not simply series segments; they specifically index components of
array partitions. These can be difficult to hear, as will be taken up in greater
depth shortly, but Babbitt expects these redundancies, too, to support appre-
hension and memory.22 Events that have been predicted or recollected are,
by virtue of redundancy, recoded. Cross-​references have another benefit: the
vast majority are brief, fewer than seven notes long. Thus, as with trichords
in pieces using trichordal arrays, the surface of the music is filled with cogni-
tively manageable series segments. Since the length of these segments slides
below the limit on immediate memory discussed by Miller (1956), the full
series can be mentally constructed out of these segments, which are recoded
by virtue of continual repetition.
Among the more interesting of Babbitt’s statements on the role of the sur-
face is a brief aside about the Mozart G-​minor Symphony, which appears
within a broader discussion of his own work. (He admits the observation
concerns his own compositions.) Babbitt notes that the first complete triad
in the first movement’s second theme, in m. 46, is G minor, a reference to the
work’s tonic; he also identifies a number of additional linear instrumental
connections at that spot that index G minor, all despite the prevailing local
key of B♭ major. Such a cross-​reference indicates “the means of keeping the
past alive in the present, of referring back to, of understanding the other-
wise incomprehensible (at least in the sense of not requiring special com-
prehension, or of being understood as a primitive itself only to be referred
to)” (Babbitt [1976b] 2003, 343). Absent the reference, Babbitt suggests—​say,
if m. 46 began the symphony or was transplanted into a piece beginning in
another key—​listeners would be required to add that triad to the list of basic
compositional materials they must keep in mind. But since Mozart repeats a
triad already well established within the piece, listeners can readily compre-
hend it using tools already at their disposal. In his own music, by matching
surface simultaneities and successions to other events in the piece, Babbitt
seeks to establish a comparable benefit.
A third technique Babbitt justified cognitively is his establishment of
a formal parallelism between pitch and rhythmic technique, most notably
his use of the time-​point system as an analogue of the twelve-​tone system.

22 As he writes in the program note for String Quartet No. 5 (1982): “The composition assumes

‘form’ and ‘shape’ by the continual, progressive, expanding interplay of recollective and predic-
tive self-​references, which should serve comparably the musical memory of the attentive listener”
(quoted in Peyser 1995, 170–​71).
56 Thinking In and About Music

As with trichordal derivation and surface references, Babbitt’s interest in

interdimensional parallelism precedes his overt concern for cognitive sci-
ence, and his constructional and organicist interests seem to have led him
to similar conclusions. Recall the statement quoted in Chapter 1: the time-​
point system provides both “holism” and the “reinforcing redundancy
of interdimensional parallelism” (Babbitt 1998a, 32). As a result of the
parallelism—​the redundancy between the dimensions of pitch and rhythm—​
the structures of each domain might, he claims, be easier for a listener to pro-
cess (see also Babbitt [1962] 2003, 138).
Babbitt is widely known as an “integral” or “total” serialist, one who ex-
tended the serial principles of the twelve-​tone system to all other musical
dimensions—​rhythm, dynamics, timbre, and so forth (e.g., Ball 2011). This
mistaken assumption likely arose as an erroneous conflation of Babbitt
and his European contemporaries, such as Pierre Boulez and Karlheinz
Stockhausen. He does not treat any dimension serially besides pitch and
rhythm. Nonetheless, this is not for the lack of a totalizing impulse. As
Babbitt ([1955] 2003, 40) explains, he was interested in composing “works
in which all components, in all dimensions, would be determined by the
relations and operations of the system,” and such a pursuit seems a natural
outgrowth of his organicist aesthetics, interest in constructional systems,
and desire for cognitive redundancy. But as he was well aware, the various
musical dimensions are not perceived identically, and this affects the kind
of parallelisms that can be meaningfully deployed compositionally (Babbitt
[1972] 2003, 284–​85; 1998a, 33). In order to clarify the differences in how
the various dimensions are perceived, Babbitt refers to the standard clas-
sification of scales of measurement.23 Only pitch-​class structure, with the
“zero” interval offered by octave equivalence, is naturally susceptible to ratio
scaling, the most specific measurement. Durations, in contrast, are only sus-
ceptible to intervallic scaling. Perhaps for this reason, in the years before

23 The editors of Babbitt 2003 provide a helpful definition of the four types:
In ascending order of increasing specificity: in a nominal scale of measurement elements
are merely assigned to categories, with no order or rank attributed to those categories
(e.g., nationality); in an ordinal scale of measurement elements are ordered with respect
to each other, though degree of difference is not defined (e.g., such two-​place predicates
as “louder than”); an interval scale not only orders the elements, but also defines a unit of
measurement descriptive of the difference between them, although the position of zero in
the scale is arbitrary (e.g., temperature Fahrenheit); a ratio scale has all the characteristics
of an interval scale except the zero point is absolute and meaningful (e.g., income meas-
ured in a defined and stable currency). (Babbitt 2003, 176–​77n.2)
Arnold and Hair 1976, 165 also discuss the significance of these concepts in Babbitt’s work.
 Construction, Cognition, and the Role of the Surface 57

Babbitt adopted the time-​point system, when he occasionally worked with

duration series, he typically makes no attempt to fully analogize rhythm and
pitch-​class structure, generally omitting aspects such as modular transpo-
sition that would be artificially imposed.24 It is only with his invention of
the time-​point system—​in which the establishment of a modulus defines a
meaningful zero interval, thus creating a robust formal analogue to pitch-​
class space—​that Babbitt began regularly using equivalent arrays for both
pitch and rhythm.25 Dynamics, register, and timbre are not even susceptible
to intervallic structuring: dynamics and register are ordinal, and timbre (as
Babbitt conceives it) is merely nominal. Since these dimensions cannot be
structured intervallically, Babbitt does not serialize them. Dynamics, reg-
ister, and timbre are linked to the serial elements of Babbitt’s music: they are
typically used for the projection of pitch or rhythmic serial counterpoint
and can also be arranged into patterns of exhaustive lists. But they are never
tasked with presenting intervallic transformations that could not be per-
ceived as such.

Organicism as Cognitive Heuristic: The “Schenker

Memorative Approach”

Cognitive science may seem a long way from Schenker and organicism. For
Babbitt, it is not. Babbitt frequently writes of Schenker as an inspiration for
his cognitive views. Particularly in his later years, following the cognitive
turn in his writings, he repeatedly discusses Schenkerian theory as a model
for musical memory:

For me, Schenker has always offered the most powerful hypotheses about
how our musical memory functions and how we can take in a piece as a
whole. . . . Schenker allowed us to view how a piece of music takes shape
on various temporal and structural levels in a cumulative way—​cumulative

24 Mead 1994, 38–​45 and 54–​123 presents a detailed survey of Babbitt’s early rhythmic practice.

Composition for Twelve Instruments (1948) is a problematic exception, as discussed in Westergaard

1965. See also Mead 2011, which discusses Three Compositions for Piano (1947), Babbitt’s first
published foray into rhythmic serialism, in the context of some of the cognitive issues raised here.
25 Babbitt ([1972] 2003, 285) claims that pitch is the dimension most susceptible to structuring.

Given that the time-​point modulus enables rhythm to be structured equivalently to pitch, he may be
referring to the fact that the time-​point modulus must be artificially imposed by a composer. It is not
a natural (or, at least, cultural) given like the octave modulus is for pitch-​class structure. It can change
from work to work or even within a single composition.
58 Thinking In and About Music

containment, or successive subsumption, if you will—​thus making it pos-

sible for musical memory to function and musical works to be perceived in
their entirety. (Babbitt 1991a, 129)26

This cognitive reading of Schenker—​which he once referred to as the

“Schenker memorative approach” (1987, 126)—​came to dominate Babbitt’s
rhetoric on Schenker, echoing other comments he made on musical structure
generally in the last several decades of his life. But his reading of Schenker
tracks his intellectual interests. In his early writings, as discussed in Chapter 1,
Babbitt lauds Schenkerian theory as a model for a musical axiomatic system
(Babbitt [1952] 2003, 23). His remarks then suggest he admired the similar-
ities between Schenker and phenomenalistic construction, such as that of
Carnap. Observing these similarities requires sidestepping substantial philo-
sophical differences; but nonetheless, both Schenkerian theory and Carnap’s
Aufbau propose a formal model for the organization of knowledge that
begins with a minimal basic assumption and develops hierarchically through
a limited set of fixed rules, taking care to observe recognition of similarity
(in music, motivic parallelism). Bringing new facts into Carnap’s system
increases “epistemic value” but preserves logical value; properly composing-​
out a Schenkerian level increases musical content but preserves contrapuntal
correctness. In Babbitt’s writing, language equating Schenkerian theory with
axiomatics fades quickly, replaced by comments on cognition and memory;
but as with his statements on other aspects of musical structure, it is probably
best not to view the gradual replacement of axiomatic rhetoric by cognitive
rhetoric as evidence of changing views, for he seems to have viewed well-​
constructed axiomatic systems as consonant with a viable cognitive model.
Logically supported hierarchy enables efficient cognition and memory.
And so Schenker provided Babbitt with a model of apprehension by way
of chunking and hierarchization. The detailed events of one level become
“contained” within another—​they are chunked, recoded, and thereby fixed in
memory, enabling them to be brought into contact with more distant events that
are similarly processed. And the organicist rhetoric Babbitt shared with Schenker
can be taken as a heuristic for this cognitive process. As discussed in Chapter 1,
such language functions as “prop oriented make-​believe,” an imaginative ex-
ercise that aids communication of a real, non-​imaginary thing. This rhetoric is
used to communicate the conditions of comprehensibility. Babbitt’s organicist

26 This sentiment is reiterated in a number of other sources, such as Babbitt 1987, 145.
 Construction, Cognition, and the Role of the Surface 59

comments on his own music should be understood in this light. Accordingly, to

approach his music with organicism in mind, as proposed in Chapter 1, may help
a listener apply the cognitive model Babbitt hoped to encourage.

“The Requirements”—​and Responsibilities—​“of

Cognitive Communication”

A few words about language are in order. Strictly speaking, this section is an
aside, since it does not concern the analysis of Babbitt’s music, but it touches
on a cognitive matter deeply important to Babbitt. Perhaps no subject appears
more frequently in Babbitt’s writings than his critique of loose, inaccurate,
and casually subjective musical discourse. His festering ire explodes into ba-
roque denunciations of journalists, historians, aestheticians, and experts in
other fields who traffic in the “never-​never land of semantic confusion, that
last resting place of all those verbal and formal fallacies, those hoary dualisms
that have been banished from rational discourse” (Babbitt [1958] 2003, 52;
see also Babbitt [1972] 2003). His anger arises from both philosophical con-
sideration and a pragmatic awareness that improper language can doom the
reception of music before a note has been heard (Babbitt [1965] 2003, 191–​
93; also Schuijer 2008, 257). In his intellectual heritage, concern for language
was overdetermined. One can hear the resonances of Carnap’s “elimina-
tion of metaphysics through the logical analysis of language” ([1932] 1996),
Schenker’s screeds against hermeneutics, and Schoenberg’s complaints about
journalists within Babbitt’s tirades.
There is much to be said about Babbitt’s views on language, but in the con-
text of this chapter, I will focus on but two aspects: what Babbitt meant by
“scientific language” and how he felt language impacted cognition. As Peles
(2012, 23) points out, the word “science,” in Babbitt’s ([1961a] 2003, 78) call
for “ ‘scientific’ language and ‘scientific’ method,” does not signify the dis-
cipline of practicing scientists. It is “science” construed broadly—​closer to
the German “Wissenschaft” (or the Latin “scientia”) than to lab coats and
Bunsen burners, in Peles’s apt formulation.27 For Babbitt, science is coexten-
sive with confirmed knowledge in any field, and therefore scientific language

27 Nor does Babbitt’s usage of “science” refer to the disciplines charged by the United States mili-

tary with winning the Cold War. Richard Taruskin (2009, 275) is mistaken to conflate the practice of
working scientists with the broader meaning Babbitt intended by the term.
60 Thinking In and About Music

is language on any subject whose assertions are founded in empirical obser-

vation and logical inference. It is a position he shares with Carnap and the
Vienna Circle.28 Recall the title of the Aufbau: Carnap seeks to provide an
epistemology fit to deal with “the world”—​that is, with all of knowledge, re-
gardless of discipline, explicitly including the study of culture. Adopting this
belief, Babbitt feels that precise language is as necessary a component of mu-
sical inquiry as it is in any other discipline.
But his worries about language run deeper. Having read Carmichael,
Hogan, and Walter (1932), Babbitt was convinced that language impacts cog-
nition and thus memory.29 His citation of this paper comes in a panel discus-
sion on music pedagogy:

If we’re discussing the capacity to hear things and the capacity to iden-
tify them verbally, we’re into the whole question of associated verbal
concepts . . . after all the verbal concept that you associate with what you
perceive is obviously going to have a great effect in the whole memorative
process. We’re back to the old Carmichael experiments again. If you show
two groups of people the same shape, tell one group of people that it’s
eyeglasses and the other that it’s dumbbells, when they’re asked to repro-
duce it, the eyeglasses people are going to reproduce it one way and the
dumbbells people are going to reproduce it another way. That’s why we
have to be terribly concerned about these attached verbal concepts. We’re
all talking about memory, about increasing memory span, and therefore
we’re back to that question of verbal responsibility; if you call this thing in
[Schoenberg’s Klavierstück, Op. 33a] a dominant seventh, you’re in trouble.
(quoted in Westergaard 1968, 69–​70)30

As Babbitt (1998a, 28) was fond of saying, there are no “immaculate

perceptions.” Perceptions are conditioned by the mindset of the observer.31
Verbal formulations affect that mindset, and therefore words mediate

28 The “boundary-​work” (in the sense of Gieryn 1983) Carnap and Babbitt each desired for their

disciplines is also similar. Just as Carnap (e.g., [1932] 1996) sought to exclude meaningless meta-
physical problems from philosophy, Babbitt dismissed musical discourse insufficiently grounded in
experience or reason (see especially Babbitt [1961a] 2003, [1965] 2003, and [1972] 2003).
29 Miller 1956, 95 also discusses Carmichael, Hogan, and Walter 1932, as evidence of the signifi-

cance of verbal recoding.

30 See Gleason 2013, 131–​35 for further discussion of Babbitt’s citation of Carmichael, Hogan, and

Walter 1932.
31 For insightful discussion of this subject, see Dubiel 2005–​6 and the works of Benjamin Boretz

discussed therein.
 Construction, Cognition, and the Role of the Surface 61

experience. If one encounters a sonority in a Schoenberg twelve-​tone piece

that happens to correspond to a dominant seventh chord, calling it a “dom-
inant seventh chord” will lead to one trying to hear it tonally. This will con-
found cognition: it will impede the apprehension of that sonority’s function
in the piece, providing misleading information about its relation to its
context. Tonal hearing is ineffective at coming to terms with Schoenberg’s
Op. 33a, so most likely, an attempt to hear the sonority as a “dominant sev-
enth chord” will result in it sounding unintegrated. Without a surrounding
tonal context, listeners will be forced to add the concept of the dominant sev-
enth chord to their list of basic materials in the piece, and the piece will be-
come harder to remember. It may even be dismissed as incomprehensible.
Babbitt is concerned about language because language affects cognition.
There are direct musical and pedagogical dangers that may result from im-
precise language. This is why musical discourse should conform to “the
requirements of cognitive communication” (Babbitt [1965] 2003, 192). It is
not, or at least not merely, for the sake of Cold War political posturing, as
some have averred (Brody 1993; Taruskin 2010, 155–​57). It is not rigor for
the sake of rigor. It is because musical discourse shapes musical cognition.
And so Babbitt, worried music he cared about (including his own) would not
be comprehended, turned his fire on those whom he felt impeded compre-
hension by equipping listeners with inadequate mindsets.

Conclusion: Babbitt’s Psychology and the Analysis

of His Music

It is probably most charitable to view Babbitt’s psychology of music as

partial, aspirational, and of its time—​the time being the years in which it
initially developed, in the 1940s, ’50s, and ’60s, during the early years of
cognitive science. It is partial in that it focuses exclusively on a narrow set
of concerns: memory, hierarchization, and association. It seems that music,
in Babbitt’s psychology, is a stream of information to be processed.32 There

32 Gamer 2012, 362 reports that Babbitt lectured on information theory, including a “detailed

explication” of Shannon and Weaver (1949) 1963, at the Princeton Seminars in Advanced Musical
Studies in 1959. While he does not cite it, Babbitt ([1972] 2003, 291) apparently alludes to Shannon
and Weaver’s ([1949] 1963, 13) finding that English is 50 percent redundant. See also Babbitt (1958)
2003, 49 for a well-​known, explicit use of information-​theoretic terminology (“efficient” and “re-
dundancy”) in relation to music. On musical applications of information theory contemporary with
Babbitt’s first encounters with it, see Bell 2019, 116–​89 and Iverson 2019, 105–​38.
62 Thinking In and About Music

is more to cognition—​and more to music, including Babbitt’s music—​than

this, as a great deal of work in recent decades has shown. From the per-
spective of the analysis of his own music, perhaps the most significant
omissions are metaphorical cognition, temporal cognition, and embodied
cognition. Later chapters of this book will explore these modes of expe-
rience. Babbitt outlines ways in which his music can be comprehended,
but experience is more than comprehension. To be affected by music, or
to physically respond to it, is to use aspects of cognition on which Babbitt
is silent. Temporality is not merely a facet of epistemology, as he describes
it; it is a part of human experience music can bend with extraordinary cre-
ativity. Thankfully, Babbitt’s music engages aspects of cognition his psy-
chology and philosophy do not.
Comprehension and memorability were primary concerns for Babbitt’s
music psychology not only because of cognitive science’s concerns at the
time, but because of music’s. As one of the self-​declared “legitimate, if aban-
doned, children of the Schoenbergian revolution” (Babbitt [1974] 2003, 333),
Babbitt felt keenly not only that the musical world he inhabited was new and
revolutionary, but also that his listeners, those who found themselves thrust
into this new world, were bound to be disoriented. He likens the problem of
hearing a new musical language with hearing an unfamiliar verbal language,
one which “you have never heard and of which you know nothing.” This is
not only a challenge for individual listeners: the problem of proliferating new
musical languages and the attendant lack of comprehension is a “crisis” for
the musical community (Babbitt 1987, 171–​72). And so, in Babbitt’s writings,
as in Schoenberg’s, comprehensibility acquired a particular urgency, even
to the exclusion of other topics (see Schoenberg [1941] 1975, 214–​15; also
Schoenberg 1995).
Yet simply as a theory of comprehension, Babbitt’s views on psychology,
when applied to his own compositions, must remain aspirational. The nub
of the problem is this: there are far more relationships in Babbitt’s music than
can be actively recognized in any single listening. Some relationships, hap-
pily, present themselves straightforwardly for perception—​it is not too diffi-
cult to hear that the alto 1 and alto 2 parts in mm. 28–​30 of Glosses develop the
same trichord ordering in inversion (Figure 1.2). But many others approach
imperceptibility. Cross-​references, the primary determinant of surface or-
ganization in Babbitt’s later music, are an egregious example, particularly
when linked to the time-​point system—​as they are in most of Babbitt’s
 Construction, Cognition, and the Role of the Surface 63

Figure 2.3. String Quartet No. 6, Violin 1, mm. 5–​7, with references and time
points indicated

pieces written after 1984. Figures 2.3, 2.4, and 2.5, from String Quartet No.
6, demonstrate.33
Using the cross-​reference technique outlined in Chapter 1, mm. 5–​7 in
the violin I part index four components of the array for violin II. The tri-
plet in m. 5 divides its time-​point interval into three, although the first note
of that triplet is silent. According to Babbitt’s scheme, division into three
indexes an array component that is three notes long, and since the reference’s
second and third pitch classes are C and B♭, the referent’s are likewise. Such a
component is found in the sixth aggregate of the second block for violin II’s
array, as shown in Figure 2.4a. The next two references, in m. 6 and the first
beat of m. 7, correspond to the next two components in the same line of vi-
olin II’s array, also shown in Figure 2.4a (as discussed in Chapter 1, Babbitt
often references consecutive components like this). At the end of m. 7, the
quintuplet signals a reference to a five-​note component whose third, fourth,
and fifth notes are D, C♯, and E. This component, too, is found in violin II’s
array, although now in the second block, as shown in Figure 2.4b. That both
referents are in violin II’s array is not coincidental. As discussed in Chapter 1,
the quartet, like many of Babbitt’s works from the 1990s, presents a pattern
in which each instrument consistently references one instrument’s array for
a given span.
But an array is an abstraction; it is not heard.34 What a listener responds
to is an array’s realization, the actual music that draws from the array. And
whatever may come through clearly in visual examination of Figure 2.4
recedes into obscurity in the music containing the referents. The excerpts

33 Aspects of this analysis originate in Colson 2015. For the fuller context of Figure 2.3 and its time-​

point aggregate, see Figures 7.8 and 7.9. A related argument is developed in Dubiel 1997, 38–​39.
34 Joseph Dubiel has argued this point at length; see particularly Dubiel 1990b.
64 Thinking In and About Music

Figure 2.4a. String Quartet No. 6, Violin 2 array block IV. Components referred
to in Figure 2.3 are shaded. Parentheses indicate pitch classes inserted between
notes of the series; brackets indicate serially expected notes that are absent.

Figure 2.4b. String Quartet No. 6, Violin 2 array block II. Component referred
to in Figure 2.3 is shaded. Parentheses indicate pitch classes inserted between
notes of the series; brackets indicate serially expected notes that are absent.

referred to are shown in Figure 2.5. The barriers to actively perceiving the
references are so numerous and daunting that I struggle to prioritize them.
The distance between the music of Figure 2.5a and that of Figure 2.3 is nearly
140 measures. The referents are contained in only a single thread of a twenty-​
four-​line texture—​four instruments each presenting six-​line arrays. The
music in both referents is itself extraordinarily active: partially a result of the
fact that both Figure 2.5a and Figure 2.5b, like every other moment in the
piece, are also referencing strings of different array components at the same
time that they are referents for Figure 2.3. And then there is the question of
the time-​point system. To hear, for instance, the triplet in m. 5 of Figure 2.3 as
a division into three requires hearing time points bracketing that triplet, even
though the sixteenth-​note pulses defining the time-​point system are hardly
 Construction, Cognition, and the Role of the Surface 65

Figure 2.5a. String Quartet No. 6, Violin 2, mm. 142–​56, with referents of
Figure 2.3 indicated

Figure 2.5b. String Quartet No. 6, Violin 2, mm. 48–​51, with referent of Figure
2.3 indicated

an explicit feature of the music, the time-​point modulus is only intermit-

tently articulated, and the first note of the triplet is not even sounded.35 And
this, again, is only one cross-​reference; there are three other instruments, all

35 It happens, in this case, that other instruments do have an attack at that time point. But many

time points are not articulated by any instrument. As can be seen in Figures 1.7 and 7.8 respectively,
the first downbeats of None but the Lonely Flute and String Quartet No. 6 are examples. For more on
the challenges the time-​point system presents to perception, see Dubiel forthcoming.
66 Thinking In and About Music

making references simultaneously, and each of their referents also simultane-

ously refer elsewhere (and all to different places).
To say that Babbitt’s cross-​reference technique strains credulity seems too
kind. As a theory of active listening, it is frankly ridiculous.
But perhaps Babbitt’s psychology does not require active listening?
Perhaps cross-​references and other challenging relationships are not meant
to be noticed? Babbitt never explicitly addresses this, and the intellectual
sources from which he draws give conflicting answers. From a cognitive per-
spective, one perceiving a signal need not be aware of its structure in order
to benefit from that structure. Most who speak and understand a language
are unaware of the redundancies that enable comprehension of that lan-
guage. When Schenker (1954, 60) notes that a “genius, like a sleepwalker,
often finds the right way,” he refers to compositional creation, not listening,
but his remark may also be read as support for a psychology that does not
require fully conscious active recognition. And Leibniz (1898, 422), too,
is encouraging: “Music charms us, although its beauty consists only in the
harmonies of numbers and in the counting (of which we are unconscious
but which nevertheless the soul does make) of the beats or vibrations of
sounding bodies.” Just as Leibniz’s listener can count and compare harmo-
nious pulsations unconsciously, perhaps Babbitt’s can unconsciously benefit
from cross-​references.
But unconscious encounters are completely inconsistent with empiri-
cism. Empiricism requires observation. If some claim is not based in obser-
vation, a logical empiricist will find it unfounded—​meaningless. Recall that
the primitive of Carnap’s Aufbau is recollection of similarity. Data cannot
simply wash over the logical empiricist; active intervention is required in
order to bind the logical connections that construct data into facts and the-
ories. To put it plainly, if a relationship—​say, between a cross-​reference
and its referent—​is not observed, it does not exist. And Babbitt, I find, is
fundamentally a logical empiricist on this point. Three kinds of comments
suggest this: he discusses how one can learn to hear musical relationships,
he mentions that a performer might bring them out, and he questions the
competency of those who cannot notice them (see, e.g., Babbitt 2001, part 3;
also Babbitt 1987, 67 and 143–​44). Such remarks imply that his psychology
requires active observation. While he is willing to acknowledge the limi-
tations of any single hearing—​“no one would say for a moment that if you
don’t hear this the first time, the piece falls apart” (Babbitt 1987, 67)—​the
 Construction, Cognition, and the Role of the Surface 67

relationships he discusses and, one may therefore assume, composes, are

surely intended to be noticed.
And so Babbitt seems to have set an unreachable standard. But this is not
to say that any given listening is a failure. There are more relationships in his
music than can be heard on a single listening, but a second listening might
bring out more, or different ones. Some relationships might never be heard
on any listening, but remain there to be observed, if ever one’s ears and atten-
tion become adequately sensitized. What Babbitt promotes, as composer and
theorist, is an ethic of striving. It is “virtuous . . . to create a standard to which
they might aspire” (Babbitt [1989] 2003, 433). We, the aspirants, can find in
his music an endless bounty of “richness of relation”—​but only if we seek it.
 3
The Seam in Babbitt’s
Compositional Development
Composition for Tenor and Six Instruments, Its
Precedents, and Its Consequences

“My Most Difficult Piece”

The years just around 1960—​approximately from the 1957 composition of

Partitions to the 1965 composition of Relata I—​were transformative years for
Babbitt’s compositional technique in three important respects. The search for
a coherent mapping between pitch and rhythmic intervals led to the invention
of the time-​point system. A deepening frustration with performative limita-
tions, especially given the new rhythmic difficulties of his time-​point music, led
to the adoption of electronic synthesis in the 1961 Composition for Synthesizer,
following some four years spent mastering the R.C.A. Mark II Electronic Music
Synthesizer. And finally, and arguably most consequentially, an interest in a more
varied and ramified contrapuntal practice gradually led him away from the four-
square (34) partitioning of the trichordal array, a process that ultimately culmi-
nated in the all-​partition array of Relata I. Balancing this increase in contrapuntal
complexity is a simplification of the lines forming each array. While in the 1940s
and ’50s the lines of Babbitt’s arrays are almost always derived series, generated
from the various segments (most often, trichords) of an underlying series, the
lines of Babbitt’s arrays henceforth would present straightforward, concatenated
iterations of series forms with identical hexachordal content.1

1 The first two of these transformations are the subject of Babbitt (1962) 2003. Babbitt’s clearest

statement on the move toward all-​partition arrays is in Babbitt (1976b) 2003, 354, as discussed subse-
quently. Dubiel 1990b reviews aspects of the evolution of Babbitt’s contrapuntal practice in some de-
tail, remarking on the “more constant and literal . . . employment of the set” in all-​partition arrays as
contrasted with Babbitt’s earlier practice (225). Mead 1994, 125 demarcates the beginning of Babbitt’s
“second period” at the year 1964, characterizing the all-​partition array as the “structural hallmark” of
that period. In this chronology, Composition for Synthesizer and Vision and Prayer (1961) are outliers,
in that they were both completed after 1960 yet contain trichordally derived lines. On the former, see

Thinking In and About Music. Zachary Bernstein, Oxford University Press. © Oxford University Press 2021.
DOI: 10.1093/​oso/​9780190949235.003.0003
 The Seam in Babbitt’s Compositional Development 69

Standing at the intersection of all of these developments is Composition

for Tenor and Six Instruments (hereafter CT6), of 1960. As will be shown, this
piece represented something of an endpoint for Babbitt. Certain consider-
ations had evidently led him along a path that he felt his current practice
could no longer support, necessitating a change. In his only two significant
published comments about the work, he appears to regard CT6 as a failure,
describing it in an almost apologetic tone that he used for no other piece of
his. These comments reveal much about the creative struggle surrounding
the work.

In perhaps the last of my instrumental works in which trichords appear in

a foreground role, Composition for Tenor and Six Instruments, I again em-
ployed an explicit unfolding of the set content, but the work is so much
more complex than the Second Quartet, the instrumental lines are so
rarely generatively univocal, and the sections so sharply contrasted in so
many respects, that—​although there might be those who would judge
its proportions more nearly divine than those of others of my works—​
convincing continuity depends crucially on invariants of order embedded
in invariants of content. (Babbitt [1976b] 2003, 354).

The piece of mine that Stefan [Wolpe] pressed me most about . . . was a
piece called Composition for Tenor and Six Instruments. He heard a per-
formance which the Group for Contemporary Music did up at McMillin
[Theater] and professed to love it. Now I must confess to you, I think the
reason he felt that was because in many ways it was my most difficult piece
both to perform and to hear. It was a piece that made many people very
angry. It had long, long, long periods of unchanging notes, or very, very
slow-​changing pitch combinations, which was not like my usual music
and which intrigued Stefan. . . . Now that piece we did go over in enor-
mous detail, for two reasons, the first being the tempo organization. It’s not
the only piece of mine in which I’ve done this, but it’s the most extreme
piece. I decided after that piece that I would have to find some sort of way of
writing music that was not as difficult. It was just too much. We also had the
problem of the tenor. The tenor in that piece used only phonemes, and the

Morris 1997, 88–​92. For the latter, the lingering trichordal approach may be due to the piece’s origins
in an uncompleted 1954 setting for soprano and piano, although that version differs substantially
from the completed electronic setting (the soprano and piano version appears in Fulmer 2016, A-​
1–​A-​8). Finally, recall that Babbitt’s engagement with cognitive science also began around 1960, as
discussed in Chapter 2.
70 Thinking In and About Music

phonemes were indeed chosen in order to either contrast or blend with the
instruments. Sometimes it worked very well, and sometimes it didn’t. . . .
He wanted to know about phonemic structure. Obviously he knew not a
great deal about vocal acoustics and vowel acoustics, and many of us were
involved in this, not merely for musical purposes, because we were involved
in electronics. He did not know about the Haskins Laboratory in New York.
I told him about it. . . . That is the piece with which I can remember the
most discussion about the organization—​spatial organization, division of
the musical space, as well as musical time, and possible analogies between
the two. (Babbitt 1983)

There is much to unpack in these quotations, and we will be returning

to them throughout the chapter. For now, let us simply observe Babbitt’s
expressed concerns about all three of the aforementioned areas in which
his compositional practice changed around 1960. There are new analogies
pursued between tempo and “space”—​that is, using a metaphor common
in Babbitt’s writings, pitch (see, e.g., Babbitt 1987, 20). Although he does
not specify it here, he is apparently referring to the piece’s use of the time-​
point system—​a technique that publicly debuted in CT62—​and, as will be
discussed, the application of the time-​point system to determine not only
local rhythm, but also global relationships among the piece’s various sections.
The rhythmic challenges this new system engenders, as well as such subtle-
ties as the matching of phonemes and instrumental timbre, created appar-
ently insurmountable performance difficulties. Either, as mentioned in these
quotations, he would have to change his compositional practice or turn to
electronic synthesis, and in the years following the composition of CT6 he
would explore both options. Finally, concerns for “convincing continuity”
led to a radical revision of his contrapuntal practice: this would, he thought
in 1976, be the last of his instrumental works built from trichordal arrays.3

2 Published catalogues of Babbitt’s works (e.g., Mead 1994, 267) place the composition of “Sounds

and Words” (1960), which also uses the time-​point system, before that of CT6. While these cata-
logues are based on a list Babbitt personally compiled (Andrew Mead, personal communication),
elsewhere Babbitt claimed an alternative chronology: “But [CT6] was very elaborate and I wanted to
[compose using phonemes] on a simpler basis. So I wrote a piece called ‘Sounds and Words’ ” (Babbitt
1986c; also 1988, 154). A piece of supporting evidence for this latter chronology is that “Sounds and
Words,” unlike CT6 and Babbitt’s pieces from the 1940s and 1950s, does not use derived series. As will
be discussed at length, CT6 marks the turning point away from derived series. Babbitt (1962) 2003
suggests that he conceived the time-​point system in conjunction with his adoption of electronic syn-
thesis, but his first electronic pieces would not appear until 1961.
3 As discussed in Mead 1984 and 1994, trichordal arrays returned to Babbitt’s practice shortly

thereafter, beginning with Minute Waltz (or) 3/​4 ± 1/​8 (1977), and would remain a central part of his
practice.
 The Seam in Babbitt’s Compositional Development 71

This chapter will examine the changes to Babbitt’s contrapuntal and

rhythmic practice in the years around 1960. Following Babbitt’s suggestion
in both quotations, and in light of plentiful textual evidence within the piece
and its predecessors and successors, CT6 is taken as the pivotal piece in his
development at this time. Nevertheless, as the competing concerns that would
come to a head in CT6 assumed significance gradually, the investigation will
perforce extend some years back. In sum, I seek to interrogate the transition
between Babbitt’s first and second periods, as they are described in Mead
(1994), with the aims of elucidating how and speculating why Babbitt’s post-​
1960 practice differs from his first decade of published compositional output.
As significant as the transformation was, however, one conclusion that can
be drawn initially is that the various changes in Babbitt’s compositional prac-
tice were motivated primarily by an essentially practical set of concerns, more
than any substantive ideological rethinking. The rhythmic considerations
stem from an attempt to solidify the pitch-​rhythm analogies he had already
been pursuing for the previous two decades. The performative challenges re-
sult, primarily, from this desire to forge a more powerful rhythmic analogy as
well as, in CT6, from an attempt to make the tenor’s timbre—​its phonemes—​
interact meaningfully with the timbre of the various instruments.4 And, fi-
nally, the changes to his contrapuntal practices were apparently motivated
by the combination of an interest in gestural and textural variety and com-
plexity, as well as a desire to explore certain patterns of invariants, and a
pragmatic, compensatory impulse to preserve the perception of hierarchical
relationships that had always interested him. In observing Babbitt pragmati-
cally balance his aesthetic goals, we can gain insight into just what those goals
are and see how his compositional technique evolved to address them.

Trichordal and All-​Partition Arrays

While the developments surrounding CT6 also touched on rhythmic tech-

nique and performative limitations, subjects to be discussed subsequently,
perhaps the most dramatic change concerned contrapuntal technique. To
quickly see the result of Babbitt’s changes in this respect, it will be helpful
to compare two arrays from the boundary points of his evolution around
1960. Figure 3.1a shows the array for the second section of Partitions, one of

4 Babbitt was aware early on that the time-​point system, in the applications he envisioned for it,

would create severe difficulties for performers. See Babbitt (1962) 2003, 133.
Figure 3.1a. Array for Partitions, mm. 9–​16

Figure 3.1b. Series for Partitions, presented in mm. 1–​3, with trichord
orderings indicated
 The Seam in Babbitt’s Compositional Development 73

the last of Babbitt’s trichordal-​array pieces from this period. It presents four
lines, each presenting derived series and each contributing a single trichord
to consistent (34) aggregates. The trichords develop orderings derived from
the underlying series, which is introduced in the piece’s opening measures;
see Figure 3.1b. As is normal for trichordal arrays, the array presents the fif-
teen combinations of four trichords arranged into complementary pairs and
a tutti.
The partitioning is perfectly regular, but the lines present distinct
trichordal and hexachordal content, as detailed in Figure 3.1a. The upper
two lines are generated from the trichord ordering <+5, −2> grouped into
C-​type hexachords, while the lower two lines are generated from <−2, −1>
grouped into A-​type hexachords. As the piece continues, the trichords and
hexachords of each line-​pair will change with each section, proceeding
through various combinations of trichord orderings and grouping those
trichords into assorted combinations of A-​type, B-​type, C-​type, and E-​type
hexachords. No two consecutive arrays present the same trichord orderings
or hexachords in the same lines.5
Figure 3.2, showing the first eleven aggregates of the first array in Relata
I, presents a starkly different approach.6 As indicated at the bottom of the
example, the partitioning is more varied than that of Figure 3.1a. Indeed, as
an all-​partition array, it is maximally varied: it traverses the partitions from
(112) to (121). Yet lines of the array are simply series forms: a listener need
infer no underlying series to understand them as related. Moreover, each line
concatenates series forms with identical discrete hexachords, by way of sec-
ondary sets. The contrapuntal whirlpool is composed of equivalent drops of
water.7
It is worth noting at this juncture that there are no technical reasons why
an all-​partition array requires each of its lines to be iterations of a series
or that each line should be restricted to a single pair of complementary
hexachords. Arrays with exhaustive lists of partitions could be assembled
with derived series, or with multiple, distinct series, or with any number of

5 For more on Partitions, see Mead 1994, 113–​18.

6 The first seven aggregates of Figure 3.2 are adapted from Babbitt (1970) 2003, 243. The opening
eight measures of the piece, described by the composer as an “introduction” (ibid., 239–​40), do not
use the array shown.
7 Not shown in Figure 3.2 are the forty-​eight instrumental lines that come together to form the all-​

partition array discussed. In the opening section, each instrumental line presents series forms; later on,
they present derived series. These lines do not independently form aggregates, and therefore the de-
rived instrumental lines are probably best considered a secondary aspect, rather than array-​constitutive
in their own right. For more on this aspect of Relata I, see Babbitt (1970) 2003, 244–​47 and 251.
Figure 3.2. Array for Relata I, mm. 8–​16
 The Seam in Babbitt’s Compositional Development 75

other imaginable configurations.8 Rather, as the first quotation cited pre-

viously has it, it is Babbitt’s understanding of “convincing continuity”—​
which is to say, apparently, his understanding of the limits of human
perception9—​that restricts the lines of his all-​partition arrays, not any in-
herent limitations of the twelve-​tone system. As discussed in Chapter 2, the
limits of perception were a topic of deep concern for Babbitt in the years
surrounding the composition of CT6 as a result of his adoption of elec-
tronic synthesis. There is no evidence that the changes to Babbitt’s contra-
puntal practice following CT6 were the result of any particular scientific
paper or experiment. However, if “every musical composition justifiably
may be regarded as an experiment, the embodiment of hypotheses as to
certain specific conditions of musical coherence” (Babbitt [1962] 2003,
109), and “coherence and continuity” depend on “principles of relatedness”
that—​structure being cognitive, for Babbitt, rather than immanent (Babbitt
1998a, 25)—​depend on the “perceptual and conceptual abilities of the lis-
tener” (Babbitt [1958] 2003, 50), one might suppose that Babbitt himself—​
his own “hypothetical other,” in any case (Babbitt [1989] 2003, 434)—​could
not hear coherence or “convincing continuity” in CT6. The music did not
support the “recollection of similarity” he required for the verification of
musical relationships. In short, expanding his practice to include a greater
variety of contrapuntal configurations results in new perceptual challenges
that, he apparently felt, needed to be offset: if the partitions were to be more
complex, the lines themselves would have to be simpler to be perceptibly
related.

“Nearly Divine” Proportions

Neither divinity nor the “divine proportion”10 figured much in the philos-
ophy of an “unreconstructed logical empiricist” (quoted in Bortz 2005, 96),
but one might guess what about the proportions of CT6 Babbitt would have
considered “most nearly divine”: the proportions of the piece’s sections are

8 The music of Robert Morris has explored several varieties of deeply un-​Babbittian all-​partition

arrays. See Alegant 1993, 148–​62 and Morris 2010, 80–​81.

9 Explicit statements clarifying Babbitt’s position that “continuity” is a matter of perception and

thus subject to the limits of perception, rather than an inherent musical attribute, appear in Babbitt
(1958) 2003, 50; Babbitt (1968) 2003, 225; and Babbitt 1998a, 25.
10 The “divine proportion” is a synonym for the golden ratio.
76 Thinking In and About Music

determined by the piece’s series. In CT6, the time-​point system determines

both local rhythm and the durations of sections. This large-​scale applica-
tion of the time-​point system seems to have been used in only one other of
Babbitt’s pieces, “Sounds and Words,” written just after CT6. In this tech-
nique, sectional duration is defined by the intervals of the piece’s series
class.11
As shown in Figures 3.3 and 3.4, in both “Sounds and Words” and CT6,
the durations of the first eleven sections of the piece correspond to the
intervals of the series. In “Sounds and Words,” the durations of each sec-
tion correspond directly to the intervals of the piece’s series measured in
seconds, while in the more extended CT6, the durations of each section,
measured in seconds, correspond to the intervals of the series multiplied
by four.12 Following these first eleven sections of “Sounds and Words,”
there is a twelfth section; apparently inserted for the sake of array comple-
tion (it allows the first half of the piece to complete three four-​aggregate-​
long arrays, each aggregate being presented equivalently and coextensively
in pitch and time points), its duration of twelve seconds is congruent, as-
suming modularity, to the interval of a unison, and therefore does not play
a role in the large-​scale presentation of the series. The first eleven sections
of the second half of the piece present the retrograde of the durations of the
first eleven sections—​an operation that corresponds to the retrograde in-
version of the original series, given that durations function as intervals in
the time-​point system—​before again concluding with a neutral section of
twelve seconds’ duration.
The overall scheme in CT6 is quite similar: the primary global structure
of the piece presents the intervals of the series (now measured in units of
four seconds) twice, with the second the retrograde inversion of the first.
Since each section in CT6 presents its own array, there is no need for neu-
tral sections. There is, however, an interlude between the two large-​scale

11 Although CT6 and “Sounds and Words” are the only two pieces completely divided into (more

or less) discrete sections whose durations are determined by large-​scale time-​point intervals, certain
later pieces such as My Ends Are My Beginnings and Melismata (1982) use time-​point intervals to
determine the duration of pitch-​class aggregates (see Mead 1987, 218–​20; 1994, 193–​95). The dura-
tional scheme of “Sounds and Words” is mentioned in Dubiel 1990a, 78n.26.
12 The word “section” is probably more appropriate, in its familiar usage, to CT6 than to “Sounds

and Words,” given both the difference in overall scale between the two pieces (14′28″ versus 2′36″, re-
spectively) and the fact that in CT6 each section presents a relatively discrete array, while the sections
in “Sounds and Words” present only single aggregates of four-​aggregate-​long arrays. But since CT6 is
the primary focus here and no other term is readily available, I will use it for both pieces.
 The Seam in Babbitt’s Compositional Development 77

Figure 3.3. Section durations, tempi, and time-​point moduli of “Sounds

and Words”

statements of the series. The durations of the interlude’s three sections, meas-
ured in seconds, each somewhat longer than the durations defined by the se-
ries, are not divisible by four, and therefore do not correspond to any interval
of the series in the established grammar of the piece. The middle section of
the interlude—​by far the longest section of the piece—​is probably the sec-
tion Babbitt referred to as having “long, long, long periods of unchanging
notes.” The first and last aggregates of that section, in particular, each take
about forty seconds (during which they each present only twelve discrete
attacks). In the first, the tenor is asked to hold a single note, 𝆑, for a breath-
taking twenty-​five seconds.
Figures 3.3 and 3.4 detail the overall durational patterns of both works.
They also include information about the time-​point modulus and tempo
of each section. Throughout “Sounds and Words,” the modulus evenly
divides the duration of the section, while in CT6 this is often not the case.13
Since the modulus in both pieces is equivalent to the written measure, this
means that each section of “Sounds and Words” is composed of measures

13 Indeed, it does not appear that the moduli or tempi are systematically determined in CT6. They

may simply be loosely determined byproducts of the desire to complete a certain amount of pitch-​
class and time-​point material within a given duration.
78 Thinking In and About Music

Figure 3.4. Section durations, tempi, and time-​point moduli of Composition for
Tenor and Six Instruments

of equal duration, while in CT6 there is sometimes a residue measure—​a

measure at the end of a section just long enough that the section will have
the correct duration. For example, the first section of CT6 divides forty-​
four seconds into a three-​second modulus, and 44/​3 = 14 r 2; accord-
ingly, the section begins with fourteen three-​second-​long measures and
ends with one two-​second-​long measure ( and  in the prevailing tempo
of 𝅘𝅥 = 60).
It appears that there is a dramatic difference in the use of tempo in the two
pieces, but this is mostly a notational illusion. “Sounds and Words” maintains
 The Seam in Babbitt’s Compositional Development 79

an even 𝅘𝅥 = 60 tempo almost entirely throughout the piece, effecting the

changing modulus with complex time signatures and tuplets. For instance,
a modulus of 2.75 seconds requires a time signature of  and a 12:11 tuplet
that spans the bar. CT6 simplifies this notationally using tempo changes.
This simplification is made urgent because CT6 uses many offset and nested
tuplets, techniques not found in “Sounds and Words.”
In the second quotation given previously, Babbitt describes CT6 as “the
most extreme piece” to use this large-​scale time-​point technique. Given its
many similarities with “Sounds and Words,” what about the earlier piece is
more extreme? One concession to practicalities that “Sounds and Words”
makes but CT6 does not regards their treatment of the shortest sections,
those corresponding to interval 1, as shown in Figures 3.5 and 3.6. In
“Sounds and Words,” accommodating the systematically expected time-​
point material in those sections would require at least four moduli. Within

Figure 3.5. “Sounds and Words,” mm. 36–​43

Figure 3.6. Composition for Tenor and Six Instruments, mm. 44–​56. The
downbeat is taken as time point 0. A single slash indicates the end of a trichordal
segment; a double slash indicates the end of a derived series.
 The Seam in Babbitt’s Compositional Development 81

the sectional duration of one second, each modulus would be one quarter of
a second and the time-​point unit interval (a twelfth of a modulus) would be
one forty-​eighth of a second—​obviously far shorter than can be accurately
performed or perceived. Babbitt simply ignores the time-​point dimension
in these sections, filling the duration of the sections with a string of five or
seven even note values. Oddly enough, the remaining sections of the piece
continue as though the expected time points in Sections XI and XIII had
been successfully completed. As with many violations of serial structure in
Babbitt’s music, the surrounding context seems oddly unresponsive to the
fact that serial expectations have been flouted.14
In Section IV of CT6, shown in Figure 3.6, which should be completed
in four seconds, Babbitt does make some concession to practicality: most of
the piece uses approximately eight pitch-​class aggregates per section, but this
section uses only four. The time-​point structure, as indicated in Figure 3.6,
is also unusually simple, presenting two successive derived series generated
by the trichord ordering <+6, −5>. These require a minimum of 141 time-​
point unit intervals, here realized as twelve moduli and a one-​unit upbeat.
The upbeat itself is a bit of a concession, extending the material of Section IV
into the time span of Section III. However, dividing four seconds into twelve
moduli means the modulus is one third of a second and the time-​point unit
interval is one thirty-​sixth of a second (thirty-​second-​note triplets at 𝅘𝅥 =
180). Clearly, despite the concession to practicality afforded by halving the
section’s expected material, this is impossible to perform.15 Babbitt may have
even recognized this and relented during the composition of the piece, for
the second four-​second-​long section, Section XXII, presents only one time-​
point aggregate, requiring only four moduli. The fact that this large-​scale
application of the time-​point system necessitated adjustments for the sake
of realistic performance may have been one of the things that led Babbitt to
abandon the technique.16

14 For more on anomalous deviations from serial expectations, see Chapters 6 and 7. A related

deviation occurs in My Ends Are My Beginnings. As detailed in Mead 1994, 193–​95, aggregates
corresponding to time-​point intervals 1 and 2 are set with durations corresponding to 13 and 14
(congruent to 1 and 2 mod 12, respectively).
15 Babbitt once sought to test the limits of human performance, finding that “15 alterations a

second is about the limit of the muscular system” (Babbitt 2002; also Chadabe 1997, 18). In his later
music, perhaps as a result of this experiment, Babbitt does not exceed this limit.
16 The large-​scale time-​point technique found in “Sounds and Words” and CT6 is remarkably sim-

ilar to the largest level of Charles Wuorinen’s “nesting” method, described in Wuorinen 1979, 149–​62
and used in many pieces throughout Wuorinen’s career (see also McConville 2011). The only differ-
ence at the global level is that Wuorinen expresses the intervals of a series using twelve sections, not
eleven, the twelfth representing the interval between the series’s last note and its first note.
82 Thinking In and About Music

Generative Polyvocality

The first quotation given previously compares CT6 to Babbitt’s String

Quartet No. 2, and the comparison is apt: both pieces develop the segments
of a twelve-​tone series with derived series before eventually presenting the
underlying series directly. But the quotation goes on to say that CT6 is “so
much more complex than the Second Quartet, the instrumental lines are
so rarely generatively univocal, and the sections so sharply contrasted in so
many respects” that profound changes in compositional technique were re-
quired. The mention of rare generative univocality appears to refer to the fact
that many instrumental lines and even single pitches play roles in multiple
simultaneous generated structures. This is hardly unprecedented; indeed,
an instance of what may be called “generative polyvocality” can be found as
early as the second aggregate of Composition for Four Instruments. As can
be seen in Figure 3.7, this aggregate presents derived series generated from
three different trichordal set classes, each articulated with distinct means.
Two of these—​the generations from (013) and (027)—​are in addition to the
prevailing trichordal array, which—​as is normal for a trichordal array—​also
presents hexachordal aggregates across the array’s first two aggregates and
linear aggregates (i.e., derived series) across the array’s first four aggregates.
All told, each note in Figure 3.7a is part of five distinct aggregate formations.
This is not an unusual density of aggregate formation in Babbitt’s music
preceding CT6.
That is, generative polyvocality in and of itself is not what makes CT6 espe-
cially complex. Rather, what seems to be innovative about CT6 in this regard
is the quantity of simultaneous generated lines (up to thirteen), the uneven
partitioning of those lines, the variety of kinds of generation (from dyads,
trichords, and tetrachords, often intermixed in complex patterns), and, es-
pecially, the fact that simultaneous generated lines unfold across sharply di-
vergent, loosely coordinated, and sometimes ambiguous time spans. Thus it
is the work’s arrays that are the primary source of its complexity—​and those
arrays are replete with generative polyvocality.

Figure 3.7a. Composition for Four Instruments, mm. 7–​9

 The Seam in Babbitt’s Compositional Development 83

Figure 3.7b. Generation of Figure 3.7a from (014), articulated using register

Figure 3.7c. Generation of Figure 3.7a from (013), articulated using temporal order

Figure 3.7d. Generation of Figure 3.7a from (027), articulated using dynamics

Extending the Trichordal Array

While trichordal arrays such as that of Figure 3.1a are by far the most
common of Babbitt’s array types in the late 1940s and ’50s, more varied and
complex structures, often introduced as expansions of the basic trichordal
array framework, appeared alongside them throughout this period. There
are essentially four techniques by which the trichordal array, and particu-
larly its regular (34) partitioning, is extended: pitch classes “swapped” across
84 Thinking In and About Music

aggregate boundaries, intrinsically asymmetrical arrays, arrays in which

derived segments systematically extend across aggregate boundaries, and
arrays with more than four lines. Each of these techniques is used in several
pieces from 1948 to 1957, with all of them featuring in All Set (1957).17 The
motivations for these techniques are difficult to generalize about apart from
the specific compositional contexts within which each appears, but one can
observe a desire for increased gestural and textural variety and an interest
in invariants unavailable in a standard trichordal array. The pursuit of these
objectives appears to have led Babbitt toward the manifold complexities of
CT6 and, from there, to the dramatic changes to his practice instantiated by
the advent of the all-​partition array.
“Swapping” alters an array by exchanging pitch classes across aggregate
boundaries, as demonstrated in Figures 3.8a and 3.8b.18 Figure 3.8a presents
an array from Semi-​Simple Variations (1956). The irregular partitioning in
this array is a result of swapping applied to the array shown in Figure 3.8b: pcs
5 and 6 have been swapped across the boundary between the first and second
aggregates and pcs e and 0 have been swapped across the boundary between
the third and fourth aggregates. Context clarifies the extent to which Figure
3.8a is heard in light of Figure 3.8b: every other array in the piece is a regular
(34) trichordal array, much like Figure 3.8b.
A result of, and plausible motivation for, these swapped pcs is increased
invariance. Effectively, two distinct trichordal mosaics are re-​partitioned to
become a single (4222) mosaic. That is, the various slots of Figure 3.8a are
filled with only two dyads and two tetrachords, but those of Figure 3.8b pre-
sent fully eight distinct trichords. As has already been observed, invariance is
a foundational desideratum of Babbitt’s compositional practice.
Swapping such as this is a feature of much of Babbitt’s early music, from
at least Composition for Twelve Instruments on.19 Whether for the sake of
invariance or textural variety, it provides a relatively unobtrusive means of
introducing irregular partitioning. As Robert Morris notes (1987, 223), the

17 I am grateful to Andrew Mead for pointing out the pertinence of All Set in this respect.
18 On the theoretical principles enabling swapping, see Morris 1987, 220–​23. For more on Semi-​
Simple Variations and its array structure, see Wintle 1976.
19 On swapping in Composition for Twelve Instruments, see Hush 1982–​83, 178–​97 and 205, where

it is referred to as “dyad exchange.” Both Dubiel 1990b, 222 and Mead 1994, 31 discuss the swapped
opening array of Partitions. Given the chronological position of Partitions—​and its suggestive title—​
it is tempting to read that instance of swapping as premonitory; Mead refers to it as a “harbinger of
the future.” While Partitions is certainly a vivid example of it, swapping was not a technical innova-
tion in 1957.
 The Seam in Babbitt’s Compositional Development 85

Figure 3.8a. Semi-​Simple Variations, array for Variation III. A single slash
indicates the end of a trichordal segment; a double slash indicates the end of a
derived series. This notation will be used throughout when segments do not
consistently coincide with aggregate boundaries.

Figure 3.8b. “Restored” version of Figure 3.8a, undoing the swapping

essential syntax of the trichordal array—​its linear structure, levels of aggre-

gate formation, and derivation from an underlying series—​remains intact.
While some irregular partitions such as those of Figure 3.8a can be con-
sidered swapped deviations from a regular array, others cannot be. A striking
early example of the latter situation comes in the final section of Two Sonnets,
as shown in Figure 3.9. Following the completion of a Petrarchan sonnet form,
“That Nature is a Heraclitean Fire and of the Comfort of the Resurrection”—​
the second of the two sonnets by Gerard Manley Hopkins set in the piece—​
concludes with three poetic “codas,” each consisting of a couplet and a
burden line.20 The array of Figure 3.9 underlies the opening of these poetic
codas. The setting of these lines consistently uses (623) partitioning: the voice
dominates the texture, contributing two complete trichords to every aggre-
gate, accompanied only by spare instrumental dyads. The instrumental lines
present trichordally derived series stretched across six, rather than the usual
four, aggregates, so that individual trichords of those series span two adja-
cent aggregates. In systematically—​that is, regularly, and not as the result of
swapping—​presenting a mismatch between derived segments and aggregate

20 “Codas” is Hopkins’s term for the added couplets and burden lines, as quoted in Johnson 1972,

236. For more on Two Sonnets, see Chapter 5.

86 Thinking In and About Music

Figure 3.9. Array for Two Sonnets, “That Nature is a Heraclitean Fire,”
mm. 88–​109

boundaries, the instrumental parts present yet another way in which this
array deviates from standard trichordal arrays.21 Finally, this array includes a
startling diversity of trichordal content. Over the course of the final passage,
the ensemble summarizes nearly all of the trichord orderings available in the
piece’s series, excepting only those that cross the hexachordal boundary. As
will be revisited in Chapter 5, a text-​setting motivation may well account for
this change of strategy.
Most of Babbitt’s arrays in the 1940s and ’50s, even those that deviate
from trichordal array norms in other respects, use only four lines. One
early exception is Composition for Twelve Instruments, which presents con-
sistent twelve-​part counterpoint.22 Despite this thick texture, there are nu-
merous respects in which Composition for Twelve Instruments is remarkably
simple. The lines of the first half of Composition for Twelve Instruments are
all standard transformations of a single twelve-​tone series. These series
are aligned in a bewildering number of aggregate formations—​as Babbitt
([1961b] 2003, 94) describes, the array “display[s]‌(12), (62), (34), (26), and
(112) combinatoriality”—​but despite this, the array still exemplifies regular
partitioning and a texture in which harmonic rhythm is consistent at each
distinct level of aggregate formation (e.g., each line completes a series form
in twelve aggregates). The second half of the piece switches to trichordally
derived series but continues to be based on regular 12 x 12 arrays, with only
the occasional instance of swapping (Hush 1982–​83, 170–​77).
In All Set, Babbitt begins to work with increasingly complex arrays with more
than four lines. Indeed, as mentioned previously, All Set is a transitional piece
in many respects. Figure 3.10a gives a remarkable instance (several others can

21 In systematically stretching trichords across aggregate boundaries, this is not unlike the

interludes of Du (1951). See Mead 1994, 111.

22 On Composition for Twelve Instruments, see Babbitt (1961) 2003, 94; Hush 1982–​ 83; and
Westergaard 1965.
 The Seam in Babbitt’s Compositional Development 87

Figure 3.10a. Array for All Set, mm. 89–​100

Figure 3.10b. Extracted and un-​swapped wind and brass parts of Figure 3.10a,
showing the underlying trichordal array

be found in Mead 1994, 118–​23). The winds and brass would, by themselves,
form a trichordal array, as shown in Figure 3.10b, but this array has been ex-
panded to make room for elements from the vibraphone’s lines. The use of a
complete array expanded to include other lines, and the resulting asymmetrical
partitioning, adumbrate characteristics of CT6.
While much of All Set uses rather conventional trichordal arrays, in sev-
eral of its sections—​particularly those for a relatively large sub-​ensemble—​it
establishes itself as the most contrapuntally intricate piece Babbitt wrote in
the 1950s.23 It incorporates each of the techniques discussed thus far, with the
large ensemble and, perhaps, a desire to create melodic licks and rhythmic ir-
regularity befitting its jazzy surface, leading Babbitt toward techniques that
went well beyond the trichordal array.
The next three years would be silent, in terms of published output, as Babbitt
wrestled with the synthesizer and thought himself most of the way toward a new
compositional period. CT6, the first result of that rethinking, would push every
contrapuntal technique even further than All Set.

23 As discussed in Mead 1994, 119–​20, the piece proceeds through the sixty-​three combinations of

its six melodic instruments. For more on the piece, see Babbitt 1987, 114–​17.
88 Thinking In and About Music

The Arrays of Composition for Tenor and Six Instruments

The series class for CT6 is unambiguously <01576e8924t3>: the ten trichord
orderings used in the piece’s arrays can only be combined into this class, the pat-
tern of sectional durations discussed previously clarifies it, and the final section,
mostly composed of forms of the series class, confirms it. The determination of
the particular series generating CT6 is somewhat less clear, but the series form
projected by the tenor in mm. 457–​59 is a good candidate. Not only is it the sole
series form articulated by the tenor soloist, it projects the sequence of intervals
realized by the durations of the first eleven sections, begins with the trichord
ordering that generates the trichordal array of the first section of the piece, and
includes set content, particularly its D-​type hexachords, that occurs frequently
throughout the piece. This series, and the trichord orderings that generate most
of the piece’s arrays, is shown in Figure 3.11.
CT6 develops the trichord orderings of Figure 3.11 using a variety of
techniques unknown to Babbitt’s earlier practice. Although the various
sections of the piece are sharply differentiated in a number of respects, as
Babbitt claims in the first quotation cited in this chapter, most of CT6’s
innovations are introduced in its first five sections. Accordingly, I will begin
with an examination of those opening sections.
Figure 3.12a presents the array for Section I of CT6.24 At a level we might
call the “foreground,” to distinguish it from slower processes behind it,
Section I realizes a mostly standard trichordal array generated from <−1,
−4>, the first trichord ordering of the series. This foreground array is clari-
fied in Figure 3.12b. The array proceeds, typically, through the fifteen com-
binations of four trichords arranged into complementary pairs and a tutti.
Each instrumental family—​the strings, the winds, the harpsichord, and the
tenor—​presents one of these trichords in each aggregate.
In addition to the trichordal array, each instrumental line in the winds and
strings is separately generated. It is in this sense that the array is generatively

24 Certain things appear ordered in Figures 3.12–​3.21 but are realized as simultaneities. I have

interpreted their ordering so that the governing trichord orderings, unambiguous whenever the
notes are presented sequentially, are clarified. Some of these simultaneities should be interpreted as
ordered from bottom to top (e.g., {C♯, F♯} in the cello in m. 27), while others should be read from top to
bottom (e.g., {D, E♭} in the viola in m. 25; see Figures 3.13 and 3.14). After Section XII, Babbitt begins
to consistently present every simultaneity such that it should be considered ordered from bottom to
top, a practice he would continue almost without exception for the rest of his career (Philomel [1964]
including the most prominent, and perhaps the final, counterexamples; see Chapter 5). This stand-
ardization of harmonic interpretation is yet another example of accommodating increasing contra-
puntal complexity through the simplification of other practices.
 The Seam in Babbitt’s Compositional Development 89

Figure 3.11. Series for Composition for Tenor and Six Instruments, as presented
in the tenor in mm. 457–​59, with trichord orderings and hexachords indicated

polyvocal. Each of the wind instruments separately completes an aggregate

generated from (015). Three out of four of the trichords in each of the winds’
instrumental lines express <+1, −5>, the fourth segmental (or second dis-
crete) trichord of the piece’s series; the exception is the third trichord for
both the flute and the oboe. Those trichords, in the fifth and sixth aggregates
respectively, use the ordering <−1, −4>, the ordering developed by the
trichordal array. In those aggregates, the coincidence of the two generating
schemes (the total ensemble generation from <−1, −4> and the separate
wind instruments’ generation from <+1, −5>) means that the (015) the winds
should contribute to the foreground trichordal array must come from first
the oboe alone, in the fifth aggregate, and then the flute alone, in the sixth ag-
gregate. This brings the two orderings of (015) into conflict: either the instru-
mental generation or the generation of the foreground trichordal array will
be violated. Apparently the trichordal array takes precedence.
The strings present the first two aggregates of a tetrachordal array generated
from two orderings of (0167).25 The set class (0167) is not itself a subset of the
piece’s series, making its appearance an unusual occurrence in Babbitt’s pieces
built with derived series. However, the two orderings of (0167)—​< −5, −1, −5>,
used in the viola and cello, and <+5, +6, −5>, used in the violin—​each concat-
enate two imbricated statements of trichord orderings that are found in the se-
ries, as the respective seventh and tenth segmental (or third and fourth discrete)
trichords in the series. The violin’s first tetrachord, <G, C, F♯, C♯>, for instance,
concatenates <G, C, F♯> and <C, F♯, C♯>, each of which represents trichord

25 This tetrachordal array bears comparison with the trichordal array in mm. 89–​100 of All Set,

depicted in Figure 3.10. As in the earlier example, the tetrachordal array is embedded within a larger
array, its elements spread out to accommodate additional lines.
Figure 3.12a. Array for Section I of Composition for Tenor and Six Instruments, mm. 1–​15. Vertical dashed lines indicate
the partitioning of the foreground trichordal array. Instrumental families are arranged in order to demonstrate hexachordal
aggregates; groups of instruments that form such column-​spanning aggregates are separated by horizontal dashed lines.

Figure 3.12b. Clarification of foreground trichordal array in Section I of Composition for Tenor and Six
Instruments
 The Seam in Babbitt’s Compositional Development 91

ordering <+6, −5>.26 In sum, the surface trichordal array presents the first dis-
crete trichord of the series, the individual wind lines present its second discrete
trichord, and the strings’ tetrachordal array presents its third and fourth discrete
trichords. In introductory fashion, Section I thus adumbrates the entire series.
A notable feature of this passage—​characteristic of CT6 but in sharp
contrast to Babbitt’s earlier practice and, for the most part, to his later
practice—​is the vastly different rates of aggregate completion in the different
instrumental lines. The tenor and harpsichord complete their aggregates
every four ensemble aggregates. The winds unfold their aggregates at half
that speed, completing them only at the end of the section. The strings are
slower still: they finish the section having completed only two tetrachords
and will complete their aggregates in the next section.
In Section II, the tenor and harpsichord continue to be generated by
<−1, −4>, the first trichord of the series, unfolding only a single aggregate
over the course of the section; see Figure 3.13. The strings complete the
tetrachordal array begun in Section I and then complete a new tetrachordal
array, presenting the same tetrachordal contents as the first tetrachordal
array in the same order, although each tetrachord has been inverted into it-
self. The winds have a new strategy: both the flute and oboe are divided into
two registers, and both present the same four trichords in an invertible coun-
terpoint pattern featuring D-​type hexachords in both instruments.27 These
registral lines in the winds do not complete linear aggregates. They present
only six notes each in Section II and do not continue into the next section.
The lines formed by these various generative strategies do not—​indeed,
cannot—​constitute a trichordal array as the various lines of Section I do. The
partitioning of each aggregate is uneven and irregular—​genuinely so, not as
the result of swapping. Furthermore, these separate lines cannot be neatly
partitioned into aggregates. In order to create aggregates with these lines,
pitch classes are repeated across aggregate boundaries.28 For instance, in the

26 The transformation relating the imbricated trichords is what Lewin (1987) 2011, 180–​81 calls

27 Three of these four trichords, {F, B♭, B}, {C♯, D♯, G}, and {C, E, F♯} are presented in the same
“RICH,” for RI-​chaining operation.

order in each instrument. It is not clear why {G♯, A, D} does not follow suit; in particular, it seems
that D and A could be swapped in the oboe part in m. 21. This would group the four trichords into
two inversionally related pairs (as is very common throughout the piece and elsewhere in Babbitt’s
trichordal practice).
28 I believe this is the first instance of this technique—​ which Morris 1993 calls “horizontal
weighting”—​in Babbitt’s practice. It would eventually become a standard (indeed, necessary) part
of all-​partition array construction, as discussed in Babbitt (1974) 2003, 317. One difference between
CT6 and Babbitt’s later practice is that in CT6, certain pitch classes repeated across aggregate bound-
aries are transposed by an octave; afterward, they are generally repeated at pitch.
Figure 3.13. Array for Section II of Composition for Tenor and Six Instruments, mm. 16–​29. Columns
represent completions of the chromatic gamut, as in all other arrays in this volume, but the final three
columns include duplications.
 The Seam in Babbitt’s Compositional Development 93

second, third, and fourth aggregates, only the upper register of the oboe and
the cello contain C. In order for C to be represented in all three aggregates, one
of the instruments must repeat that pitch class across an aggregate boundary,
as the cello does between the second and third aggregate. The result of these
repetitions is that even though the various lines present eight aggregates’
worth of material, by the end of the section there are more remaining pitch
classes than can be sorted into aggregates. Therefore, the last three columns
of Figure 3.13 contain more than twelve pitch classes. They are aggregates
in the sense of chromatic completion (every pitch class is represented) but
not precise exhaustion (some pitch classes are duplicated). As can be seen in
Figure 3.14, which reproduces the final three “aggregates” of Section II, these
duplications are generally presented as unison simultaneities—​most often as
simultaneous attacks. It is unclear how this fact should be interpreted. One
the one hand, presenting these duplications as simultaneous unisons has the
effect of condensing them. From the perspective of the ensemble, there are
still basically twelve pitch-​class events in each of the final three aggregates.
On the other hand, unison simultaneities (except those generated by harp-
sichord coupling29) are relatively unusual in CT6 and, indeed, in Babbitt’s
output generally up to this point, and thus attract special attention.30
As shown in Figure 3.15a, every line in Section III is trichordally derived,
presenting six distinct trichord orderings. The viola, tenor, and harpsichord
use the only two directly adjacent segmental trichords in the series that can
be combined to form D-​type hexachords, and indeed the hexachords formed
here, within and between instruments, are the specific D-​type collections
found in the primary series form of the piece. The registrally separate lines in
the winds, violin, and cello present the first six notes of a trichordally gener-
ated aggregate that will continue through Section IV and will be completed
in the fourth aggregate of Section V.

29 Babbitt is permissive on this point, directing simply: “Harpsichord registration is to be chosen

in order to achieve the designated dynamics and durations as accurately as possible by means of the
particular instrument employed.”
30 This type of “aggregate” (if that term is still applicable) is resonant with a general interest of

Babbitt’s throughout the 1960s in reintroducing octaves and unisons into his compositions (although
earlier examples do exist, most strikingly in String Quartet No. 2). Nearly all of his pieces from that
decade include octaves or unisons, sometimes as anomalous one-​off duplications of serially expected
notes and sometimes within a consistent, orchestrationally doubled projection of a contrapuntal
line (as discussed in Babbitt [1970] 2003, 244–​54, as well as subsequently: the final section of CT6
contains several complete unison or octave duplications of series statements). This interest would
eventually be fully incorporated into his compositional syntax in the form of weighted aggregates
(beginning with String Quartet No. 4 [1970]) and superarrays (beginning with Reflections [1975]).
On weighted aggregates, see Babbitt (1974) 2003, 318–​20.
Figure 3.14. Composition for Tenor and Six Instruments, mm. 25–​29.
Duplicated pitch classes are boxed.
Figure 3.15a. Array for Section III of Composition for Tenor and Six Instruments, mm. 30–​44
96 Thinking In and About Music

Figure 3.15b. Mosaic content in array for Section III of Composition for Tenor
and Six Instruments. Underlines indicate pcs exchanged between otherwise
identical mosaics.

These trichords are partitioned into dyads and tetrachords such that every
aggregate and, over the course of the section, every instrument present pairs
of dyads and tetrachords (except for the harpsichord, which is represented
twice). One result of this partitioning is an increase in invariance: despite the
wide variety of trichords used, there are basically only two distinct mosaics of
dyads and tetrachords, as shown in Figure 3.15b.31
Figure 3.16 shows the pitch-​class array for Section IV (the music in ques-
tion is reproduced in Figure 3.6). As mentioned previously, Section IV has
only four aggregates. The winds, violin, and cello continue with the same
trichordal content they had used in Section III, presenting the third trichord
of linear aggregates that begin in Section III and are completed in Section
V. In this section, there are a great number of pitch classes held across aggre-
gate boundaries. This technique is familiar from Section II, but the apparent
justification for these repetitions is different from what it had been in the
earlier section. Each of the twelve lines contains three pitch classes, but not
every pitch class is represented exactly three times (for instance, F is found
only in the low registers of the flute and violin, while A is found four times, in
the high register of the oboe and harpsichord and the low register of the cello
and viola). In order to stretch this material across four aggregates, there must
be twelve repetitions, weighted toward the underrepresented notes. There
are, however, thirteen repetitions, as well as an anomalously inserted C♯ in
the final aggregate in the violin, resulting in two duplications (G and A♭) in
the final column. As in Section II, the duplications are presented as simulta-
neous unisons. But unlike in Section II, it is easy to imagine a recomposition
that would not result in G and A♭ being duplicated in the final aggregate. If
the harpsichord did not repeat its A♭, then either the cello or oboe A♭ could
be shifted back an aggregate, and if the oboe did not repeat its G in the third

31 More specifically, there are four distinct mosaics, but those four can be divided into two pairs

that are identical except for a single pair of pitch classes (this is shown in Figure 3.15b). The strategy
used in this section resembles that of Variation III of Semi-​Simple Variations (see Figure 3.8a)—​in
both sections, irregular partitioning results in increased invariance.
 The Seam in Babbitt’s Compositional Development 97

Figure 3.16. Array for Section IV of Composition for Tenor and Six Instruments,
mm. 44–​56. The parentheses indicate a pitch class inserted into the section’s
trichordal scheme.

aggregate, then either the violin or viola G could be shifted back an aggre-
gate. The inserted C♯ completes the final aggregate, but it is also easy to im-
agine a recomposition that would avoid the insertion: it could be removed if
the C♯ in the viola were repeated. The purpose of the duplications and inser-
tion is mysterious; they seem to represent an uncharacteristic disorderliness
in Babbitt’s array construction.
As can be seen in Figure 3.17, Section V presents trichordal lines in
thirteen-​part counterpoint, developing the same six trichord orderings
presented in Section III. This is the high-​water mark for contrapuntal com-
plexity in this piece and in Babbitt’s output up to this point. Furthermore,
we see here a limiting case for another innovative aspect of the piece: its
diverse and sometimes ambiguous rates of linear aggregate completion.
The registrally discrete lines in the winds, violin, and cello present the final
trichord of the linear aggregates they began in Section III before going on
to present an additional trichord.32 These additional trichords do not con-
tinue to create further linear aggregates, but do mean that the flute, oboe,
violin, and cello parts as a whole present aggregates within Section V. The
viola, contrastingly, completes the linear aggregates it had begun in Section
IV. The harpsichord does not complete any linear aggregate but does com-
plete two hexachordal aggregates: of the three trichords it presents in

32 The cello’s high A♭ and E♭ in m. 65 appear to be reversed (whether intentionally or not): the

resulting <+5, +5> trichord ordering is not contained within the piece’s series.
98 Thinking In and About Music

Figure 3.17. Array for Section V of Composition for Tenor and Six Instruments,
mm. 57–​70

each of its registers, the first create aggregates in combination with the
harpsichord’s trichords in Section IV, while the second and third combine
to complete a new aggregate. The tenor presents ten notes of an aggregate
derived from the final two segmental trichords of the series, arranged into
D-​type hexachords (as in Section III); this aggregate would be completed
by E♭ and A. A was the last note the tenor sang before Section V (back in
Section III), and E♭ will be the first new note he sings in Section VI, after
repeating B♭, but viewing the tenor’s material as an aggregate beginning
with that distant A and finishing with the succeeding E♭ would obscure its
trichordal and hexachordal derivation. To wit, <G♭, A♭, D>, the first three
tenor notes of Section V, create a familiar array trichord, <−2, +6>, while
a trichord incorporating the A at the end of Section III (<A, G♭, A♭>) does
not. It is therefore unclear how to interpret the tenor’s incomplete aggre-
gate. Such complications regarding the spans within which the various
lines complete aggregates are a distinctive feature of this piece. It may well
be one of the aspects that Babbitt found problematic. There is no other
Babbitt piece in which aggregate structure is this opaque and creates such a
tangle of contradictory grouping patterns.
Sections I through V demonstrate many of the innovative complexities
found in CT6’s array construction. For the most part, the remainder of the
piece develops various segments of the series using the techniques intro-
duced in these sections. Figures 3.18, 3.19, and 3.20, however, present three
 The Seam in Babbitt’s Compositional Development 99

additional array strategies—​simpler than those in the first five sections, yet
still distinctive—​found later in the piece.
Section VIII, as shown in Figure 3.18, presents a typical trichordal
array on the surface, progressing, as is usual, through the fifteen combin-
ations of four trichords. However, the trichords of this array are distrib-
uted such that each instrument completes exactly one aggregate—​except
for the tenor, which completes two (presenting two lines undifferenti-
ated in their projection). Figure 3.19 depicts the array for the first section
of the interlude—​the three sections in the middle of the piece that do
not play a role in the large-​scale durational scheme discussed previously.
Both of the outer sections of the interlude, uniquely in the piece, pre-
sent twelve aggregates. This array develops an unusual property of the
series; namely, that its discrete dyads present only three interval classes,
in the sequence <ic1, ic2, ic5> in both hexachords. Each line presents
one of these intervals in each aggregate. Finally, Figure 3.20 presents the
array for Section XXII, the second four-​second-​long section, which like
Section IV uses only four aggregates. This array presents four standard
trichordal lines. Two of these lines are isolated registrally in the harp-
sichord, one is divided between the violin and cello, and one is shared
between the flute, oboe, viola, and tenor. This shared line presents yet
another, especially diffuse, instance of generative polyvocality: the four
instruments together present a single line generated from <−2, +1>,
while they each individually also present <−5, −2>.

Figure 3.18. Array for Section VIII of Composition for Tenor and Six
Instruments, mm. 102–​15
Figure 3.19. Array for Section XII of Composition for Tenor and Six Instruments, mm. 159–​86. Brackets
indicate a serially expected note that is absent.
 The Seam in Babbitt’s Compositional Development 101

Figure 3.20. Array for Section XXII of Composition for Tenor and Six
Instruments, mm. 393–​96

“Invariants of Order Embedded in Invariants of Content”

In the first quotation cited in this chapter, Babbitt remarks on the complexity
of certain aspects of the piece, especially that the instrumental lines are “so
rarely generatively univocal” and that the “sections are so sharply delineated
in so many respects.” The foregoing discussion should have amply demon-
strated both of these points. From this, Babbitt evidently felt that he had
sacrificed “convincing continuity,” which, he claims, “depends crucially on
invariants of order embedded in invariants of content.” From this point for-
ward, the lines of Babbitt’s arrays would typically be forms of a single series—​
“invariants of order”—​proceeding, via secondary sets, through series forms
with identical hexachords—​“invariants of content.” In order to accommo-
date the increase in contrapuntal lines and partitional variety, he apparently
felt that he had to simplify each line and make the relationships between the
lines more straightforward. Babbitt would work with increasingly elabo-
rate contrapuntal structures in the years to come as he expanded into all-​
partition arrays, weighted aggregates, and superarrays, but the lines of these
arrays would remain simple. Even when he returned to derived series some
seventeen years hence, they never appear in anything more complex than a
basic trichordal array, such as that shown in Figure 3.1a. The vortex of gener-
ative strategies found in CT6 has no successors in Babbitt’s output.
Babbitt’s break away from derived series begins, arguably, in the final sec-
tion of CT6, depicted in Figure 3.21. This section is not a coda according
to the durational scheme—​it extends for the expected forty-​four-​second-​
long duration—​but it is strikingly different in its pitch organization from
102 Thinking In and About Music

any section presented thus far. In the first quotation cited previously, Babbitt
compared this piece to his String Quartet No. 2, writing that both employ
“an explicit unfolding of the set content.” Like the earlier piece, the full last
section of CT6 is built nearly entirely from statements of the piece’s under-
lying series, conjoining series forms that all share identical hexachordal
content. As shown in Figure 3.21, these series forms are primarily found
in the strings and in two distinct registers of the harpsichord. Each of these
five lines presents four series forms, progressing through the four classical
transformations (S, I, R, and RI); these series forms are presented in an
array-​like pattern, cycling through the ten ways in which the four ordering
possibilities can be presented in solos or pairs.33 Since there is a mismatch
between the number of series required to complete these ten combinations
(sixteen) and the number of series required for each of the five lines to pre-
sent all four orderings (twenty), there are four duplications in which two
instruments present identical series simultaneously. Generally—​as with
the earlier duplications in the piece, such as those in Section II—​doubled
series forms are presented in tightly coordinated unisons or octaves or a
close approximation thereof. Although not participating in the array of se-
ries forms, the winds and tenor also occasionally enter to double notes found
within the array. Ultimately, the winds and tenor present derived series,
or what appear to be the beginnings of derived series, generated from the
two orderings of the series of (015), <−1, −4> and <+1, −5>, compiled into
the D-​type hexachords also found in the section’s series forms.34 The tenor
duplicates not just a trichord but a complete series form, S4. This is the only
series form duplicated outside the array of series forms, and the only series
form presented by the tenor. As has been discussed, for these reasons among
others, S4 may be taken as the series of the piece.
Figure 3.22 presents the last six measures of the piece, encompassing the
final two series statements and the final simultaneity. The tenor and oboe
duplicate the harpsichord: the tenor, S4; the oboe, six non-​consecutive pitch
classes, forming instances of <−1, −4> and <+1, −5>. After the completion of
the array of series forms, the full ensemble returns for a climactic final chord.

33 There is an exception to the first of these principles: the viola has no R form and the cello

presents two R forms. Apparently, the cello presents the viola’s R form (whether intentionally or not is
unclear).
34 Additional material is sometimes also projected by these duplications. For instance, the oboe

part in mm. 433–​34 is arranged registrally into instances of <−1, −4>, as shown in Figure 3.21, but
temporally into instances of <+2, −1>.
Figure 3.21. Array for Section XXV of Composition for Tenor and Six Instruments, mm. 433–​62. Columns represent
completed series forms; there is no consistent aggregate construction.
104 Thinking In and About Music

Figure 3.22. Composition for Tenor and Six Instruments, mm. 457–​62

This sonority complements the second hexachord of It, sustained simulta-

neously in the harpsichord; it is one of the complementary hexachords used
in series forms throughout the final section. It is mostly compiled by having
instruments repeat the last notes they played, as can be seen in the oboe
part in Figure 3.22; the only exception is the tenor, which sings F—​needed
to complete the hexachord—​rather than repeating its last note (which is
not a member of the hexachord). Emphasizing this restatement is the fact
that the dynamics for each instrument in the final sonority maintain the dy-
namic level with which those pitches were last sounded in the instrument—​
including the tenor, whose previous F, in m. 458, is also 𝆐𝆑.35
The simplicity of this final section, in comparison with the convolutions
of the rest of the piece, belies its significant implications. Babbitt’s
work over the next few years can be seen as a variety of attempts to

35 As discussed in Chapter 6, repeated notes are a characteristic closing technique for Babbitt.
 The Seam in Babbitt’s Compositional Development 105

Figure 3.22. (Continued)

synthesize the contrapuntal complexity developed in the preceding twenty-​

four sections with the linear cogency of this final section. The result was the
all-​partition array.
Section XXV of CT6 also foreshadows a substantial change in the hierar-
chical structure of Babbitt’s music henceforth. In an array with derived lines,
the basic model is that the lines are understood to be derived from an under-
lying series—​a specific ordering of six or twelve pitch classes that the content
and order of the derived material can be understood as reflecting. In an array
in which each line states an iteration of the series, such as that seen in Section
XXV, there is ordinarily no reason to suppose that any particular form of the
series is hierarchically prior. In these cases, often the most specific object that
can be regarded as the hierarchical source is a series class, not any particular
106 Thinking In and About Music

series form. The result is that even though later pieces present considerably
more direct statements of their series than can be found in trichordal-​array
pieces (some of which contain no statements of their series whatsoever), the
underlying source is more abstract: an equivalence class, not a series of pitch
classes.
This increased abstraction at the level of the hierarchical source would
eventually be balanced not only by simpler array lines but also by an increased
concreteness at the level of surface organization. The primary determinant
of surface organization in music throughout Babbitt’s career is that the sur-
face should be built from materials also found in the array. In pieces with
trichordal arrays, the result is generally that the surface will be filled with
the trichord orderings developed in the arrays (see Mead 1994, 54–​123). In
pieces with all-​partition arrays, on the other hand, particularly from 1982 on,
surface references are generally to specific partitions of the aggregate found
elsewhere in the array. That is, the practice with trichordal arrays is for the
surface to be composed of hierarchically significant intervallic patterns. In
later pieces with all-​partition arrays, references are to specific sequences of
pitch classes.
As Dubiel (1990b) has it, the invention of the all-​partition array marked
a “watershed” moment for Babbitt (222): although his compositional prac-
tice would continue to evolve, later developments might well be described as
“comparatively superficial” (225).36 The array’s strengths, in light of Babbitt’s
general systematic concerns, are clear. The all-​partition array does more than
incorporate the contrapuntal variety that was an increasingly important in-
terest of Babbitt’s, it codifies it. It extends the principle of exhaustive com-
pletion latent in the twelve-​tone series—​the basic principle undergirding
hierarchy in Babbitt’s music—​to the dimension of contrapuntal variety. By
integrating Babbitt’s hierarchical concerns with his interest in contrapuntal
variety, and in achieving a continuity he found convincing as a result of linear
redundancy, the all-​partition array became the foundation of the remaining
forty years of Babbitt’s compositional life.

36 However, Dubiel’s (1990b, 225) further characterization of Babbitt’s development, that “there

is an unmistakable tendency . . . to place the twelve-​tone set further and further in the background,”
strikes me as more problematic: the transition to using forms of the series as material for lines in the
array presents the series in a fashion much more directly available to perception than that which
it occupied in Babbitt’s trichordal pieces, in which the underlying series often has to be inferred.
Furthermore, the (121) partition that appears in most all-​partition arrays presents the series literally
and monophonically.
 4
The Surface and the Series in Composition
for Four Instruments

For all the sophisticated organicist, constructional, and cognitive sources of

his theorizing, the “first step” Babbitt proposes for one listening to a twelve-​
tone work is startlingly direct:

For the listener, the most efficient and revelatory first step towards
acquiring familiarity with and comprehension of this, or any other twelve-​
tone work, is that of identifying and examining that composition’s twelve-​
tone set, . . . which . . . supplies the total pitch material of the composition
and so endows the composition not merely with its local sonic character-
istics but with its structural properties at every stage and level of composi-
tional unfolding. (Babbitt [1968] 2003, 223)

A listener must identify and examine the series. And given Babbitt’s intel-
lectual heritage, one may presume the benefits such an exercise might have.
From a constructional perspective, the series of a work likely enables the
most effective relation descriptions: if the “total pitch material” can be related
to it, then the various events of the piece can be defined and understood as
its manifestations. From a cognitive perspective, a mental image of the se-
ries can enable efficient recoding of events that can be recognized as series
statements or subsets, and knowledge of its “structural properties” can en-
able comprehension of syntax. A listener can thus form expectations about
the sequence of events and relations they are likely to encounter. And from
an organicist perspective, a series provides access to a work’s imagined hier-
archical development as the source of its “compositional unfolding.”
Sometimes identifying the series is a straightforward matter. There are
pieces, such as Glosses, that begin with their series, and an attentive listener
may grasp how subsequent events are relatable to that which they have al-
ready heard (see Figure 1.1). In others, Babbitt sets up a “revelatory” nar-
rative, to use Mead’s (1994, 107) apt phrase for pieces that present their

Thinking In and About Music. Zachary Bernstein, Oxford University Press. © Oxford University Press 2021.
DOI: 10.1093/​oso/​9780190949235.003.0004
108 Thinking In and About Music

series only after first developing its subsets—​“telling you the butler did it,”
as Babbitt said about String Quartet No. 2 (quoted in Straus 1986, 22). Many
pieces with all-​partition arrays present a version of this narrative, revealing
their series, typically as a (121) partition, only after developing other, more
complex partitions. Series identification in such cases is a somewhat trickier
task, at least on first hearing. Babbitt’s description of the cognitive process
involved is “disambiguation” (quoted in Straus 1986, 22), which captures
the sense that the confusion of earlier passages, which may seem to present
material that hangs together only loosely, fades as that material is demon-
strated to be relatable to a single source.1 A listener must wait for the even-
tual reveal, but fortunately Babbitt often facilitates the process by presenting
the series in a marked manner: in all-​partition array pieces, for instance, the
(121) partition will typically be projected in a single register or instrument, a
noticeable deviation from the prevailing contrapuntal texture. But in other
pieces—​including Composition for Four Instruments, which is not only one
of Babbitt’s most well-​known pieces2 but is also representative of issues that
arise repeatedly in his work—​the underlying series is never directly stated. In
such works, there are derived series, but there is no revelation of the source
from which they are derived. The listener’s task thus becomes quite a bit more
difficult. An examination of Composition for Four Instruments will serve as a
case study for the manner in which series can be identified in such works and
more broadly, and more importantly, as a study of the ways in which a series
may function in relation to the surface of the music it can be said to underlie.
Babbitt’s comments on the series of Composition for Four Instruments in
1976 are revealing about both its compositional purpose and the process of
its apprehension:

As early as in my Composition for Four Instruments, I attempted to use the

set as an ever and multiply effective compositional determinant, if only cu-
mulatively inferable, in its effects, as an entity, not thereby to obscure or
conceal it cryptographically, but to create individual ambiguities of refer-
ence which only finally and dispositionally, in the course of the events of the
piece, become referentially unequivocal. I did not conceive of the first six
measures of the Composition for Four Instruments to be so initially credible

1 As noted in Chapter 2, in the construction of such narratives, whether deliberately or not, Babbitt

approaches Schoenberg’s concept of the “musical idea.”

2 Witness its prominent billing in, for instance, Dubiel 1992, 82–​91; Mead 1994, 55–​76; Perle 1991,

82; and Taruskin 2010, 143–​48.

 The Surface and the Series 109

as a “set statement” as to resist dispositional reassessment for more than a

matter of measures, for no longer than, say, the F-​major triad at the opening
of the Beethoven First Symphony, or the A-​flat-​major of the opening of the
Chopin G-​minor Ballade. (Babbitt [1976b] 2003, 348)

Like an off-​tonic opening of a tonal piece, the piece’s first aggregate is even-
tually shown to be a local elaboration. The true underlying series is only “cu-
mulatively inferable” over the “course of the events” as ambiguities gradually
fade. It can be taken as a “compositional determinant” by the observation
of “its effects.” Babbitt does not clarify what those effects might be, except
to promise they are multiple and, considered as a whole, render the series
unequivocal. The series may be “continuously, thoroughly, and utterly influ-
ential” (Babbitt 1987, 27), but how, and on what, it exerts its influence is left
unstated.
Inferences being what they are, it is perhaps not surprising that there is po-
tential for disagreement. Babbitt may not have intentionally hidden the series
“cryptographically,” but the many published series for the Composition for
Four Instruments attest to the challenge of identifying any single source of the
“total pitch material” in the piece.3 Yet despite these various interpretations,
Babbitt (1997, 133) later reiterates that the series for Composition for Four
Instruments is, by the end of the piece, unambiguous, writing that “the four
different trichords (different either in structure or order) are genidentically
related as the four discrete trichords of the emergent referential series, and
the harmony and polyphony of the aggregates which they singly and together
move and shape reflect the properties of that series and—​ultimately—​of only
that series, omnipresent and yet never present.” In this statement, Babbitt
reveals a further nugget: the four trichords developed in the piece’s arrays
are the discrete trichords of an underlying twelve-​tone series. But despite his
insistence that the piece defines a single series, the fact remains that until
Babbitt ([1976b] 2003, 349–​50) declared <B, E♭, C, B♭, D, D♭, G, F, E, A, F♯,

3 There are some nine distinct published series for the piece, in Arnold and Hair 1976, 159; Babbitt

(1976) 2003, 349–​50; Lester 1989, 228; Mead 1994, 67; Perle 1991, 82; and Taruskin 2010, 143. Most
of this unhappy saga is told in Dubiel 1990b, 252–​53n.29. As Dubiel recounts, Perle 1991, 82 and
Cone 1967, 37; 1976, 305 incorrectly claim the series is presented at the end of the piece; the claim
is repeated by Lester 1989, 228 and Taruskin 2010, 143, in what are essentially paraphrases of Perle
1991. Although the piece’s final array is built from all four of the work’s trichord orderings, they are
not presented in direct sequence. In addition to these published sources, an unnamed source Babbitt
(1987, 27; 1998a, 26–​27) mentioned several times apparently tried to take the first twelve notes of
the piece as its series. In general, the richest sources on Composition for Four Instruments are Dubiel
1992, 82–​91 and Mead 1994, 55–​76, both of which examine many aspects not discussed here.
110 Thinking In and About Music

G♯> to be the underlying series of the piece, even knowledgeable scholars dis-
agreed. The confusion that persists suggests that the question is not only one
of identity but of understanding how an underlying series might be clarified
within a piece in which it is never stated. At this point, the reader could be
forgiven a sense of despair. Why attempt to reduce the piece to a single un-
derlying series when the evidence of existing scholarship points to irresolv-
able ambiguity? As Dubiel (1997, 45) asks, “What is the special attraction of
telling people that there is a single theoretical entity back there pulling all the
strings?”—​particularly, one might add, when the choice of any single entity
seems arbitrary if not perverse?
I propose that the series for Composition for Four Instruments is indeed
unequivocal, that it is in fact exactly what Babbitt claims it is, that the process
of identification is analytically worthwhile, and moreover that the process is
revealing about Babbitt’s output as a whole. While there are indeed a number
of possibilities for how the piece’s trichords and hexachords may be com-
piled, there are surface aspects that sufficiently clarify their relationship, lim-
iting the choices such that a single series emerges as a most efficient cognitive
model—​or most explanatory organic source—​for the piece’s events. This is
a role distinct from that usually ascribed to surface aspects, a fact that will
occasion reflection on what it means to be the “surface.” Furthermore, there
are surface aspects that clearly conflict with the series, which is particularly
notable given that the series is dependent on the surface for its clarification.
This conflict draws attention to these aspects, facilitating the rhetorical func-
tion that seems to justify their inclusion.
The analytic assertion of declaring a succession of twelve pitch classes the
underlying series of a piece is meaningful in direct proportion to the amount
that can be learned about the various events of a piece by making such an
assertion. Adopting a constructive framework, a series to which more of a
piece refers can be taken as a more effective logical primitive. From the per-
spective of cognitive science, the ideal series is that which best facilitates hi-
erarchization by relating to more events and thus providing the means for
recoding those events. As has been noted, these approaches converge for
Babbitt. Logic and cognition support, for him, the same sort of music and
the same sort of analytical statements about music, and one may reduce their
conclusions into a single guiding analytical principle: the best series for a
piece is that which takes account of as much as possible. Organicism leads to
the same principle: the most plausible organic source is that whose offshoots
flower most widely.
 The Surface and the Series 111

In light of this principle, I assume that the extent to which an unstated se-
ries should be considered the source of a piece’s trichordal arrays is the extent
to which that series is an effective embodiment of a piece’s harmonic contents,
the sequence in which those contents appear in the piece, and the relationships
among those contents. Babbitt’s statements quoted previously are concordant
with this assumption. A series in Babbitt’s music is not only “omnipresent,” it
is a “multiply effective compositional determinant.” Accordingly, a series that
can be plausibly construed to be a compositional determinant of more aspects
of a piece is a stronger model for a piece than another series.
Figures 4.1 and 4.2 list all of the trichords and hexachords used in the
arrays of Composition for Four Instruments. The trichordal vocabulary is ex-
traordinarily limited: two set classes, both of which are represented by only
four distinct sets, arranged into two distinct orderings each. In accordance
with Babbitt’s statements, it can be assumed that the underlying series will
include one representative from each trichord ordering. This restriction, in
and of itself, does not decide the matter with much specificity: many dis-
tinct series have discrete trichords that present all four of the piece’s trichord
orderings. (This is probably the most important factor behind the many
published series for the work.) In contrast, the diverse trichordal contents of
more complex pieces, such as Composition for Tenor and Six Instruments and
Two Sonnets (see Chapters 3 and 5), limit the available options, since few se-
ries include all of those pieces’ trichords as segments.

Figure 4.1. Trichords used in the arrays of Composition for Four Instruments.
Each appears both as listed and under retrogression. Trichord orderings are
given in reference to the piece’s underlying series.

Figure 4.2. Hexachords used in the arrays of Composition for Four Instruments
112 Thinking In and About Music

The principle that the series should model the piece as closely as possible
suggests that the discrete hexachords of the series should be hexachords used
in the piece’s arrays. But of the six possibilities in Figure 4.2, which pair of
hexachords is the best candidate for membership in the series? The answer
is clear: although all three pairs of hexachords appear in various combin-
ations throughout the piece, the A-​type hexachords appear in every one of
the piece’s arrays, either linearly or harmonically (as is clear in Mead 1994,
66). Given this omnipresence, {E, F, F♯, G, G♯, A} and {A♯, B, C, C♯, D, D♯}
should be taken as the discrete hexachords of the piece’s series.
Moreover, those hexachords should be partitioned into trichords included
in Figure 4.1. One consequence of this, as a quick comparison of Figures 4.1
and 4.2 will reveal, is that the hexachords will be partitioned into two
trichords of the same set class. The series, that is, will start with two members
of (014) or two members of (013), arranged to reflect the four distinct tri-
chord orderings of Figure 4.1.
At this point, we have reached the state of knowledge embodied in Figure
4.3, reproduced from Mead (1994, 67). Mead recognizes that there are mul-
tiple ways to arrange the trichords and hexachords of the piece into a single
series and therefore lists four distinct “underlying rows” as a “catalog of the
ordered trichords of the piece, as well as of the principal hexachords” (65–​
67). There is much that is indisputably important in this diagram: it contains
all of the trichords used in the piece arranged into the omnipresent A-​type
hexachords. Nonetheless, ambiguities remain. That Mead lists multiple series
is indicative of the fact that, from the perspective of trichordal derivation, it
is unclear whether the series should start with (013) or (014). Furthermore,

Figure 4.3. The “underlying rows” of Composition for Four Instruments shown
in Mead 1994, 67
 The Surface and the Series 113

each trichord given in Figure 4.3 could be independently retrograded, and

the trichords within each hexachord could be swapped—​in any of the given
series—​without affecting the ability of each series to be considered the source
of the piece’s trichords and hexachords. Figure 4.3 is indeed effective as a “cat-
alog” of the piece’s trichords and hexachords, but it does not have the pre-
cision of an underlying series that can be taken as a model not just of the
trichordal and hexachordal content, but of the sequential development of the
piece’s arrays and of details of the piece’s surface.
Perhaps the most strikingly unusual aspect of the surface organiza-
tion of Composition for Four Instruments is its coordinated presentation of
simultaneities—​ not just simultaneous attacks but all instances of pitches
sounding together. In what amounts to a rigorously constrained contrapuntal
approach, each section of the piece, with the exception of the last—​“section”
here defined by changes in instrumental ensemble, as shown in Figure 4.6—​uses
no more than two interval classes between registrally adjacent sounding notes.4
Sometimes those interval classes sound simultaneously, creating a trichord and
perhaps a third interval class between the outer notes of that trichord. While
the pattern of simultaneities does not define an exhaustive list, no two sections
use exactly the same simultaneities, giving each a distinct harmonic quality.
(Sometimes, consecutive sections have no simultaneities in common, resulting
in a sharp harmonic break; in other cases, smoother transitions are effected by
the use of common intervals.) The piece’s entire vocabulary of simultaneities
is given in Figure 4.4. The highly constricted vocabulary of the first fourteen
sections is visible in the figure, as is the relative freedom of the final section. In
what may be taken as a rhetorical summation, the final section recapitulates
nearly all of the piece’s harmonic vocabulary, including every interval class.
According to the guiding principle, a series that includes more of a
piece’s simultaneities as segments may be considered a better model of the
piece than a series that contains fewer. Moreover, as discussed in previous
chapters, such a conception of the series fits Babbitt’s compositional prac-
tice: it is customary for simultaneities in his pieces to reflect those pieces’
series. It is actually impossible to construct a twelve-​tone series that contains
all of the simultaneities in Figure 4.4 as well as the trichords and hexachords
in Figures 4.1 and 4.2, but one can come very close. Permuting the pos-
sible trichords, hexachords, and trichord orderings reveals that one series
class, <041e32865t79>, contains every simultaneity, from the perspective of

4 A slightly imprecise statement of this principle is given in Dubiel 1992, 125n.15.

114 Thinking In and About Music

Figure 4.4. Simultaneities in Composition for Four Instruments. Interval classes

in black appear as registral adjacencies; interval classes in grey appear only
between outer notes of trichords. Set classes in boldface are those that cannot
be considered to be part of the piece’s series. Trichord orderings are given in
relation to the underlying series when possible.

set-​class content, except for (027) and (0157).5 The set class (027) is the fly
in the ointment in a number of respects—​it is also the only subset of (0157)
that cannot be included—​and we will be returning to it shortly. But including
every other simultaneity determines the series class. Indeed, there is a near-​
perfect match between the simultaneities and the segmental dyads and
trichords of the series. Each of the segmental dyads and trichords of the series
defines a set class that appears as a simultaneity somewhere in the piece, and
no other dyadic or trichordal set classes are used, except for (027). Moreover,
in addition to set-​class content, this series class encompasses almost every
trichord ordering used in the piece’s simultaneities. The only exceptions are

5 The set class (027) cannot be included without abandoning at least (06) and (026). The tritone is

a considerably more prominent part of the piece’s surface—​both its simultaneities and successions—​
than (027), and it is therefore preferable for the series to include it.
 The Surface and the Series 115

Figure 4.5. Partitioning of trichordal arrays used in Composition for Four

Instruments (retrograded after section IV). Dots denote trichords sounding in a
given instrumentally or registrally defined line.

Figure 4.6. Global distribution of instrumental ensembles used in Composition

for Four Instruments. Dots denote instruments playing in a given ensemble. This
is equivalent to Figure 4.5.

Figure 4.7. Global distribution of trichordal content used in Composition

for Four Instruments. Dots denote trichord orderings used within each array.
Shaded columns deviate from the pattern of Figures 4.5 and 4.6.

the ordering of (016) that appears in section III and two trichord orderings—​
one of (016) and one of (026)—​in the final quartet.
Of the possible members of the series class <041e32865t79>, some begin
with (014)s as their first discrete trichords, and some begin with two (013)s.
While a series beginning with either trichord would still be an adequate source
of the piece’s trichords, hexachords, and simultaneities, an examination of the
distribution of trichordal content within the piece provides disambiguation.
The pattern of trichords contained within the opening trichordal array (Figure
4.5), progressing through the fifteen combinations of four trichords arranged
116 Thinking In and About Music

Figure 4.8. Possible series for Composition for Four Instruments

into complementary pairs and a final tutti, is replicated by the pattern of instru-
mental combinations that extends across the entire piece (Figure 4.6). This same
pattern is also—​nearly—​replicated by the pattern of trichord orderings used in
each of the piece’s coextensive arrays (Figure 4.7).6 The differences are subtle,
but they have important implications. Notably, from the perspective of series
disambiguation, section VI begins with the two (014) trichord orderings rather
than ending with them, as it would have had it followed the patterns set out in
Figures 4.4 and 4.5. The result is that every complementary pair of arrays begins
with at least two lines articulating (014)—​including, prominently, the first pair,
which begins with (014) alone. Ten out of fourteen trichord orderings that ap-
pear in arrays at the beginning of a complementary pair are members of (014).
Conversely, every complementary pair ends with (013), and (013) is generally
more weighted toward the end of these pairs. Accordingly, a series that begins
with (014) is a better model of the piece than a series that begins with (013).
At this point, we have narrowed the possible series to those listed in Figure
4.8. These are the only members of series class <041e32865t79> that in-
clude the trichords and A-​type hexachords of Figures 4.1 and 4.2 and begin
with (014)s. Of these options, one has particularly strong surface support.
As Dubiel (1990b, 253.n29) has noted, the discrete trichords of the series
marked “a)” in Figure 4.8 present the first three trichords played by each in-
strument, in the order of initial instrumental appearance (see Figure 4.9).
Given this support, <B, E♭, C, B♭, D, D♭, G, F, E, A, F♯, G♯> is confirmed as the
best candidate for the series of Composition for Four Instruments.
The introduction of the series as the first three notes in each instrument
appears to contradict Babbitt’s claims that it only becomes clear over the
course of the piece. But as this discussion has shown, it has taken a consider-
ation of virtually the entire piece to narrow the choices down to the options
in Figure 4.8. Choosing the single series among these four that happens to

6 Dubiel 1992, 125n.15 notes the correspondence between these partitions without noting its

imperfection.
 The Surface and the Series 117

take particularly good account of the opening of the piece is not equivalent
to basing the entire determination of the series on just the four initiating
trichords. The series is principally defined by the varied evidence offered by
array contents, simultaneities, and the arrangement of trichords, and not
simply by the piece’s opening. That is to say, to turn the analogy in Babbitt
([1976b] 2003) on its head, hearing the first sound of Beethoven’s Second
Symphony and immediately declaring it to be the tonic is unwarranted,

Figure 4.9. Composition for Four Instruments, mm. 1–​59, with (027)s marked.
Score in C.
118 Thinking In and About Music

Figure 4.9. (Continued)

 The Surface and the Series 119

Figure 4.9. (Continued)

even though it happens, eventually, to be correct. Confirmation—​referential

unequivocality—​is gradual; it is not conclusive until the end of the piece.
Indeed, there are simultaneities introduced in the last two sections of
Composition for Four Instruments that are essential to the disambiguation
of its series. The clue offered by the opening trichords of each instrument is
helpfully indicative, but confirmation takes place over the piece as a whole.
In organicist accounts it is common, as evidenced in various statements by
Babbitt cited previously, to describe twelve-​tone works as outgrowths of their se-
ries. This may well be a useful mindset for twelve-​tone composers to adopt during
at least part of their compositional process. But unlike composers, listeners are
120 Thinking In and About Music

(presumably) not given access to a twelve-​tone series in advance of their experience

with a piece. Indeed, even the characterization of a work as twelve-​tone is a matter
of individual construal. If listeners find that a work lends itself to such construal, it
will be because they find it easiest or most productive to comprehend the work in
relation to an underlying series, and their process in making such a determination
may little resemble a step-​by-​step ascent of the composer’s logical hierarchy. To a
composer, or an analyst representing a work with an organicist or constructional
model, a series may represent a developing kernel or a logical axiom. But for a lis-
tener, an underlying series is necessarily an emergent phenomenon, defined and
conditioned by the data experienced during the course of listening. Even if at some
point a listener adopts an organicist mindset as an explanatory heuristic, this will
not ultimately change the emergent nature of serial perception.
This fact is sometimes elided in organicist rhetoric. For instance, it is typ-
ical to explain surface details as “references” to array counterpoint (a usage
I admittedly have adopted). The implication is that surface details are de-
pendent on the array, and that although they may direct a listener’s ears
toward particular aspects of the array—​predicting or recalling events else-
where in the piece, for instance—​they are nonetheless in a subordinate hi-
erarchical position. In this fiction, the series and the array came first. But
listeners attempting to find a model that will help them efficiently compre-
hend the events of a piece are grabbing onto any piece of evidence available
to them, and the segments of the series constituting the surface of Babbitt’s
music are likely to be the aid most readily at hand. From this perspective, sur-
face elements are not simply references to the series. They define the series. In
a sense, they constitute it. The series has no existence apart from them.7
The heightened status surface elements acquire by this new role in se-
ries disambiguation is reminiscent of Richard Cohn’s (1992) discussion of
motives in Schenkerian theory. Despite Schenker’s insistence that motives
arise as a result of organic development, Cohn describes how their role in
determining hierarchy in many Schenkerian analyses grants motivic associ-
ation an independence and potential hierarchical significance at odds with
the image Schenkerian theory presents of a succession of prolongations
elaborating the Ursatz. As with motives and Schenkerian hierarchy (in
Cohn’s reading, at least), the foregoing discussion of Composition for Four
Instruments calls into question the determining-​determined binary puta-
tively governing array hierarchy and the surface. If the series in Composition

7 Dubiel 1997, 39 questions the concept of “references” in Babbitt’s music from a somewhat dif-

ferent perspective. An important claim in his discussion is that the interpretation of surface figures as
references is a choice; it is not compelled by the technique.
 The Surface and the Series 121

for Four Instruments is—​partially—​determined by the piece’s simultaneities,

then simultaneities have an autonomy apart from the series.
But in turn, surface events can acquire functionality in relation to serial
disambiguation. Listeners seeking out the series come to distinguish between
events that disambiguate the series and those that do not. A practical conse-
quence of this realization is the light into which it thrusts events that con-
trast with the series, for they confound listeners attempting to comprehend
the work on the basis of those events. In Composition for Four Instruments,
the most salient problematic set class is (027), the simultaneity that cannot
be considered part of this referential network of series disambiguation. In
the terms of Babbitt’s ([1976b] 2003, 343) comments on Mozart discussed in
Chapter 2, (027) is a “primitive” requiring “special comprehension.” Listeners
must hold it, in addition to the series, in mind, in order to comprehend the
piece in Babbitt’s terms. But this loss of efficiency is compensated for: that
(027) is brought into relief by its position apart from the piece’s hierarchy
contributes to its rhetorical effect. The fact that it is cognitively problematic
makes it rhetorically productive.8 The difficulty organicism would have in
justifying the sonority—​it cannot reasonably be said to be an outgrowth of
any aspect of the series—​contributes to the same effect.
Even before any simultaneity has occurred, Babbitt draws attention to
(027). As can be seen in Figure 4.9, the trichord jumps out of the first aggre-
gate in the opening clarinet solo, marked by loud dynamics and a large leap
crossing the clarinet’s break that precedes its first note. Following this dra-
matic intrusion, (027) underlies the dynamic structure of the second aggre-
gate: as shown in Chapter 3 in Figure 3.7d, the second aggregate features four
dynamically discrete levels, each of which realizes a member of (027).9 As the
rest of Figure 4.9 shows, (027) maintains prominence throughout the opening
clarinet solo and the ensuing trio, both in simultaneities and successions.
Following the end of this opening section, (027) essentially disappears for
most of the rest of the piece. For over three hundred measures, it never appears
as a simultaneity and can hardly be found as a plausible grouping of adjacen-
cies. But at the end of the piece, shown in Figure 4.10, it once again features
prominently, including twice in succession in m. 369, at the beginning of the
final section. The last simultaneity in the piece—​{A♭, C♯}—​is composed of the

8 Such rhetorical intrusions will be taken up at greater length in Chapters 6 and 7.

9 Dynamics in Composition for Four Instruments are often used to articulate trichord orderings
other than those used in the arrays, although (027) is the most salient dynamically projected set class
that is not found in the work’s series. Some aspects of the work’s dynamic organization are discussed
in Lester 1989, 269–​71. I am grateful to Daniel Colson for discussion on this subject.
122 Thinking In and About Music

Figure 4.10. Composition for Four Instruments, mm. 368–405, with (027)s
marked. Score in C.
 The Surface and the Series 123

Figure 4.10. (Continued)

124 Thinking In and About Music

Figure 4.10. (Continued)

two pitch classes shared between the (027) heard in the second measure and
the first (027) in m. 369.
The trichord (027) is thus used to link the beginning and ending of the
piece. As will be discussed in Chapter 6, such circularity is a technique
Babbitt used frequently to signal closure. The fact that (027) is hierarchically
problematic contributes to the effectiveness of this device. Put simply, (027)
sounds distinct from the piece’s harmonic language. Its pair of ic5s—​open,
consonant, and resonant in the wide spacings found in Composition for Four
Instruments—​contrast sharply with the dissonant (013) and (014) trichords
comprising the piece’s arrays, neither of which contains an ic5. The unique
harmonic quality of (027) is encapsulated in its role as the trichordal proto-
type for diatonicism, in Ian Quinn’s (2007, 28) terms. Indeed, (027) is dis-
tant from the other trichordal simultaneities used in the piece according to a
number of published measures for modeling set-​class similarity. It is, for in-
stance, the only trichordal simultaneity in the piece appearing in the bottom
two ranks of Joseph N. Straus’s (2003, 337) trichordal voice-​leading space,
which measures the number of voice-​leading shifts required to transform
one trichord into another, and is at the opposite end of the space from the
omnipresent (013) and (014). Its reprise at the end is thus recognizable as a
return to something not heard throughout the bulk of the piece. Shortly after
(027) reappears, an attenuation of the piece’s rhythmic momentum—​the re-
sult of a rhythmic coda, as will be discussed in Chapter 6—​coincides with the
change in harmonic quality. The work is brought to a gentle close.
 6
Completeness and Temporality

Introduction

In numerous remarks throughout the final decades of his life, Babbitt re-
peated a statement that described what he felt were not only fundamental
goals of his compositional techniques but also basic requirements of mu-
sical comprehension. Although this statement varies somewhat in its many
manifestations, the core concepts appear both fully and with unusual conci-
sion in one of its first appearances, from the program notes to an early per-
formance of Dual:

It is just the progression from the local to the global in relational implications
which should provide the listener with the means of achieving that cogni-
tion of cumulative containment and successive subsumption which human
memory in general, and musical memory in particular, requires for a mu-
sical work to be entified, eventually, as a unified, closed totality—​as an all of
a piece of music.1

Babbitt’s statement indicates that he thinks processes and limitations

of human cognition determine the requirements for a complete piece.
Cognition, he feels, requires a nested hierarchy in which each level is
“contained” or “subsumed” within a deeper level of structure. It is through
these means alone that musical memory can “entify” a musical work—​that
is, conceive of it as an objective whole. These cognitive demands have com-
positional consequences: hierarchy must be created by progressing outward
“from the local to the global,” generating a sequence of nested levels uni-
fied by “relational implications.” A later program note for Dual clarifies and

1 Quoted in Sandow (1982) 2004, 254. The occasion for which these program notes were written

was a February 22, 1982, performance by the Group for Contemporary Music. Further iterations of
this statement can be found in Babbitt (1984) 2003, 386; 1987, 144–​45; 1991a, 129; and (1991b) 2003,
441 and 446. Although the specific formulation of the ideas presented here seems to have originated
around 1980, it has numerous precedents, including Babbitt (1952) 2003, 25; (1961) 2003, 84; (1970)
2003, 254; and (1979a) 2003, 372.

Thinking In and About Music. Zachary Bernstein, Oxford University Press. © Oxford University Press 2021.
DOI: 10.1093/​oso/​9780190949235.003.0006
 Completeness and Temporality 197

intensifies this statement: “relational implications” is replaced by “relational

replications,” indicating the extent to which Babbitt feels self-​similarity is a
necessary component of this hierarchy.2
This claim demonstrates the intersection of two of Babbitt’s primary
influences: Schenker and cognitive psychology. The idea that a piece develops
outward in hierarchical, self-​similar stages is quite directly Schenkerian, but
the justification for this ontology is its ability to facilitate cognitive chunking
and hierarchization. The hierarchy crucial to both Schenkerian organicism
and the cognitive model that Babbitt propounds does indeed suggest, as the
quotation intimates, what defines a complete piece, a “closed totality,” for
Babbitt. Pieces require a progressively derived hierarchy built on self-​similar
principles. These principles will determine what defines a level—​and, ac-
cordingly, what defines the largest, piece-​spanning level—​as complete. But
examination of Babbitt’s music reveals that hierarchical logic actually does
not, in general, determine a piece’s completeness. Therefore, the challenges to
systematic completeness in Babbitt’s music result in challenges to the func-
tion and status of both hierarchical organicism and the cognitive approach
outlined in Chapter 2 as models for understanding his music.3
The basic level-​defining principle in Babbitt’s compositional hierarchies
is the principle of exhaustive completion. The principle, abstracted from the
initial exhaustive list that is a twelve-​tone series, is extended to every aspect
of Babbitt’s compositional technique. Accordingly, exhaustive completion
provides us with a reasonable hypothesis for what defines a completed piece,
a “closed totality,” for Babbitt. Exhaustive lists are “cumulatively contained”
within one another, and when the largest list is complete, the piece should
be over.
The exhaustive lists further suggest a model for understanding temporal
progression in Babbitt’s music. A listener may be able to track these lists, per-
haps following the narratological implications some of them seem to inspire.
Babbitt himself occasionally described the varying clarity or ambiguity of a
piece’s series, a process embedded in all-​partition arrays—​among his most
frequently deployed kinds of exhaustive lists—​in a quasi-​narratological way.
In the (121) partition, the series is simply laid out monophonically, and in
other partitions with long segments, the series is, similarly, relatively clear.

2 This program note was prepared for a March 14, 1984, performance at the Library of Congress

(Milton Babbitt Collection, Box 5, Folder 4).

3 As discussed at the end of Chapter 1, other writers have come to broadly similar conclusions

about the use of Babbitt’s writings in the analysis of his music. See particularly Dubiel 1997.
198 Thinking In and About Music

In the more even partitions—​(112), for instance—​the series might be almost

completely obscured. As a result of this fluctuating serial clarity, Babbitt
(1997, 133) explains that “disambiguation [of the piece’s series] is itself an
element in the overall automorphic progression.” He describes Reflections
as embodying a “certain kind of musical progression” in which a “con-
fused” beginning initiates a process of “disambiguation,” which gradually
“unweaves” into a “very explicit statement of the materials of the piece”
(Babbitt 1998b). Other sorts of piece-​spanning exhaustive lists may also lend
themselves to narratological interpretation. Dubiel’s (1992, 83–​91) analysis
of Composition for Four Instruments, for instance, constructs a narrative de-
scription of that piece’s exhaustive list of instrumental combinations.4 To the
extent that listeners are aware of the principle of the largest list—​admittedly
more likely with the list of instrumental combinations in Composition for
Four Instruments than with the list of partitions in an all-​partition array—​
progression through the list could organize their temporal experience. They
could track their progress through the piece as the list unfolds, coming to
expect that the completion of the largest list will herald the completion of the
piece as a whole.
Often enough, however, the principle of exhaustive completion does not,
in fact, determine the overall scope of a piece. In several pieces, a large-​scale
exhaustive list is either preceded or followed by other material. As these
pieces seem to extend beyond their systematically normative boundary
points, they will be called “over-​complete.” In other pieces, multiple suc-
cessive exhaustive lists span the piece but are not contained within an over-
arching exhaustive list. Accordingly, these pieces will be called “multiply
complete.” Finally, many other pieces present the beginning of an exhaus-
tive list that is cut off before its completion; these, in turn, will be called
“under-​complete.”5 Over-​completeness, under-​completeness, and multiple

4 It should be noted, however, that Dubiel’s analysis of Composition for Four Instruments is

presented within a discussion of what he calls the “animation of lists,” which complicates the image
somewhat. Roughly, Dubiel aims both to use the list of instrumental combinations to construct a
narrative through the piece and simultaneously to discuss how the piece challenges, subverts, or
transcends this list. Likewise, Dubiel’s (2008, 130) analysis of Allegro Penseroso seeks to attribute an
“aspect of succession, of progression” to the piece’s exhaustive list of registral combinations while
recognizing that “its significance for our experience is intermittent and inconsistent” and cautioning
that one should resist “imposing . . . a degree of teleology on this music that the music may not
sustain.”
5 Babbitt was not the first composer to begin permutational patterns and cut them off before com-

pletion. As described by Daniel Harrison (1988, 43), “The majority of Bach’s works which use triple
counterpoint do not employ all six possible arrangements.” Nonetheless, the stakes are considerably
higher in Babbitt’s music. Harrison continues, “Rather, triple counterpoint is generally a secondary
means of organization in the service of other structures—​harmonic, thematic, and so forth.” Yet in
 Completeness and Temporality 199

completeness are not mutually exclusive: a piece can, for instance, be under-​
complete in one dimension and over-​complete in another. These situations
are common enough to cast doubt on the normative status of these lists, the
hierarchy that contains them, and the linking of a piece’s completeness with
the completion of exhaustive lists.

Over-​, Under-​, and Multiple Completeness

Babbitt’s late piano piece Allegro Penseroso (1999) presents both over-​
completeness and under-​completeness in one or another of its dimensions.
As shown in Figure 6.1, the piece begins by presenting the sixty-​three com-
binations of six registers arranged into thirty-​one complementary pairs and
a final tutti (see Dubiel 2008, 124–​31). The piece’s pitch structure is idio-
syncratic, when compared to Babbitt’s usual practice, containing neither a
trichordal array nor an all-​partition array: the only level of aggregate for-
mation is that found between the pairs of lines within each register. In each
section, each sounding register contributes one aggregate, a Schoenbergian
hexachordal aggregate in which each series contributes six notes, and this is
the only contrapuntal technique of aggregate formation in the piece. Without
an encompassing array structure that organizes the piece using nested levels
of aggregate structure and an exhaustive list of partitions, the sequence of
registral combinations becomes the piece’s sole over-​arching exhaustive list.
However, rather than ending with the exhaustion of this list, the piece
concludes with six additional combinations of registers, shown at the bottom
of Figure 6.1. This makes the piece appear over-​complete. These final, extra
combinations might be considered a coda, coming after the primary formal
structure is complete and serving rhetorical ends that contribute to a sense of
closure. The piece begins with the highest register in isolation, then proceeds
to fill in the remaining registers and, eventually, all combinations of those
registers. This process culminates in the tutti, with the fragmented registers
finally coming together after having been held apart for the entire piece up to
that point. The final, additional registral combinations that follow gradually
bring the piece back to its starting point. Moving more slowly now, with each

Babbitt, there are no significant large-​scale structures apart from the exhaustive lists. The widespread
under-​completeness in Babbitt’s music represents not the sidelining of a “secondary means of organi-
zation” in favor of a primary one, but a challenge to the primary means.
200 Thinking In and About Music

Figure 6.1. List of registral combinations in Allegro Penseroso (adapted from

Dubiel 2008, 128). Each point represents a pitch-​class aggregate.

register contributing two aggregates (which is to say, each line completing a

series form), the piece retreats toward the top of the keyboard.
Meanwhile, and characteristically, the rhythmic material is precisely
equivalent to the pitch-​class material, and just as the pitch-​class counterpoint
is delineated by six registers, the time-​point counterpoint is delineated by six
dynamics. There is one change from usual practice here. Often in Babbitt’s
later pieces, pitch-​class material articulated by high registers corresponds to
rhythmic material articulated by loud dynamics, as in None but the Lonely
Flute (see Chapter 1). In Allegro Penseroso the scheme is inverted. This allows
the opening to be isolated in the highest register and at softest dynamic,
heightening the sense of “starting from 1,” as Dubiel (2008, 124) describes it.
As shown in Figure 6.2, the dynamics begin exactly the same pattern as the
registers, reflecting the equivalence of the pitch and rhythmic structures. But
 Completeness and Temporality 201

Figure 6.2. List of dynamic combinations in Allegro Penseroso (adapted from

Dubiel 2008, 128). Each point represents a time-​point aggregate.

over the course of the piece, the dynamics only present the first twenty-​four
of the sixty-​three combinations in the pattern. Thus, if exhaustive lists are the
measure of completeness, then from the perspective of pitch structure and
register Allegro Penseroso appears over-​complete, while from the perspective
of rhythmic structure and dynamics it appears under-​complete.
Composition for Four Instruments provides another example of a
piece that is over-​complete in one dimension, in this case the dimen-
sion of rhythm. The pitch material for Composition for Four Instruments
is presented in fifteen successive trichordal arrays, the last seven retro-
grading the partitioning of the first eight. The rhythmic structure is based
on analogous material, with transformations of the duration series <1, 4, 3,
2> combined and partitioned identically to the trichords of the trichordal
array (see Mead 1994, 67–​68). These rhythmic arrays unfold much more
slowly than do the pitch arrays, such that the whole piece presents just two
full statements: an original followed by its retrograde. This construction
echoes, in its retrogression, the arrangement of partitions in the trichordal
arrays.6
The second rhythmic array concludes in m. 381, twenty-​five measures
from the end of the piece. As the rhythmic material that follows is not de-
rived from the hierarchy of exhaustive lists that form the array, the rhythmic

6 Structural palindromes like this, in which the second half of a piece uses the retrograde of the

array of the first half, are quite common in Babbitt’s output (although they are not composed out as
literal palindromes). The large-​scale durational patterns of Composition for Tenor and Six Instruments
and “Sounds and Words,” discussed in Chapter 3, provide additional examples. Palindromes can be
considered a proper, if simple, application of the principle of exhaustive lists, in the sense that they
present all possible ways of presenting an array using only the R transformation. They also relate to
another important concern of Babbitt’s, to be discussed later in this chapter: that he views a return to
a piece’s opening as a technique for signaling closure.
202 Thinking In and About Music

structure of the piece as a whole appears over-​complete. The rhythms of the

final twenty-​four measures, shown in Figure 6.3, are a drastic simplification
of the earlier material, a simple succession of basic transformations of the
original rhythmic pattern presented monophonically—​one after another—​
and not superimposed or augmented. The result is that the final measures
sound simple, clear, and calm, a gentle denouement after the tumult of the
rest of the piece. One is tempted to call this a coda, although this term would

Figure 6.3. Composition for Four Instruments, mm. 382–​405, with

transformations of the rhythmic series <1, 4, 3, 2> indicated. Score in C.
 Completeness and Temporality 203

Figure 6.3. (Continued)

properly apply only to the rhythmic dimension: the pitch array structure is
precisely complete, finishing with the last note of the piece.7
A Solo Requiem is also rhythmically over-​complete, but unlike the over-​
completeness seen in Composition for Four Instruments or Allegro Penseroso,
A Solo Requiem begins with material appended to the array rather than

7 In the second movement of Three Compositions for Piano, written just before Composition for

Four Instruments, Babbitt uses a similar technique. The introduction, interludes, and coda all present
straightforward, monophonic presentations of the piece’s duration series (see Mead 2011, 17–​22).
204 Thinking In and About Music

Figure 6.4. Sketch for A Solo Requiem, p. 1 (Milton Babbitt Collection, Box 16,
Folder 2). Used by kind permission of the Estate of Milton Babbitt.

ending with it. Accordingly, the extra material might be described as an “in-
troduction”—​although, once again, only as viewed from the rhythmic per-
spective. From m. 4 until the end of the piece, the time points complete two
full all-​partition arrays, corresponding precisely to the arrays supporting the
pitch material of the first two settings in the cycle. But the first three meas-
ures use time-​point material not drawn from the array. Figure 6.4 gives the
first page of Babbitt’s sketch for the piece: the bracketed upper staves outline
the piece’s two piano parts, while the lowest line indicates the time points
used in this passage. As can be seen, there is no time-​point array here, but
rather only a single line, <A, B, C, E, D, G, C♯>, proceeding through the first
seven notes of a member of the piece’s series class.
The sketch reveals that at one point Babbitt planned this opening sequence
of time points to be one “half step” lower. The crossed-​out line of notes directly
above the time-​point line that was actually used is <A♭, B♭, B, E♭, D♭, G♭, C, A>.
This is the beginning of the series form that appears in the (121) partition of
Figure 6.5. A Solo Requiem, mm. 1–​4. Introductory time points are indicated in
the manner of Babbitt’s sketch of the passage: pitch-​class intervals correspond to
durations measured in sixteenth notes.
206 Thinking In and About Music

the piece’s first array. (The significance of the crossed-​out line above that, di-
rectly under the sketch of the second piano part, is obscure to me; presumably
it represents an even earlier attempt at sketching the time points of this pas-
sage.) Babbitt may have discarded this series form—​despite its significance
later in the piece—​in order to establish a link to the array proper. The first time
point of the time-​point array appears with the entrance of the voice in m. 4.
This is nine sixteenth notes after the final attack of the “introduction,” corre-
sponding to ordered interval 9 between the seventh and eighth time points
of this series. That is, the next time point in the series form presented at the
bottom of Figure 6.4 is B♭, and the voice’s first time point is notated as B♭.
One plausible motivation for this introduction in the time-​point dimen-
sion is that it gives Babbitt an opportunity to construe dynamics broadly,
apart from their usual function of delineating time-​point lines. Indeed,
as can be seen in Figure 6.4, Babbitt sketches dynamics directly into the
piano parts.8 This is unlike his usual practice during the 1970s, ’80s,
and ’90s, in which dynamics are indicated by the time-​point array at the
bottom of each system (see Figure 1.6). It is not apparent that there is any
systematic rationale guiding these choices of dynamics. What is apparent,
however, is that Babbitt abandoned this dynamic scheme sometime be-
tween this sketch and the completed manuscript. The entire opening
passage, up through the entrance of the voice, is 𝆏𝆏 in the finished score
(see Figure 6.5). Although the result is reminiscent of the 𝆏𝆏 beginning of
Allegro Penseroso, the decision seems more likely to be motivated by the
piece’s text and genre. The Requiem begins with a mysterious shimmering
scattered across the two pianos’ registers, every note sustained throughout
the passage. Its ghostly effect is heightened by the diminished seventh
chord that opens the piece, an unusual sonority in Babbitt’s practice (al-
though common in this piece) emphasized through the immediate reit-
eration of each of its pitches. Although the sketch in Figure 6.4 reveals
that Babbitt did not initially write this time-​point “introduction” with a
consistent 𝆏𝆏 in mind, it is not an effect he could have achieved within the
bounds of the time-​point array.

—​—​—​

8 In Figure 6.4, “f ,” “f ,” “f ,” and “p ” appear to be shorthand for   𝆑𝆑𝆑, 𝆑𝆑𝆑𝆑, 𝆑𝆑𝆑𝆑𝆑, and 𝆏𝆏𝆏. Babbitt
3 4 5 3
uses this shorthand elsewhere in his sketches.
 Completeness and Temporality 207

The hierarchical program sketched in Chapter 1 depends on the nesting of

exhaustive lists within larger exhaustive lists. In certain pieces, however,
multiple exhaustive lists are presented in succession at the largest level, not
contained within any global exhaustive list. An example of one of these “mul-
tiply complete” pieces is My Ends Are My Beginnings.9 Over the course of the
piece, the time points present just one all-​partition array. Therefore, from the
perspective of rhythm, the piece is precisely complete. Meanwhile, the pitches
complete three distinct all-​partition arrays (of which the first is equivalent to
the time-​point array). This multiply complete arrangement appears to have
been motivated, or at least treated, programmatically: as detailed by Mead
(1994, 189), the hexachordal distribution of the three pitch arrays forms a
large-​scale ABA pattern—​reinforced by the use of clarinet for the first and
third arrays and bass clarinet for the second array—​reflecting the piece’s title.
Multiple completeness can also appear in compositions that combine
multiple types of arrays. Babbitt began to combine all-​partition arrays and
trichordal arrays in 1979, with Paraphrases, and continued in numerous
pieces over the next two decades. In the earlier of these works, such as The
Head of the Bed, Groupwise, Lagniappe (1985), and Consortini (1989), al-
though the trichordal arrays and all-​partition arrays are each separately
based on exhaustive lists, they are not subsumed by any larger, piece-​
spanning exhaustive list. The title of the first of these works suggests that
the trichordal arrays are derived from, and thus “paraphrase,” the series
presented in the all-​partition arrays.10 In this respect these pieces remain
conceptually unified. But they no longer present hierarchies of exhaustive
lists.11
—​—​—
​
The discrepancy between the treatment of pitch and rhythm seen in many of
the aforementioned examples—​namely, that the rhythmic material unfolds at

9 The piece’s title echoes and pluralizes Machaut’s Mon fin est mon commencement. Ways in which

the work’s “ends” reflect its “beginnings” are explored in Mead 1994, 202–​3.
10 Babbitt’s notes on Paraphrases explain the pertinence of the work’s title thus (Milton Babbitt

Collection, Box 12, Folder 2). For more on Paraphrases, see Mead 1984, 326–​31.
11 One might similarly consider as multiply complete certain of Babbitt’s earlier pieces that concate-

nate different types of arrays, such as Three Compositions for Piano, Composition for Twelve Instruments,
Du, Woodwind Quartet (1952), String Quartet No. 2, Two Sonnets, Vision and Prayer, and Philomel.
Glosses also combines different kinds of arrays—​arrays of series forms and trichordal arrays. In these
pieces as well, the discrete arrays are not subsumed within a global exhaustive list. Regarding the later
pieces that combine trichordal and all-​partition arrays, such as Manifold Music and When Shall We
Three Meet Again?, the rotating pattern of references discussed in Chapter 1 may be taken to provide a
large-​scale, subsuming list that renders these pieces properly (rather than multiply) complete.
208 Thinking In and About Music

a slower rate than the pitch material—​is a standard feature of Babbitt’s music,
with only a handful of exceptions. Babbitt never explained why this should
be the case, but plausible explanations touch on both systematic concerns
and preferences regarding texture. Texturally, this discrepancy would appear
whenever there are more pitches sounded simultaneously than there are time
points reiterated with different pitches (as noted in Schubert 1994, 73–​75). As
Babbitt systematically adopted the technique of cross-​references discussed
in Chapters 1 and 2, the problem intensified, for the technique demands that
numerous pitches fall between structural time points. The combined result
of these concerns is that even though pitch and rhythmic structures typically
present equivalent material, the pitches tend to move through this material
considerably faster. Therefore, at least one dimension will usually appear ei-
ther under-​complete, over-​complete, or multiply complete. The precise coor-
dination of pitch and rhythm, which would facilitate having both dimensions
present complete structures, is attempted in a few pieces from the 1960s (in-
cluding “Sounds and Words,” Vision and Prayer, and Post-​Partitions [1966];
on the latter, see Mead 1994, 173), but Babbitt did not make it a general fea-
ture of his practice.
Throughout the 1970s, in pieces such as String Quartets No. 3 (1970)
and No. 4, Tableaux (1972), Arie da Capo (1974), A Solo Requiem, and More
Phonemena (1978), Babbitt dealt with this discrepancy between pitch and
rhythm by using the approach seen in My Ends Are My Beginnings: the time
points and pitches both complete arrays, but time points complete fewer
arrays than the pitches.12 Typically, the time points present arrays equiv-
alent to the first one or two pitch arrays. There is usually a fixed ratio be-
tween the rates of unfolding of the pitch and rhythmic arrays (see Scotto
1988, 9–​12). But by the mid-​1980s, Babbitt changed his priorities. The
pitches in these later pieces, with a handful of exceptions to be discussed
shortly, express complete arrays, but the time-​point arrays, proceeding at
a rate slower—​often many times slower—​than the pitches, are left under-​
complete. When the pitch arrays are completed, the piece is finished, re-
gardless of whether the time-​point arrays have been completed.

12 With the exception of My Ends Are My Beginnings, the multiple pitch arrays in these pieces are

subsumed within a larger exhaustive list, and therefore these pieces are not multiply complete. For
instance, as Dubiel 1991, 107–​8 describes, the six consecutive pitch arrays of A Solo Requiem are per-
muted such that each of the six active instruments or registers project each of the six combinatorial
pairs of the piece’s array once in the piece.
 Completeness and Temporality 209

Figure 6.6 shows the complete superarray for pitches in Danci: each of
three registers presents a trichordal array.13 Figure 6.7 shows the superarray
for the piece’s time points; characteristically, it is precisely equivalent to the
pitch array.14 However, twenty-​seven time points at the end of the array (in-
dicated with strike-​throughs in the figure) simply do not appear. The pitch
array is completed in the middle of the time points’ final composite aggre-
gate,15 and the piece simply ends there.

Figure 6.6. Pitch-​class superarray for Danci. Parentheses indicate pitch classes
inserted between notes of the series; brackets indicate serially expected notes
that are absent.

13 The piece-​spanning exhaustive list governing Danci’s pitch structure is a rotating pattern of

references, as in String Quartet No. 6 (see Chapters 1 and 2).

14 Or more precisely, the pitch and time-​point superarrays can be considered equivalent if one adds

4 to each time-​point class beginning on the second quarter note of m. 32. It is as though there is a
quarter note missing in that measure, and indeed a simple notational mistake might well be to blame.
Shifts like this appear in many of Babbitt’s time-​point pieces. It is unclear if they are intentional.
15 A “composite aggregate” occurs, in a superarray, when the separate arrays each complete an aggregate.
210 Thinking In and About Music

Figure 6.7. Time-​point superarray for Danci, presuming the piece begins on
time point 4 and adds 4 to each time-​point class (i.e., shifts forward a quarter
note) beginning on the second quarter note of m. 32. Strikethroughs indicate
time points not used in the piece.

Danci is a characteristic example of Babbitt’s practice from the mid-​1980s

until the end of his life: the pitch arrays are usually complete, and the piece
ends when they have been exhausted, leaving the slower time-​point arrays
under-​complete. However, Danci is interesting, and somewhat uncharacter-
istic, in that the time points almost complete their array. Had the pitches been
unfolded only slightly more slowly relative to the time-​point array, perhaps
because more pitches were repeated or fewer presented in simultaneities,
the time-​point array would have been complete. Such a move would not be
without precedent: as Peter Schubert (1994, 67, 75) has noted, Babbitt ends
Three Cultivated Choruses (1987) with several measures of repeated pitches
 Completeness and Temporality 211

in order to use the remaining time points. In Danci, however, Babbitt makes
no such effort.
More typically, particularly later in Babbitt’s career, a piece will conclude
long before the time points finish their array. Phonemena (1969), apparently
the earliest example to use an under-​complete array, uses fifty-​nine time-​
point aggregates of a seventy-​seven-​aggregate all-​partition array (Kuehn
1995, 24). In Mehr “Du”, there are sixteen of an expected twenty-​nine time-​
point aggregates—​or rather just over sixteen, as the piece also uses the
first three time points of what would have been the seventeenth aggregate.
In Preludes, Interludes, and Postlude (1991), there are forty-​seven of an ex-
pected fifty-​eight time-​point aggregates. In None but the Lonely Flute, there
are twenty of an expected fifty-​eight time-​point aggregates. In Around the
Horn (1993), there are twenty-​three of an expected thirty-​four time-​point
aggregates. In Soli e Duettini for flute and guitar (1989), there are twenty-​
nine of an expected fifty-​eight time-​point aggregates. And over the course of
the nearly half-​hour-​long String Quartet No. 6, there are just over thirty-​one
of an expected fifty-​eight time-​point aggregates.
Several of these pieces follow a practice begun as early as Composition for
Tenor and Six Instruments, in which the time points present a simplified ver-
sion of the pitch arrays. Often, the pitches will present multiple arrays (suc-
cessively or combined in a superarray), but the time points present only one
of the first of those arrays. This was noted previously in My Ends Are My
Beginnings and A Solo Requiem. In String Quartet No. 6, for instance, each
instrument presents two full passes through the piece’s fifty-​eight-​aggregate
array, while all of the time points together contribute to only a single array,
one equivalent to the cello’s first array. Nonetheless, the time-​point array is
still significantly under-​complete.
Perhaps the most extreme example of under-​completeness in the time-​
point dimension can be found in Soli e Duettini for two guitars (1988). The
superarray for this piece is a long crab canon: each guitar presents an array
with fifty-​eight aggregates, and these arrays are related by retrogression, as
well by the M transformation (Mead 1994, 256). The time points present an
equivalent superarray, with the loud dynamics corresponding to the material
in Guitar 1 and the soft dynamics corresponding to the material in Guitar
2. However, the time-​point arrays each use only nine aggregates of the ex-
pected fifty-​eight. Notably, since they do not reach the mid-​point of the array
(or, indeed, anywhere close to it), it is not evident, from examination of the
212 Thinking In and About Music

time points alone, that they are even presenting retrograded versions of the
same array.
There does not appear to be any systematic reason why the under-​complete
arrays end where they do. They do not tend to end at any particularly significant
moment in the array such as an aggregate boundary or at the completion of a
series form. Nor, in any of these later pieces, is there a fixed relationship between
the rates the pitch array and rhythmic array unfold that might be thought to de-
termine when the time-​point array will be truncated. The practice seems to be
that the unfolding of the various arrays is not systematically determined. Thus,
Babbitt’s early formulation of the relationship between pitch and rhythm as one
of “structured rhythmic counterpoint” overstates the logical rigor that this rela-
tionship assumes in his later practice. The governing rule seems to be that the
arrays unfold independently, and when the pitch array is over, the piece is over.
This suggests that in pieces in which there are multiple arrays in various
dimensions, Babbitt prioritizes one, such that completion in that domain is
considered a more essential determinate of completeness. Consistently in
these later pieces, pitch is prioritized over rhythm (although this is not the
only example of prioritized domains, as will be demonstrated shortly). The
project outlined in Babbitt ([1962] 2003) and elsewhere of a rhythmic struc-
ture that is equivalent to pitch structure yet functionally independent appears
to have been at least partially abandoned. The desires for “holism” and cogni-
tive reinforcement, which he cites as reasons for creating equivalence between
the dimensions, appear not, in the end, to have been sufficiently powerful
concerns to justify completing the independent yet equivalent arrays.
The prioritization of pitch over rhythm seen in the aforementioned
examples of under-​complete time-​point arrays can also be seen in other
aspects of Babbitt’s compositional practice, and even in certain of his theo-
retical claims.16 First, there are a number of pieces—​notably, the so-​called
pulse pieces, to borrow Richard Swift’s term, as well as many vocal pieces—​
that simply do not have an independent rhythmic structure.17 But excepting
Homily (1987), for snare drum, Babbitt’s only piece for solo unpitched

16 Consider his statement that “[pitch] is . . . the most important of the musical dimensions, since

its susceptibility to musical structuring includes and exceeds that of any other dimension” ([1972]
2003, 285). Babbitt refers here to the scales of measurement discussed in Chapter 2.
17 Swift’s coinage is cited in Mead 1987, 215. “Pulse pieces” are based around a fixed pulse (always a

quarter note) that is subdivided according to the system of cross-​references discussed in Chapters 1
and 2 in relation to None but the Lonely Flute and String Quartet No. 6. Babbitt uses this technique
in every piece written in 1982 and 1983 (The Head of the Bed, String Quartet No. 5, About Time
[1982], Canonical Form [1983], and Groupwise), as well as in the much later Concerti for Orchestra
 Completeness and Temporality 213

percussion, there are no examples of pieces that have a rhythmic array but
no pitch array. In the pulse pieces, just as in the many other pieces that sys-
tematically pursue cross-​references, rhythm is directly determined by a pitch-​
structural concern (i.e., the references to the pitch-​class array determine the
duration and placement of the notes), but there are no examples of a rhythmic
concern directly determining pitch. Finally, there are a number of ways in
which pitch structure appears to be more carefully organized than rhythmic
structure. For instance, the technique of cross-​references, in which pitches
simultaneously participate in the underlying array counterpoint and in
references to other parts of the array, has essentially no correlate in rhythmic
structure. One can occasionally observe patterns in the surface organization
of time points: in None but the Lonely Flute, for instance, the time points begin
with three consecutive series segments.18 But for the most part, it appears to
be Babbitt’s typical practice that time points are simply used as necessary to
support the pitch material and to achieve desired gestural effects. Accordingly,
it seems consistent with other aspects of his practice that at the conclusion of
the pitch array in many later pieces, the time-​point array is simply cut off.
The best sketch evidence for the theory of prioritized arrays comes not
from an under-​complete time-​point array but from the under-​complete
pitch arrays in Piano Concerto No. 2. In this piece, the orchestra is divided
into duos and trios, each group presenting a separate all-​partition array. The
solo piano part, meanwhile, presents its own set of arrays. In an interview
about the concerto, Babbitt (1998b) discusses the priority of the piano part,
describing it as the “focal instrument,” and the whole piece as “virtually an
accompanied cadenza.” Cross-​references support this, he explains: there is
“nothing that happens in any instrumental part that hasn’t happened, or isn’t
about to happen, or isn’t happening, or isn’t destined to happen, in the piano.”
This focus on the piano part may explain why the piano arrays are complete
while the orchestral arrays are all left under-​complete.
It is evident from his sketches that Babbitt typically prepared arrays in
advance and then checked off notes as he used the pitch classes they de-
note.19 The arrays created for the piano part of Piano Concerto No. 2 are,

(2004). Other pieces that derive their rhythms entirely from their pitch-​class arrays, using various
techniques, include Playing for Time (1977), My Complements to Roger (1978), and the final section
of Paraphrases (Mead 1987, 214–​20). On the lack of rhythmic serialism in some of Babbitt’s vocal
pieces, see Chapter 5, n44.
18 See Figure 1.8. The segments are <016872>, <e4359t>, and <56e170>.
19 Babbitt does not appear to have checked off notes in the array as he composed time points, per-

haps because the notation of time points in his sketches allows for easier comparison with the array.
214 Thinking In and About Music

accordingly, all checked off. But even though the arrays were prepared in
full for the various orchestral duets and trios, the checks stop after twenty-​
nine aggregates. And indeed, only the checked-​off notes appear in the fin-
ished score. Figures 6.8 and 6.9 give two representative pages from the array
prepared for the vibraphone and marimba; Figure 6.8 is from the first half

Figure 6.8. Vibraphone and Marimba array chart for Piano Concerto No. 2, p. 4
(Milton Babbitt Collection, Box 13, Folder 1). Used by kind permission of the
Estate of Milton Babbitt.
 Completeness and Temporality 215

Figure 6.9. Vibraphone and Marimba array chart for Piano Concerto No. 2, p. 6
(Milton Babbitt Collection, Box 13, Folder 1). Used by kind permission of the
Estate of Milton Babbitt.

of the array, with checks indicating Babbitt’s progress as he composed; and

Figure 6.9 is from the second half, showing a page Babbitt prepared but did
not use. Babbitt’s sketches reveal that at one stage in the composition, he cre-
ated complete arrays for each of the groups of orchestral instruments. Their
216 Thinking In and About Music

under-​completeness was, apparently, not anticipated. However, having com-

pleted the piano part—​the prioritized part—​the rest of the ensemble simply
stops. Indeed, the release of the piano’s last note marks the end of the piece.
The foregoing examples of prioritized dimensions illustrate the most
common kind of under-​completeness in Babbitt’s music: situations in which
one dimension—​the prioritized dimension—​is still complete even if the sub-
sidiary dimensions remain under-​complete. However, in at least four cases,
no array whatsoever is completed. Each of these compositions recycles all-​
partition pitch-​class arrays Babbitt had prepared for earlier pieces. But al-
though each piece begins these familiar arrays, none finishes them. Lagniappe’s
final section uses twenty-​eight aggregates of a seventy-​seven-​aggregate pitch-​
class array (see Mead 1994, 240). “Pantun” (2000) uses thirty-​seven aggregates
of a fifty-​eight-​aggregate array. “Now Evening after Evening” (2002) uses, over
its entirety, fifty-​three aggregates of a seventy-​seven-​aggregate array. Concerti
for Orchestra is in some senses the most confounding example, because the
structures begun rather clearly at the beginning (including, but not limited to,
all-​partition arrays in each group of instruments) seem to gradually dissolve,
eventually fading into an extended silence and a brief section of music at the
very end, unconnected to any array but gently recalling some of the material
heard near the very beginning (see Morris 2012, 407).
In sum, given the pieces that appear over-​complete, under-​complete, or
multiply complete, we should set aside the idea that Babbitt’s music derives
a sense of completeness solely from the exhaustion of lists. Moreover, the
model of perceived formal progression predicated on tracking these lists
must similarly be revisited: in my experience and, to the best of my know-
ledge, the experience of others, pieces with under-​complete arrays are not
perceived as incomplete. The extensively worked-​out organic hierarchy with
its carefully constructed nested layers built on a single principle does not, in
the end, explain why one of Babbitt’s pieces is complete.
Indeed, it is not apparent that there is any general systematic model for
completeness in Babbitt’s music. Compositional evidence suggesting other-
wise is lacking, particularly in pieces with under-​complete arrays: as shown
in the sketches for Piano Concerto No. 2, for example, it appears that Babbitt
had no preconception of how or when the piece, or at least its orchestral parts,
might end. The failure of exhaustive lists to adequately explain complete-
ness in Babbitt’s music, the fact that no alternative systematic explanation
seems forthcoming, and the sketch evidence that in early stages of compo-
sition Babbitt himself was unsure of significant aspects of a piece’s overall
 Completeness and Temporality 217

development, suggest that, in looking for a systematic and logical answer to

the problem of completeness, we may have been using the wrong approach.

Signals and Verticality

The indeterminacy in Babbitt’s preconception of the end of his Piano Concerto

No. 2 fits well with my experiences listening to his music—​particularly his
instrumental music. It has been observed by certain commentators that
Babbitt’s music seems non-​teleological. I admire this quotation of Joel
Krosnick, cellist in the Juilliard quartet and long-​time champion of Babbitt’s
music: “Dual . . . is not progressive in traditional ways, it is a state of mind, a
philosophical state.”20 This is my impression as well, for Dual and for many
of Babbitt’s works. In my own experience, I would say that there is typically
no basis for predicting whether a piece is two minutes or twenty minutes
from completion. I could go even further and say that there is often not even
anything about the music that would inspire listeners to make this sort of
prediction. One simply enters his music’s “state of mind” and stays there for a
while, leaving because the musicians have stopped playing, not because one
has arrived at a goal or completed any sort of logical argument. Thus, from
the perspective of large-​scale form, Babbitt’s music exemplifies the notion of
“vertical time” described by Jonathan D. Kramer:

A vertically conceived piece, then, does not exhibit large-​scale closure. It

does not begin but merely starts. It does not build to a climax, does not pur-
posefully set up internal expectations, does not seek to fulfill any expecta-
tions that might arise accidently, does not build or release tension, and does
not end but simply ceases. (Kramer 1988, 55)

As Krosnick suggests, certain performers and listeners have perceived

Babbitt’s music in this way for a long time.21 The present study confirms that

20 Included in the program notes for a March 14, 1984, performance of Dual (Milton Babbitt

Collection, Box 5, Folder 4).

21 See Bailey 2002, 1–​42; Scotto 2002, 320–​21; and Boretz 2006, 10–​11 for related discussions of

temporality in Babbitt (and, in the case of Bailey 2002, a number of other composers). A mode of
listening intriguingly resonant with “vertical time,” featuring a heightened experience of individual
events and their association with other events, rather than their subsumption into a hierarchy, is that
experienced by listeners with autism, as detailed in Straus 2011, 163–​67. Relatedly, see Straus’s (2018,
125–​54) discussion of autism and serialism as mutually reinforcing forms of cultural modernism.
218 Thinking In and About Music

this perception is supported at the deepest level of Babbitt’s compositional

techniques.
Beyond the matter of array completeness, two additional aspects of
Babbitt’s compositional techniques suggest a non-​directed, “vertical” con-
ception of temporal progression. In Babbitt’s analyses of Schoenberg’s twelve-​
tone music, perhaps his most frequently repeated comment regards the
“analeptic” and the “proleptic,” the ways in which a passage of Schoenberg’s
music can forecast developments to come or recall moments that have passed
(e.g., Babbitt 1987, 63–​84). Although Babbitt observes instances of analepsis
and prolepsis in the music of many composers, in Schoenberg’s twelve-​tone
music these discussions usually focus on a single aspect: hexachordal areas.
He demonstrates how a passage using one hexachordal area can imply an-
other hexachordal area, thereby suggesting a goal toward which the music
is moving. However, Babbitt conspicuously avoids this means of progres-
sion in his own practice. In Babbitt’s all-​partition arrays after 1970, all pos-
sible hexachordal area appear all of the time. That is, the hexachordal areas
of an array’s combinatorial pairs are another way in which all-​partition
arrays present exhaustive lists: the number of combinatorial pairs in an array
is determined by how many distinct hexachordal areas are available given
the hexachord of the array’s series.22 Babbitt thus closes himself off from a
principal “basis for moving through a piece” in twelve-​tone music (Babbitt
1987, 68).
Babbitt’s technique of cross-​references presents a similar challenge to di-
rectional progression. He describes these references as fostering analepsis
and prolepsis, either recalling or predicting array segments that appear
elsewhere in the piece (Babbitt 1998a, 28–​29). But there seems to be no de-
terminate pattern governing whether any particular reference is analeptic
or proleptic, or any pattern or principle regulating the distance between
references and referents.23 As a result, Babbitt does not appear to wield the
system of references in order to create a progressive, linear, teleological path
through the piece. The technique ensures that everything, in some sense,
echoes something else; “the menagerie is also a hall of mirrors,” as Dubiel

22 This number is inversely related to a hexachord’s symmetry. Therefore, arrays using first-​order

all-​combinatorial hexachords—​the A-​type, B-​type, and C-​type—​use six combinatorial pairs, arrays
using second-​order D-​type hexachords use three combinatorial pairs, and arrays using third-​order
E-​type hexachords use two combinatorial pairs.
23 A similar point is made in Dubiel 2012, 9.
 Completeness and Temporality 219

(2008, 136) evocatively describes it. And as in a hall of mirrors, the reflected
images serve only to be reflected again. They do not point to a goal.24
My hearing of Babbitt’s music is comparable to the listening strategy phi-
losopher Jerrold Levinson (1997) recommends for the apprehension of tonal
music in the common-​practice period. Levinson promotes a position he calls
“concatenationism,” the view that perception is focused on a chained suc-
cession of immediate contexts of up to a minute or so, rather than anything
that takes place over a larger timespan such as overall form: “Understanding
music is centrally a matter of apprehending individual bits of music and
immediate progressions from bit to bit” (13). Musical value and enjoy-
ment, he posits, are similarly dependent on local events (14). Levinson’s
concatenationism is not Kramer’s verticality—​there seems to be temporal
progression within music’s small-​scale bits as he understands them, although
this is not an aspect he focuses on—​but the two conceptions share a focus on
the immediate, rather than on large-​scale trajectories or formal schemes.
Overall, the position Levinson stakes out is more radical than I think is
sustainable—​among other things, his denial of the perceptual reality of
large-​scale form in common-​practice tonal pieces is contradicted by my own
experience. But Babbitt’s casual attitude toward the completion of large-​scale
array structures suggests that a listener may do well to focus on small-​scale
effects rather than large-​scale progressions through exhaustive lists—​that is,
to listen as a concatenationist. Such a listening strategy suggests a role for the
exhaustive lists besides that of securing hierarchy. Heard in this light, the lists
might contribute primarily to the music’s variety as it passes through the dif-
ferent partitions, registral patterns, instrumental patterns and so forth. This
variety exists whether or not the lists are completed—​something a purely
concatenationist listener will not be attending to anyway. Moreover, the aes-
thetic commitments of concatenationism indicate how a listener may value
a piece of Babbitt’s music with under-​complete arrays. A hierarchically ori-
ented listener is likely to be frustrated. A concatenationist listener attending
to immediate contexts may not be.

24 After describing how a hexachordal area in the middle of the first movement of Schoenberg’s

String Quartet No. 4 is prepared by a registral aspect of its opening, Babbitt (1987, 67) says, “that
there will be absolutely no doubt in your mind that somehow, somewhere, each group of six has been
associated before.” “Somehow, somewhere” is an intriguing concession: it is the fact of relation he
claims is perceived, not the specific means or referent. The non-​teleological deployment of references
in his own compositions, and the highly variable distance between references and referents, might
reflect a similar understanding of perception.
220 Thinking In and About Music

Several important qualifications to the general image of Babbitt’s music

as “vertical” or non-​linear should be made. First, although Babbitt’s pieces
generally do not trace a complete motion from beginning to end, they are
constantly suggesting directional movement of some kind or another, from
striking local gestures to longer motions stretching out over a minute or
more; Chapter 7 will take this up in detail. These local motions, in partic-
ular, are captured well by a concatenationist perspective. That they tend not
to extend over entire pieces or suggest predictions about a piece’s overall span
does not challenge their aesthetic significance.
A second qualification concerns the experience of moving through a list.
Particularly with large-​scale lists of registral patterns, instrumental com-
binations, or playing techniques, listeners will easily observe when Babbitt
moves from one combination to another. Even if they have not grasped the
underlying principle of the list or observed how far into the list the piece
has gotten, they will register the fact that, say, the clarinet has reentered,
or the viola just switched to pizzicato. Such moments inflect the temporal
flow: listeners know that where they are is different from where they were,
and therefore time has progressed. Within the duration of each combination,
there may be verticality, and verticality may structure temporal perception
on the broadest scale (both subject to the other qualifications discussed), but
at the moment of the change in active register, timbre, or playing technique,
an awareness of temporal progression becomes available.25
Third, even if most of Babbitt’s music is “vertical,” there are a number of
striking exceptions. Most of these date from the late 1940s to the early 1960s
when he worked primarily with trichordal arrays. These pieces often seem to
be goal oriented, as detailed thoroughly by Mead (see Mead 1994, 54–​123).
String Quartet No. 2, which gradually assembles its series before presenting
it in full near the end of the piece, provides the canonical example of this
early tendency towards directedness. It appears that Babbitt did not commit
to a non-​teleological perspective until the mid-​1960s, coinciding with his
embrace of the all-​partition array.
Furthermore, occasional glimpses of overall formal shapes can still be
found in a number of pieces written after this anti-​teleological turn. These
shapes often hinge on one moment in particular, the unique (121) partition in
an all-​partition array. Numerous analysts of Babbitt’s pieces with all-​partition
arrays have emphasized the importance of the (121) partition of a piece’s

25 I am grateful to Andrew Mead for discussions on this matter.

 Completeness and Temporality 221

pitch-​class array, and for good reason: its serial clarity and monophonic tex-
ture make it a highly marked moment. Babbitt will occasionally emphasize
the arrival of the (121) partition of the pitch-​class array through an instru-
mental playing technique not heard in the rest of the piece, by extending it
for a longer duration than the aggregates that surround it, or, conversely, by
compressing it into a virtuosic flourish.26 In several cases, the techniques
used to underscore this partition produce a perceptually salient passage sug-
gestive of a climax. Analyses by Mead (1983, 104; 1994, 153) and Daphne
Leong and Elizabeth McNutt (2005; 2019, 271–​72) describe the appearance
of the (121) partition of a pitch-​class array in the line assigned to the highest
register, thus resulting in an extended registral highpoint, as climactic. While
these passages do not function as the peak of any broad dramatic arc, they
nonetheless provide a startling disruption of a piece’s continuity. The (121)
partition of a piece’s time-​point array also occasionally receives special em-
phasis. This is especially true of pieces in which Babbitt assigns the loudest
dynamics to the line of the time-​point array that contains the (121) partition,
causing the (121) partition of the time-​point array to stand out with several
consistently loud measures. My Ends Are My Beginnings, Homily, and the
second half of A Solo Requiem all offer striking examples of this technique.27
Furthermore, Babbitt highlights the sense of climax available to this partition
through his placement of it: typically, toward but not at the end of the array.
As there is generally no systematic reason to present an array in any partic-
ular order, it is plausible that Babbitt would choose the order of his arrays pre-
cisely because of this resonance with traditional climax structure. However,
even in pieces in which the (121) partition provides a striking, climactic, or
crystallizing moment, it tends to remain just that—​a moment, not a goal that
had been approached over the course of the piece or that leads to a piece’s
conclusion. Moreover, in the many pieces with multiple all-​partition arrays
(whether superimposed or successive), the (121) partition is not unique.
Indeed, in most pieces with superarrays, it is hardly even noticeably marked.

26 The importance of the (121) partition to Babbitt is also evident in his sketches. The sketch ma-

terial for many pieces includes a list of (121) partitions with indications as to the instrumental or
registral lines that project them and the measure numbers in which they appear. The compositional
use of these lists is obscure: since they include measure numbers, Babbitt must have prepared them
after the indicated aggregates had been composed. It appears these lists were compiled for the sake of
record keeping, as guides to these crucial moments.
27 These passages are from mm. 308–​13 of My Ends Are My Beginnings, mm. 110–​16 of Homily,

and mm. 246–​49 of A Solo Requiem. For more on this passage of My Ends Are My Beginnings, see
Bernstein 2013, 294–​95 and Mead 1987, 230; for more on this passage of Homily, see Leong 2011.
222 Thinking In and About Music

Finally, even though one generally has no basis for determining, in the
midst of things, the degree to which one has progressed toward a piece’s con-
clusion, one can often sense that a piece is within a few seconds of ending. Or,
conversely, after one has heard the end of a piece, the final seconds often do,
in memory, indeed sound like an ending—​even if a piece’s arrays have not
been properly completed. This is because Babbitt uses a number of relatively
traditional closing techniques. Four of these techniques appear with great
frequency:

• motion to an extreme (for example, an extreme dynamic—​usually ex-

tremely soft—​extreme register, or an extremely thick or thin texture);
• a summation or clarification of a piece’s most important structural
materials;
• a return to one or more aspects of a piece’s opening; and
• repetition of notes, often while fading out.

Each of these techniques has a long historical association with closure in the
Western musical tradition.28 In this sense they might be considered—​almost
uniquely in Babbitt’s compositional practice—​rhetorical. They can be un-
derstood as signals to an audience, communicating that a piece is ending.
Each of these signals suggests, in their own way, that further development
in a piece will not occur. A motion to an extreme, however local, initiates a
process that cannot be continued; it can thus serve, synecdochically, to indi-
cate the end of the piece as a whole. Both summations of important materials
and reprises of the opening are inherently backward-​looking, indicating that
forward progression (such as it is) in the piece has stopped. The recapitula-
tory nature of these techniques is also effectively rhetorical, in the Classical
sense.29 Repetition—​admittedly a simple closing technique, but a very wide-
spread one in Babbitt’s practice—​echoes the traditional device of tonic af-
firmation. It can also be considered an indication that the pitch material of
the piece has been exhausted, that no new pitches are forthcoming.30

28 A number of these techniques, or close correlates thereof, are discussed in relation to common-​

practice music in Agawu 1987. On “structural framing” devices linking beginning and endings in
nineteenth-​century music, see Alegant and McLean 2007. But of course, there remain differences
between Babbitt and his predecessors. For instance, despite the conventional use of registral extremes
in Babbitt’s closing rhetoric, none of Zohar Eitan’s (1997, 8–​14, 20–​25) “secondary hypotheses” re-
garding melodic highpoints appear to be consistently supported in Babbitt’s music, although there
are momentary instances that may be taken to correspond to them.
29 See, e.g., Quintillian 1907, 407–​8.
30 The final composite aggregate of Whirled Series (1987), stretched out over twenty-​four meas-

ures, is perhaps the most luxuriously expansive demonstration of the technique of closing a piece
 Completeness and Temporality 223

Babbitt himself notes the second and third of these techniques in various
analytical comments. The third, the signaling of an ending by returning to
the beginning, ranks among the attributes of pieces he observed most fre-
quently: in his writings, he notes this in pieces by J. S. Bach, Schoenberg,
Bartók, Stravinsky, Varèse, Webern, Luigi Dallapiccola, and himself. It
is an attitude embodied by his quip on the second movement of Webern’s
Variations for piano, Op. 27 (1936): “And he’s back where he started, on B♭
and A♭, which is good enough to end any piece” (Babbitt 1987, 38). Both the
use of a return to the opening and the device of ending (and opening) with
an extreme—​indeed, the hint of a dramatic impulse that rarely emerges from
the hardened exterior of Babbitt’s prose—​are described in this comment on
his own Relata I: “Cymbal, tam-​tam, three drums, and wood block are used
only, and—​thus—​extravagantly, in the opening eight measures, and the cor-
responding final seven measures” (Babbitt [1970] 2003, 239).
These closing signals are used whether or not an exhaustive list is com-
plete. They are generally independent of serial hierarchy.31 Figure 6.10 shows
the final measures of “Now Evening after Evening,” which, as mentioned,
uses only fifty-​three aggregates of its seventy-​seven-​aggregate all-​partition
array. Although the array is left under-​complete, the piece still concludes
with two typical closing signals: a drop to 𝆏𝆏, the softest dynamic in the piece,
and a series of gentle, slow repeated notes in the piano.
The closing signals are not only independent of twelve-​tone hierarchy, they
can also sometimes even be understood to override that hierarchy. The “codas”
at the end of many over-​complete pieces generally appear to be designed to high-
light one or more of these techniques. The six extra combinations of registers in
Allegro Penseroso, as shown in Figure 6.1, instantiate a motion toward an ex-
treme register—​or rather, a gradual focusing on an extreme register, as the lower
registers successively drop out—​and a return to the register of the opening,
also at the top of the piano. Similarly, as seen in Figure 6.3, the last twenty-​four
measures of Composition for Four Instruments, a coda from the perspective of

with repeated notes (see Mead 1994, 226–​27). Jason Eckardt (1997, 97) compellingly relates the ex-
tension of this composite aggregate to cadential stability.

31 The independence of closing signals from pitch-​structural processes calls to mind the twelve-​

tone practice of Schoenberg to some degree. Like Babbitt, Schoenberg indicates closure through
parameters other than pitch-​class structure. Richard Kurth’s (2002) discussion of cadences in
Schoenberg’s Fourth String Quartet is illuminating on this point. Moreover, Schoenberg, to a much
greater degree than Babbitt, also relies on rhetorical devices borrowed from tonal practice, as has
long been noted (e.g., Boulez [1952] 1968).
224 Thinking In and About Music

Figure 6.10. “Now Evening after Evening,” mm. 51–​55

rhythmic structure, present a clarification of the basic duration series (as well
as, at the end, a sly repeated note). Even the mysterious ending of Concerti for
Orchestra can be understood as instantiating several of the closing signals.
An exceptionally clear example of an over-​complete piece whose coda
realizes Babbitt’s closing signals can be found in Homily, as shown in Figure
6.11.32 The coda, which begins at the indication “snares on, snare sticks” on
the final page of the printed score, succinctly enacts all four of the closing
signals. Until this section, there have been only three rolls in the piece, at the
end of the third, fourth, and sixth blocks of the piece’s all-​partition array. The
rolls that fill the entire coda thus dramatically expand on a device already

32 The final section of Homily is recognized as a structural coda in Leong 2011.

 Completeness and Temporality 225

Figure 6.11. Homily, concluding measures

associated with closure within the piece. They also present a rather extrava-
gant demonstration of the technique of ending a piece with repeated notes.
The last note of the piece is struck with both sticks at the same time. This
technique is used just six other times in the piece: on the very first note of the
piece, on the first note of blocks four, six, seven, and eight, and, anomalously,
on the downbeat of the third measure. The reprise of the technique at the
very end thus connects the ending not only with the beginning of the piece
but with important beginnings throughout the piece.
The dynamics enact a motion to extremes by expanding from 𝆑 and 𝆏 in
the first three measures of the coda to 𝆑𝆑 and 𝆏𝆏 in the final four measures.
Indeed, the use of both sticks on the final attack presumably has an accentual
effect, heightening the extreme dynamic: the final note is the only note in the
piece struck with both sticks and marked 𝆑𝆑, making it perhaps the loudest
moment in the piece. These dynamics articulate two time-​point series forms,
one presented in loud dynamics and one in soft, as shown in Figure 6.12.
After the piece’s all-​partition array, the straightforward presentation of two
226 Thinking In and About Music

Figure 6.12. Time-​point series in the coda of Homily, presuming the downbeat
of each measure is time point 0

Figure 6.13. “Sounds and Words,” mm. 82–​84

series forms partitioned into simple hexachordal aggregates realizes the last
remaining closing signal, the clarification of a piece’s basic materials.
Another striking example of a closing signal comes at the end of “Sounds
and Words.” Throughout the song, the voice and accompaniment proceed
throughout the piece without pitch-​class intersection. A surprising excep-
tion to this rule occurs at the very end of the piece, shown in Figure 6.13.
The last sound of the piece is not explained by the piece’s array. The piano,
in double octaves, echoes the final vocal pitch. Inexplicable serially,33 es-
sentially unrelated to any previous event in the piece,34 the motivation for
this lagniappe seems to be the realization of yet another of Babbitt’s favored
closing techniques. Luminously, and perhaps somewhat mischievously, the
piano echoes the final vocal pitch.35

33 The work’s array structure is discussed at length in Dubiel 1990a, 62–​75.

34 There are other octaves in the piano part of “Sounds and Words,” as can be seen in Figure 3.5, but
they always appear as duplications of array-​correct pitches.
35 The wit on display here, in which the final note upends the scheme otherwise developed so care-

fully throughout the piece, might lend itself to the psychological explication suggested in Babbitt
1997: “One can reconstruct the personality of a composer who would choose his final note in order to
confound any known construal of the piece until that point” (132). It is impossible to know whether
the quip is self-​effacing, but one hopes so.
 Completeness and Temporality 227

Babbitt also uses a number of characteristic and even familiar techniques

for beginning a piece of music. These are somewhat more varied than his
closing techniques, but certain general observations can be made. His de-
scription of the “opening of the main body” of Relata I—​that is, the begin-
ning of the all-​partition array in m. 8, following a brief introduction (see
Figure 3.2)—​is his most detailed discussion of the formal and rhetorical
requirements for beginning a piece of music. He describes this passage as
having “the obvious attributes of a ‘beginning.’ ” Chief among these attributes
is that this passage comprises a “minimal statement in many dimensions”; it
makes use of a minimal registral span in a “ ‘nonextreme’ octave,” a limited
dynamic range in the “midrange” of 𝆐𝆏 and 𝆐𝆑, and a simple hexachordal ag-
gregate construction. From this “minimal statement,” the piece develops out-
ward only gradually (Babbitt [1970] 2003, 240–​42).
Although there are other pieces that begin with restricted and mid-​level
values in their various dimensions—​Composition for Four Instruments, Swan
Song No. 1 (2003), and Concerti for Orchestra, for example—​Babbitt’s compo-
sitional practice shows considerably more variety than his statement on Relata
I suggests. Allegro Penseroso, for instance, as illustrated in Figures 6.1 and 6.2,
also opens with straightforward hexachordal aggregates, a minimal registral
span, and a limited dynamic range. However, Allegro Penseroso begins in an
extreme register and dynamic, two features common to the opening of many
of Babbitt’s pieces. Other pieces seem to represent an inverse perspective,
opening not minimally but maximally, with a sudden demonstration of the
piece’s potential. Perhaps the two pieces that begin most explosively in this
sense are Post-​Partitions and Septet but Equal (see Figures 7.3 and 7.5), but
the many pieces that begin with a quick flash through all available registers
also exemplify this tendency. As with endings, one can generalize to say that
Babbitt often starts his pieces with extremes of one kind or another, of which
an extremely “minimal” statement is one possibility.
Also as with endings, Babbitt’s beginnings of pieces frequently lay out a
piece’s most important structural materials. Many pieces, for instance, begin
directly with a statement of their series. “The Widow’s Lament in Springtime,”
Du, Two Sonnets, Semi-​Simple Variations, Partitions, Composition for
Synthesizer, Vision and Prayer, Philomel, Relata I, More Phonemena, Images
(1979), Paraphrases, It Takes Twelve to Tango (1984), Glosses, Piano Concerto
No. 2, and Pantun all state their series clearly at the outset. In many works that
use the time-​point system, the time-​point modulus is indicated with unusual
clarity by the repetition of the opening time point (see Mead 1987, 195–​97).
228 Thinking In and About Music

Pieces that employ a superarray, a structure that tends to obscure the aggre-
gate, often begin by presenting one aggregate separately, apparently in order
to establish the normative status of the aggregate before it is subsumed into
the piece’s counterpoint. As can be seen in Figures 6.1, 6.2, 6.6, and 6.7, Allegro
Penseroso and Danci exemplify this technique (see also Mead 2012, 385).
Just as material appended to the end of a piece’s overarching exhaustive list
can function as a coda that realizes Babbitt’s characteristic closing techniques,
so can introductions generally be understood as realizing his characteristic
opening techniques (although introductions are much less common in his
music than codas). This is true of the introductions of Three Compositions for
Piano I and II, Relata I, and Ars Combinatoria. Each of these introductions
presents its piece’s basic structural materials in a simplified or clarified form.
This is evidenced by Babbitt’s statement on the first eight measures of Relata
I, which he describes as “constitut[ing] literally and pervasively an intro-
duction, by introducing the main features of the pitch structures of the total
work,” as well as the main features of its temporal, dynamic, and instrumental
structure (Babbitt [1970] 2003, 239). Furthermore, each of these pieces ends
with a coda that largely recapitulates its introduction, thus instantiating the
closing technique of linked beginnings and endings. The time-​point intro-
duction to A Solo Requiem discussed previously might be also understood
as introducing the piece’s structural materials, in that it presents most of a
series form monophonically. More striking, however, is the flexible use of
dynamics this introduction makes possible, which opens the way for the use
of 𝆏𝆏 throughout the entire first three measures. This emergence from a soft
dynamic—​an appropriate gesture given the work’s programmatic function as
a requiem and the hesitant, apologetic tone of the Shakespeare sonnet with
which it starts—​constitutes yet another characteristic opening technique.
The opening and closing techniques establish the second major compo-
nent of perceived temporal progression—​and thus perceived completeness—​
in Babbitt’s music. After a piece’s beginning, the impression throughout the
bulk of it is generally that of “vertical time,” in the sense that the piece tends
to give no clues as to its overall scope. But then, at the end, there comes a
signal—​constructed using the aforementioned closing techniques—​and one
understands the piece to be complete.
 Completeness and Temporality 229

Conclusion: “Sounds of Relations”

In his writings, Babbitt promotes a hierarchical view of musical comprehen-

sion, a view reflected in a compositional technique based around nested ex-
haustive lists. A piece’s completeness, accordingly, would be defined, for both
composer and listener, by the completion of lists. But as we have seen, only
certain of his compositions, and only certain aspects of others, are bound
by this definition of completeness. Many extend beyond their systematically
normative endpoint, many others are cut off before the endpoint has been
reached, and still others consist of multiple discrete units, which do not form
a single, “closed” musical whole. The ways in which Babbitt’s music does
not accord with the implications of his theories—​ways in which his pieces
challenge notions of completeness suggested by those theories—​has led to
a reevaluation of their role and significance in his music. Most significantly,
the theory of listening Babbitt sketches is upended by these complications.
If one is to hear his pieces as self-​contained works, it will have to be in non-​
hierarchical terms. Approaching his music as largely vertical—​defined by a
broad temporal stasis composed of vivid gestures and inflected with opening
and closing rhetoric—​is likely to be more rewarding than coming to it with
teleological expectations would be.
This model for listening appears to be a far cry from the claims about or-
ganic hierarchy, cognitive chunking, and “unified closed totality” with which
this chapter began, and in a sense it is. Nonetheless, it still might be possible
to take Babbitt’s statements on these subjects seriously. The neat systematic
nesting of exhaustive lists and the comprehensive extensions of the twelve-​tone
system can be thought of as providing a kind of internal coherence or consist-
ency. They create a carefully defined field of activity, a world within which cer-
tain actions can receive a particular kind of attributive meaning in the context
of their surroundings. To the extent that there is completeness, it is of the kind
described by Dubiel (1991, 117): “There are sounds of relations, and sounds of
‘limitlessly profligate’ individuality and completeness of the moments so re-
lated and thereby so constituted, and the works they make up.” Closure in this
sense becomes a result of the interconnectedness of a piece’s contents, rather
than being defined by the attainment of a goal.36 Totality is a state, not the result

36 Although discussing tonal music, Mark Anson-​Cartwright (2007, esp. 10–​12) offers support for

the notion that closure may be a function of the interrelatedness of a piece’s constituent parts. As he
explains, “closure may inhere in a piece” (11). Citing Frank Kermode’s ([1966] 2000, 6ff.) classic study
of literary closure, Anson-​Cartwright notes that closure may be immanent—​present throughout a
work’s temporal span.
230 Thinking In and About Music

of a process; it is a condition of being, not of having been. And when the closing
signal comes—​whenever it comes—​the musical thread is severed; the contin-
uous present of vertical time is cut off. In the words of Scott Burnham (2016,
155–​56), “What a moment ago was vivid presence is now lived experience
(Erlebnis); the piece turns into the past in an instant.”
 5
Poetic Form and Psychological Portraiture
in Babbitt’s Early Texted Works

Introduction

Babbitt wrote five texted works between the years 1950 and 1964: “The
Widow’s Lament in Springtime” (1950), a setting of a poem by William
Carlos Williams; Du (1951), a song cycle on seven poems by August Stramm;
Two Sonnets (1954), a setting of a pair of poems by Gerard Manley Hopkins;
Vision and Prayer (1961), on a poem by Dylan Thomas; and Philomel (1964),
on a text jointly authored by John Hollander and the composer. He would
continue to compose large-​scale texted works for another twenty years and
texted works of smaller dimension for the remainder of his life. Music with
text is, in short, a central part of Babbitt’s compositional career.
Two topics will be explored throughout this chapter: Babbitt’s projection
of poetic meaning and his concern for poetic form—​the sonic, syntactic, and
visual aspects of poetry. The aspects of poetic meaning most consistently de-
veloped are twofold. The first, which is typically so blatant as to obviate the
need for commentary, is the use of madrigalism—​word painting. Babbitt’s
texted music throughout his whole life is filled with examples. “Screams,” in
Philomel, receives high notes followed by descending glissandi. “Says,” in The
Head of the Bed, is spoken. “False relation,” in “The Virginal Book” (1988),
coincides with a tonal false relation. “Eeee,” in Philomel, is set on E. Most of
his texted pieces have such madrigalisms, and as one may expect from the
composer of Four Play (1984), a fair number are groaners.
A more interesting aspect of his engagement with poetic meaning is the
representation of the psychology of characters in the texts—​a technique
I will call psychological portraiture. It is this aspect, in particular, that jus-
tifies the focus on these early works: in “The Widow’s Lament,” Du, and
Philomel, Babbitt uses devices that listeners may interpret as depicting the
psychological development of the poems’ characters. The derived series used
in these works were a fruitful source of metaphor, and Babbitt’s turn away

Thinking In and About Music. Zachary Bernstein, Oxford University Press. © Oxford University Press 2021.
DOI: 10.1093/​oso/​9780190949235.003.0005
126 Thinking In and About Music

from derived series in the 1960s also matches an apparent decline in his in-
terest in psychological portraiture.1
The aspects of poetic form most clearly articulated in these works in-
clude syntax, poetic meter, line length, rhyme scheme, and, more generally,
phonemic parallelism (that is, sonic echoes among the phonemes of a text).
Babbitt notes his concern for poetic form, and his desire to musically project
and reinforce aspects of poetic form that can be obscured by musical setting,
in this telling quotation:

When just the “musical” aspects of the texts—​the sonic, durational, accen-
tual rhythms and their progressive successions—​become part of a total mu-
sical structure, those attributes of poetic coherence are so reoriented as to
be attenuated and even obscured as characteristics in themselves; but their
structural functions can be restored and reinforced in the new amalgam
by the associative means of the correlated musical materials. It is with such
considerations that my settings begin. (Babbitt 1979b, 45)

It is a goal of this chapter to explore the “correlated musical materials” and

their contribution to the resulting text-​music “amalgam.”
Reflecting his compositional interest in the sounds of poetry, Babbitt had a
life-​long fondness for texts with a high degree of phonemic parallelism; that
is, with many patterns of rhymes—​particularly those that are not straight-
forward end rhymes. One can witness this in his numerous comments on
the sonic richness of particular texts or lyrics (see, e.g., Babbitt [1985a] 2003;
1997, 131–​32; 2005), or in his choice of heavily rhyming texts by Hopkins and
Stramm, but a more direct way to observe the kinds of texts that interested
him is to look at a lyric Babbitt wrote himself, shown in Figure 5.1. “As Long
as It Isn’t Love,” included in the collection Three Theatrical Songs, was written
in 1946 for an unproduced musical, Fabulous Voyage, during Babbitt’s brief,
youthful stint on Broadway.2
In this lyric, there are end rhymes, internal rhymes (“merest mash”; the
repetitions of /​n/​in the third stanza), and several instances of a clever sonic
word play characteristic of the Tin Pan Alley tradition Babbitt draws from
(“passion,” “mash in”; “enter any,” “enterprise”). The meter of the lyric is

1 Although Philomel postdates Composition for Tenor and Six Instruments, it uses derived series of

a different kind, as will be discussed.

2 The musical is an adaption of The Odyssey. “As Long as It Isn’t Love” is sung by Calypso. The line-

ation of Figure 5.1 concords with Forte 1997, 70.

 Poetic Form and Psychological Portraiture 127

Figure 5.1. Lyric for “As Long as It Isn’t Love,” by Milton Babbitt. From Three
Theatrical Songs.

subtle and intricate: an iambic pattern with numerous variations. One might
imagine how a lyricist such as this could favor, in a more serious mood, the
resonance of Hopkins and the metric play of Williams.
The works addressed in this chapter cover a relatively narrow span of
Babbitt’s output, but in their concern for setting of local textual parallelisms,
they are characteristic of his career as a whole; it is principally in their focus
on psychological portraiture and in the techniques deployed for projection
of large-​scale poetic form that they differ from his later texted works. Two
examples from opposite ends of Babbitt’s career demonstrate his sustained
interest in matching sonic parallelism to musical parallelism. The first, shown
in Figure 5.2, is the setting of the first stanza of Figure 5.1. Two features are
worth noting. First, musical patterning reflects the poetic meter. When there
are regular iambs, the music settles into quarter notes with stressed syllables

Figure 5.2. “As Long as It Isn’t Love,” mm. 1–​7, vocal part. From Three
Theatrical Songs.
128 Thinking In and About Music

Figure 5.3. “A Waltzer in the House,” mm. 11–​14, vocal part

on beats. When there is variation (the opening inversion, the closing ana-
pest), the musical rhythm is more flexible. Second, a repeated descending
major seventh brings out a phonemic parallelism. The rhyme between “pas-
sion” and “mash in” is obvious enough without musical emphasis, but the
repeated seventh also links this rhyme to “a lust-​”—​whose vowel sequence
is a near-​rhyme with those of “passion” and “mash in.” In this way, musical
parallelism brings out a textual parallelism that would likely otherwise pass
unnoticed.
“A Waltzer in the House” (2003) was composed almost sixty years later but
demonstrates Babbitt’s continued interest in some of the same concerns; see
Figure 5.3. The song is a brief and amusing setting of a poem by Stanley Kunitz
for soprano and vibraphone. In the excerpt shown, Babbitt again brings out
the poem’s meter through placement of stressed syllables on beats. The lines
are iambic, but both begin with a parallel bit of catalexis: the opening weak
syllable is removed so that each line starts with the interjection “O.” A de-
scending major second at the start of each line underscores this repetition.
Moreover, this example demonstrates Babbitt’s taste for madrigalisms of a
rather direct, even silly, kind, as wide oscillations illustrate “swaying” and
“bobbing.” While Babbitt’s text setting is not always as straightforward as in
these examples, sensitivity to poetic meter, correspondence between poetic
parallelism and musical parallelism, and a compulsion toward madrigalism
are characteristic of virtually all his settings.
Beyond observation of the occasional madrigalism (e.g., Babbitt [1964a]
2003, 156), or a note about a text’s pertinence to a particular occasion,
Babbitt was almost silent on issues of poetic meaning.3 But his comments
on his own works—​generally in unpublished program notes—​nearly always

3 A rare exception is his program note for The Head of the Bed (in the Milton Babbitt Collection,

Box 8, Folder 2), which incisively comments on the “dual role of narrator and protagonist” the singer
embodies.
 Poetic Form and Psychological Portraiture 129

mention the relationship between poetic form and music. Unfortunately,

these comments are typically so vague and elliptical as to be virtually useless
to an analyst seeking to investigate that relationship in any detail,4 but they
can be taken, at the least, as an indication of the significance of such matters
to the composer. His observations on others’ texted music, too, touch on po-
etic form, and are often more revealing about his compositional interests
than his writings about his own works. A brief essay on Jerome Kern (Babbitt
[1985a] 2003) is unusually insightful in this respect. In one intriguing re-
mark, Babbitt notes admiringly that the lyrics to Kern’s “In Love in Vain,”
written by Leo Robin, have “just two terminal rhymes, both rhyming with
the title”—​the remarkably “fluid fit” of the words to Kern’s melody is a re-
sult of the “less apparent means of sonic association, such as assonance and
consonance within and between lines” (397). As this quotation indicates,
Babbitt was not interested in more blatant rhymes; he had a preference for
subtler phonemic parallelisms. Correspondingly, his own settings concen-
trate on bringing out medial rhymes, slant rhymes, and other phonemic
parallelisms more intricate and interesting than straightforward end rhymes.
Babbitt reveals that his background in popular music directly inspired this
interest in a number of remarks.5 Despite his evident admiration of Kern,
however, Babbitt is willing to criticize him—​particularly on the grounds of
inadequate text setting. In two songs, Babbitt finds that Kern places musical
emphases—​a registral peak in one instance, a large leap in another—​that
contradict the lyrics’ syllabic stresses (Babbitt [1985a] 2003, 397). As can
be seen in the preceding examples, Babbitt was concerned to avoid similar
metric contradictions in his own work.

4 A typical example, from the program note to “The Widow’s Lament” (Milton Babbitt Collection,

Box 8, Folder 2): “The subtle, intricate sonic and syntactical structure of the poetry further deter-
mined harmonic choices, registral dispositions, and polyphonic deployments, large and small scale
references and recurrences, and even the structure of the underlying all-​interval, hexachordally de-
rived twelve pitch-​class series.”
5 One amusing instance bears quotation:

[Allen Forte, commenting on Babbitt’s Three Theatrical Songs] makes us mindful of or

reminds us of the “lyricists,” not primarily those Broadway fashioners of the polysyllabic
rhyme or the clever, catalogue one-​liner, but those less of Broadway than of Hollywood
and “Tin Pan Alley” who carefully crafted the sounds of words, the relations of words
to music, and of words to words. The uses not of just rhyme, terminal or internal, but
of degrees of assonance, consonance (direct and cognate), chiasmus, and more are
precedents which have affected my settings of Du, of The Head of the Bed, and—​indeed—​
all my settings. Surely it must have been an awareness of such verbal elegances that
allowed Ruby Keeler satisfiedly to accept Dick Powell’s limited physical offering in I Only
Have Eyes for You. (Babbitt 1997, 131–​32).
130 Thinking In and About Music

Babbitt’s engagement with poetry can be understood, in part, as a con-

tinued expression of the aesthetic and cognitive views that governed his en-
tire musical aesthetic. In crafting correspondences between text and music,
his music evidences, as it does in the purely musical realm, a desire to maxi-
mize meaningful relationships. When he can create a serial cross-​reference,
he does so; and when he can create a madrigalism, a musical parallelism un-
derlying a textual parallelism, or a musical form that interacts purposively
with a poetic form, he will generally do likewise. The added dimension colors
but does not change the underlying principle that musical richness and com-
prehensibility are functions of interrelationships.
But it should also be noted, especially in these early works, just how signif-
icantly he would bend his musical practice in order to accommodate the texts
he set. Text introduces artistic criteria separate from the contextual logic of
serial hierarchy. In some situations, Babbitt responds by merely coordinating
serial events with textual events. But in others—​notably Du and Philomel—​a
desire to reflect poetic meaning led him to musical structures that diverge
from the practice and principles developed in his instrumental work.

“The Widow’s Lament in Springtime”

“The Widow’s Lament in Springtime,” a setting of a poem by William Carlos

Williams for soprano and piano, is the first texted serial piece Babbitt
published. The song is remarkable both for its sensitive lyricism and for the
fact that it introduced a number of Babbitt’s techniques for text setting. It
is also Babbitt’s first collaboration with Bethany Beardslee, the soprano for
whom Babbitt would compose nearly all of his texted works over the next
twenty-​five years. Figure 5.4 provides the poem along with a few annotations
depicting aspects of Babbitt’s setting.
Williams presents an ironic juxtaposition of a widow grieving among the
bursting new life of spring flowers. Over the course of the poem, the widow
observes the bright, primary colors of spring, and finds no comfort in them
to assuage her grief. By the end of the poem, she wishes to disappear into the
flowers—​to find death where she had previously found joy. Such doleful sub-
ject matter would reappear in many of Babbitt’s texted works, particularly
in the first half of his career. He favored poems engaging themes of loss and
madness. But in this case, the circumstances of the poem’s composition may
have at least partially inspired the choice of text: as Babbitt describes in his
Figure 5.4. “The Widow’s Lament in Springtime” text. Superscript numbers indicate the beginnings
of sentences; superscript Roman numerals indicate the beginnings of array sections.
132 Thinking In and About Music

program note for the song (included in the Milton Babbitt Collection, Box 8,
Folder 2), it was premiered at a memorial for Roy Dickinson Welch, his de-
partment chair at Princeton, in the presence of Welch’s widow.
In addition to its topical appropriateness, Babbitt’s program note claims
that his interest was piqued by the rhythmic and accentual characteristics
of Williams’s verse. Williams’s short, free-​verse lines typically receive two
or three strong stresses and one or two lighter stresses, although the total
number and weight of stresses is variable. These stresses are arranged into
various patterns, from liquid iambs to rigid spondees, with the result that the
poem’s fluctuating momentum reflects the widow’s shifting emotional state,
from observation and nostalgia to declarative resolve.6 Such shifts are also
reflected by the variable length of the poem’s sentences. The first sentence,
as the widow introduces her new, fallen state, is of an unremarkable length.
But the next sentence, “Thirty-​five years /​I lived with my husband,” which
identifies her as a widow, has an amputated terseness, suggestive of pains too
deep to express. The third sentence, too, is laconic: she observes the white
plum tree, but in a distanced, purely observational manner. It is only in the
expansive fourth sentence that she gives full expression to her situation and
its cruel dissonance with her surroundings. The final sentences, as the widow
imagines oblivion, return to a less remarkable length.
The brevity of the poem’s lines contributes to its heavy enjambment. This
effect is audible, if a bit subtle, in Babbitt’s setting, as he uses conventional
means (such as rests or long sustains) to indicate the ending of lines. The en-
jambment and irregular rhythm carry the poem along, such that there are
only three strong breaks: after the first sentence, after the fourth sentence,
and between the last two sentences. Babbitt’s setting only slightly articulates
the break between the fourth and fifth sentence, but the other two breaks
are marked by the only rests of a measure or longer. As will be discussed,
Babbitt also uses serial means to set off the outer sentences, with expressive
consequences.
The soprano begins with the song’s six-​note series.7 The vocal part under-
lying the entire first sentence of the poem—​that is, the first section of the
piece’s array—​is then composed of straightforward transformations of this

6 Winters 1937, 118–​19 and Hobsbaum 1996, 114–​16 provide valuable commentary on the poem’s

accentual structure. As will be discussed subsequently, Babbitt knew Winters’s analysis and seems to
have been directly influenced by it.
7 An overview of serial structure corresponding to that given here can be found in Morris 2010,

45–​47. A few pages of sketch material for the piece are available in the Milton Babbitt Collection, Box
18, Folder 2.
 Poetic Form and Psychological Portraiture 133

Figure 5.5. The beginning of each of the three main sections of “The Widow’s
Lament in Springtime.” Text by William Carlos Williams. Copyright © 1950
Mobart Music Publications. Copyright © renewed. All Rights Reserved. Used by
permission of European American Music Distributors Company, sole U.S. and
Canadian agent for Mobart Music Publications.

series. Meanwhile, the piano reflects the voice’s material: within each aggre-
gate the piano’s three registers complement the voice’s trichord with the three
other orderings of that same trichord.
The middle of the song is composed using trichordal arrays based on de-
rived series, developing the segmental trichords of the primary series. As
shown in Figure 5.5, the derived arrays begin by directly replicating the
opening trichord of the series. Within each of the four arrays, each line-​pair
successively develops the four trichords presented in the piece’s series. This
is shown, along with the hexachordal progression, in Figure 5.6. Following
a somewhat different pattern from the trichords, the linear hexachords of
each line-​pair cycle through the four all-​combinatorial hexachords that can
be generated by those four trichords (these are “linear” in that they are the
discrete hexachords of each line; note the A-​type hexachords in the derived
series of Figure 5.5). As a result of the shifting trichordal and hexachordal
134 Thinking In and About Music

Figure 5.6. Trichords and hexachords in the arrays for “The Widow’s Lament
in Springtime.” “A,” “B,” “C,” and “E” refer to the familiar “types” of all-​
combinatorial hexachords.

content, the division of the piece into its various arrays is reasonably au-
dible: harmonic changes, like from the chromatic A-​type hexachord to the
diatonic C-​type hexachord, mark sectional breaks. In particular, the pres-
ence of tonal material—​(037) triads or C-​type hexachords—​is notable, and
may even be read to have a semiotic function, indicating heightened expres-
sivity. The two sections (III and IV) presenting triad-​derived series coincide
with the long fourth sentence, in which the widow describes the color of
the flowers around her, as well as expressing her grief in unusually poignant
terms. As her focus turns away from the flowers and toward her death in the
final sentences of the poem, tonal material returns to the more concealed
presentation found also in the song’s opening (as discussed in Straus 2018,
Example 6.12).
“Harmonic” hexachords—​hexachords formed by the total content of
each line-​pair within each aggregate—​are less obviously patterned than the
trichords and linear hexachords. They do not create any exhaustive list and
on that basis might be taken as having less form-​generating significance than
the trichords or linear hexachords. However, it is notable that in sections II
and V, the two sections in which the linear hexachords omit the B-​type hex-
achord, the harmonic hexachords include it. In this way, every section of
the piece uses, in one fashion or another, the hexachordal type of the piece’s
 Poetic Form and Psychological Portraiture 135

series. As a result, even though Babbitt’s setting features harmonic contrast—​

most significantly, four distinct all-​ combinatorial hexachords—​ it also
presents a high degree of continuity. This reflects the poem, which focuses
throughout on only one image.
For the setting of the final sentence of the poem, Babbitt returns to com-
plete statements of the primary series. The setting of this final sentence is
shown in Figure 5.7. As will be discussed, Babbitt’s return to the primary
series and the original tempo has implications for both poetic form and
meaning.
Most of the middle of the song is set off by a slight increase in tempo. This
nearly matches the piece’s serial structure: the increase in tempo comes with
the beginning of the third sentence, rather than the second, which had initiated
the piece’s derived interior. The return to the opening tempo comes precisely
with the return to presentations of complete series forms at the beginning of
the sixth sentence. The quickened pace in the middle of the song corresponds
with a rhythmic change in the poem: the heavy spondees of the first sentence
(“own yard,” “new grass,” “cold fire”) give way to a somewhat livelier verse,

Figure 5.7. The vocal part of “The Widow’s Lament in Springtime,” mm. 42–​49,
and a reproduction of the analysis of corresponding lines in Winters 1937, 119.
Winters’s accent symbols have been added to Babbitt’s score. Text by William
Carlos Williams. Copyright © 1950 Mobart Music Publications. Copyright
© renewed. All Rights Reserved. Used by permission of European American
Music Distributors Company, sole U.S. and Canadian agent for Mobart Music
Publications.
136 Thinking In and About Music

with more unaccented syllables coming between stressed syllables. The final
sentence of the poem is again heavy, somewhat like the opening—​including
a spondee in the very last line (“marsh near”; see Figure 5.7). The disjunction
between the first tempo change and the transition to derived series is a bit of
a puzzle, but it might be explained in practical terms. The break between the
first and second, fourth and fifth, and fifth and sixth sentences coincides with
a boundary between arrays. In this respect, serial structure articulates poetic
form. Very little marks the break between the third and fourth sentences, but
this can be explained by Williams’s repetition of “masses of flowers,” which
bridges the end of the third and the beginning of the fourth sentences. Perhaps
the absence of musical articulation can be understood to support this in-
stance of anadiplosis. But this leaves the break between the second and third
sentences, which also does not coincide with the end of an array (nor, practi-
cally, could it: the brief second sentence could not be stretched across the eight
aggregates of a trichordal array without either a melismatic or contrapuntal
technique inconsistent with the rest of the setting). The placement of the tempo
change provides musical articulation for this sentence ending.
In order to observe the poetic significance of the use of different kinds of se-
rial material, it will help to return to the coordination of sentences and array
boundaries, shown in Figure 5.4. Superscript Arabic numerals indicate the
beginning of the poem’s six sentences, while Roman numerals indicate the be-
ginning of the six sections of the array. The coincidence of the total number of
array sections and sentences in the poem is suggestive. The first, fifth, and sixth
sentences of “The Widow’s Lament” are precisely coextensive with sections of
the array. For these sentences, the piece’s serial structure directly projects the
poem’s syntax. The irregularities in the syntactical construction of the middle
of the poem are matched by a loss of coordination between poetic and con-
trapuntal structure: the terse second and third sentences are presented within
the beginning of the second array section, while the expansive fourth sentence
rambles across the end of the second, third, and fourth sections of the array.
The layout of the song’s arrays—​the fact that it begins and ends with com-
plete series while its interior is derived—​is richly suggestive from a herme-
neutic perspective. The song’s trichordal derivation can be imagined as an
examination of its series: a focused lens zooming in on each of its trichords.8
Similarly, we might take the relation between the parts of the piece as one of

8 Hush 1982–​
83, 159 employs a similar metaphor in an analysis of Composition for Twelve
Instruments.
 Poetic Form and Psychological Portraiture 137

surface and depth, the primary series presenting a broad Gestalt that each
derived series dives into and dissects. We may even read the relation as one
between present and past: the primary series giving a synoptic overview
from a distanced, present-​day perspective, while the derived series provide
recollection of individuated moments.
All these possibilities are suggestive of readings of Williams’s poem. The
first sentence sets out the situation. The middle of the poem provides the
reason for the widow’s lamentation—​her husband’s death—​and features her
observation of the contrast between springtime’s bursting flowers and her
own sorrow. The last sentence, then, presents her response to her tragedy as
she yearns to disappear into the flowers.
The outer sentences are firmly in the present tense. The middle section,
although it includes present-​tense glances at the flowers surrounding the
widow, are framed by her recollections. Therefore, Babbitt’s choice to set the
outer sentences with the primary series and the inner sentences with derived
series provides a reading of the psychological action of the poem. Listeners to
the song are carried, with the widow, into reflection and examination.
Finally, the setting of the poem’s metrical structure deserves comment.
Babbitt mentions that he learned of the poem in an analysis by Yvor Winters
(1937), part of which is shown in Figure 5.7.9 Winters analyzes the irregular
meter of the poem by marking heavily stressed syllables with two dots and
lightly stressed syllables with one dot. He marks lines that have other than the
usual two heavy stresses with asterisks. Babbitt’s setting matches Winters’s
analysis quite closely: double-​dotted syllables tend to be on downbeats, to
have long durations, or to initiate a slur, while single-​dotted syllables receive
some musical emphasis but a bit less than the double-​dotted syllables. The
setting of the final sentence, shown in Figure 5.7, is particularly faithful to
Winters’s analysis—​every downbeat is matched to a double-​dotted syllable.10
In Babbitt’s output as a whole, “The Widow’s Lament” might be best un-
derstood as experimental. It is his first serial vocal piece. In some respects,
perhaps accordingly, it is an outlier. There is, as Igor Stravinsky (1982,
100) says, a “distinctly American and very lovely pastoral lyricism” in “The
Widow’s Lament” that would fade in the coming years. But all the techniques

9 See Babbitt’s program note in the Milton Babbitt Collection, Box 8, Folder 2.
10 There is, however, one double-​dotted syllable in this example that is not set on a downbeat:
“go.” This appears in the only line in the piece’s final sentence to receive only one heavy stress. The
meter seems to be relevant here: “go” appears on the second beat of the only measure in the section
in .   Therefore, two beats separate “go” and the next downbeat—​an interonset interval equivalent to
that which usually separates downbeats in this passage.
138 Thinking In and About Music

developed for the projection of poetic form and psychological portraiture—​

articulation of syntactical units, attention to the poem’s meter, and explora-
tion of the interpretive possibilities of derivation—​would be developed in
the works to come.

Du

Du followed shortly on “The Widow’s Lament” and was premiered by the

same duo, Bethany Beardslee and pianist Jacques Monod (Beardslee with
Proctor 2017, 373). The song cycle is a setting of seven poems by August
Stramm selected from a collection of the same title, published posthumously
in 1917. As Morris and Alegant (1987, 94–​95n.32) have pointed out, the cycle
represents an expansion and elaboration of techniques introduced in the ear-
lier song. The first and last songs of Du, like the first and last sentences of “The
Widow’s Lament,” are set using complete six-​note series forms, while its inner
songs and interludes use trichordally derived lines. There are poetic similarities
as well—​themes of loss and distress as well as unusually short lines and much
enjambment. But one probably should avoid pushing too hard at the works’
similarities, either poetically or musically. Williams is subtle and restrained;
Stramm depicts the sputtering ravings of a madman. The scale of “The Widow’s
Lament” is intimate; Du is grandiose and metaphysical. The approach to lan-
guage could also hardly be more different: Williams’s unobtrusive vernacular
contrasts starkly with Stramm’s obscurities, ambiguities, and neologisms.
Babbitt’s settings of the poets’ work are sensitive to these differences: the lyrical
character of his setting of “The Widow’s Lament” is replaced by volatility and
angularity throughout most of Du. (This difference is even more noticeable in
the accompaniment than it is in the vocal part.) And perhaps most intriguingly,
while both “The Widow’s Lament” and Du begin and end with statements of
complete series, in “The Widow’s Lament” it is the same series class that opens
and closes the song, while in Du, the outer songs present distinct series with
distinct harmonic content. This is, in fact, Babbitt’s only work with multiple
distinct underlying series classes, in apparent representation of the personal
and cosmic rift depicted in Stramm’s poetry.11

11 Several late works require the invocation of broader-​than-​usual series classes incorporating the

M operation (see Mead 1994, 233–​37 and 255–​63), but this provides a different, and less problem-
atic, set of issues than Du, whose series cannot be transformed into each other by any standard serial
operation.
 Poetic Form and Psychological Portraiture 139

Like much of the poetry Babbitt set, including all of the remaining
examples in this chapter, Du is rife with phonemic parallelism.12 The end of
“Wankelmut” presents a striking example, an almost sub-​linguistic cataract
of rhyme: “Wirr /​Wirren /​Wirrer /​Immer wirrer /​Durch /​Die Wirrnis /​
Du /​Dich /​Ich!” In future settings, Babbitt would take exceptional pains to
project such rhymes and alliterations; Du is less systematic in this regard. But
perhaps the preponderance of these parallelisms makes special means of pro-
jection unnecessary. They come through clearly without special intervention.
Du may be Babbitt’s most committed attempt at realizing narrative and
psychological portraiture through serial means. For this reason, I will dwell
on it at some length.

—​—​—​

Stramm’s Du is, as he subtitles the collection of thirty-​one poems, a set of

Liebesgedichte—​love poems.13 And like any set of love poems, its subject is
the relationship between one character and another; here they are, at once
intimately and anonymously, referred to as simply “Ich” (“I”) and “Du”
(“You”), although biography and context suggest that Ich is probably male
and Du probably female. The situation is one of mad, unrequited love by Ich
for Du. Ich searches (“Mein suchen sucht!” [“My seeking seeks!”]), Du oc-
casionally comes within sight, but to no end. Stramm’s work, as Christopher
Waller (1986, 25) describes it, “is a poetry founded in a desperate sense of
isolation.”
Ich sees Du as invested with tremendous powers, even divinity (“Du trägst
Antwort . . . Du siegst Gott!” [“You carry [the] answer . . . You are victorious
God!”]). Numerous scholars of Stramm have commented on the mystical
forces that haunt his poetry and on Du’s apparent command of those forces.
As Christoph Herring notes about “Werben,” a poem in Stramm’s Du that, al-
though not set by Babbitt, is representative: “Cosmic law places Man and the
mysteries of life in a position of waiting. Woman alone, the great mediator,
could grant communication. But there will be no fulfillment. . . . The struggle
between the two erotic principles will never end” (Herring 1961, 48). The
metaphysical stage on which Stramm’s Du is set ensures that Ich and Du can

12 This is a characteristic of Stramm’s poetry generally. Perhaps for that reason, Babbitt returns to

Stramm’s poetry twice, in A Solo Requiem (1977) and Mehr “Du” (1991).
13 The entirety of Du has been reprinted in Stramm 2013.
140 Thinking In and About Music

never be reconciled. This is no simple case of unrequited love. Ich and Du

represent the eternal cosmic grappling of Stramm’s dark vision.
Babbitt chose seven of the thirty-​one poems in Stramm’s cycle to set in
his own Du, and there is therefore something of a distinction between the
effects created by Stramm’s entire cycle and those created by Babbitt’s se-
lection. Moreover, Babbitt’s ordering of the seven poems he selected is sig-
nificant in a way that the arrangement of Stramm’s Du evidently was not, at
least to Stramm.14 The result is that by their ordering as well as by musical
aspects of their setting, Babbitt has imposed a narrative structure onto these
seven poems.
One consequence of Babbitt’s selection is that the range of situations
in which Stramm depicts Ich and Du is somewhat diminished. As Waller
(1986, 31) describes, sometimes, in Stramm’s cycle, Ich’s quest is successful,
leading to poems that culminate “in an ecstasy of sexual fulfillment.” Waller
names the poems “Erhört” and “Erfüllung” as exemplifying this tendency.
But Babbitt set neither of these poems, nor any other poem in which Ich
and Du can reasonably be regarded as interacting anywhere outside of Ich’s
feverish imagination. From within the perspective of the seven poems in
Babbitt’s Du, the ending of “Schwermut,” the last poem in Babbitt’s cycle—​
“Tief /​Stummen /​Wir” (Deeply /​Grow mute /​We)—​presents the only fact
we know about the two characters’ relationship. In regard to each other, they
are both, deeply, silent. The seven poems Babbitt set are reproduced and
translated in Figure 5.8, with several annotations pertaining to the analysis
to follow.15
The anonymity of referring to Ich and Du by pronouns alone reflects a
confusion about their identity that has been taken to be a central concern
of the cycle: as John Rahn (1976, 79) writes, “[The poems’] theme seems to
be the emotional relations, confusion, distinction, and identity among Du,
Ich, Dich, Mich, and Wir.” Waller (1986, 31) summarizes the possibilities for
Du’s identity that have been put forth by various scholars: “The ‘Du’ has been
variously interpreted as Stramm’s wife, as God, as the cosmos, as somebody/​
anybody of the opposite sex, as the opposite sex, as somebody/​anybody out-
side the poet’s self.” All these interpretations seem more or less plausible at

14 According to Jeremy Adler (1980, 124), Stramm did not order Du, “but left that task to his friend

and publisher Herwarth Walden.”

15 I am grateful to Sharon Krebs, Harald Krebs, and Tobias Tschiedl for their assistance with this

translation.
 Poetic Form and Psychological Portraiture 141

Figure 5.8. Texts set in Babbitt’s song cycle Du, with annotations, selected from
Stramm 2013. Translation my own.
142 Thinking In and About Music

Figure 5.8. (Continued)

various points in the cycle, and indeed a reading that allows for a protean Du,
one who sometimes embodies realistic attributes of femininity and is some-
times characterized more abstractly, is probably closest to the mark. From
the perspective of the analysis to follow, the fact of confusion and ambiguity
over identity is more important than any particular reading.16

16 As Richard Sheppard (1985, 281) points out, ambiguity over identity is compounded by syntag-

matic ambiguity, in that the pronoun “Du” might refer to Du, the persona, or might refer to proxi-
mate nouns in particular poems that can be construed as pronoun antecedents: “ ‘Hölle Teufel /​Du
 Poetic Form and Psychological Portraiture 143

The separation between Ich and Du that defines the cycle is reflected
by a divergence within serial structure.17 The trichordally derived lines
in the interior of the cycle use six distinct trichord orderings, but these
trichords, unlike those of “The Widow’s Lament,” cannot be understood
as segments of the same underlying series. Instead, the most efficient pos-
sible source for the “total pitch material of the composition” is a pair of
distinct series that appear in the vocal part at the beginning and ending
of the piece. Figure 5.9a presents the piece’s trichordal vocabulary. Figure
5.9b gives the beginning and ending of the piece’s vocal part, the most
plausible candidates for the work’s series, with their segmental trichords
marked.18
The fact that the piece can only be reduced to a pair of series, not a single
one, can be interpreted as an “enactment,” in David Lewin’s (2006, xiiff.)
sense, of the disjuncture between Ich and Du. Roughly, the A-​type series
and its trichordal subsets are associated with Ich and Ich’s internal preoccu-
pations and hysteria, while the C-​type series and its subsets are associated
with Du and, generally, states outside of Ich’s tortured reality.19 Both array
construction—​the trichords and hexachords that appear in the arrays used
in each song—​and surface details support this enactment of the cycle’s cen-
tral, fractured relationship. Figures 5.10 and 5.11, which show the work’s
trichords and linear hexachords, will be referenced throughout the ensuing
discussion.
The trichordal and hexachordal structure of Du is touched on by Babbitt
([1976b] 2003). He notes the distinction between the opening and closing
hexachord and describes the relationship as instantiating a kind of “progres-
sion.” In his telling, the arrival of (037) at the beginning of “Traum” is the
moment at which the closing hexachord becomes “ ‘the’ derived hexachord”

siegst Gott’ can be read as: ‘You, oh my God, are victorious’; ‘You [i.e. a second (human) person] are
victorious after the manner of a God’; ‘You [i.e. a second (human) person] are victorious over God’ or
‘You [referring back to the Devil] are victorious over or after the manner of God.’ ” Any translation,
including the one given here, necessarily conflates some of these ambiguities and thereby loses an es-
sential aspect of Stramm’s original.

17 See Rahn 1976; Mead 1994, 107–​12; Morris and Alegant 1988, 89–​99; and Sullivan 2005 for

more on Babbitt’s Du.

18 Figure 5.9a does not include <−2, −1> (013), shown in the second series in Figure 5.9b, because

that trichord ordering is the retrograde of <+1, +2> (013), introduced already in the initial series.
19 Facilitating the perception of the contrast between these series is the great difference in sound

between the chromatic and diatonic hexachords. This qualitative difference is encapsulated by their
position as generic prototypes in Quinn 2007.
144 Thinking In and About Music

Figure 5.9a. Array-​generating trichord orderings in Du

Figure 5.9b. The two series of Du, as presented in the vocal part in mm. 1–​2
and mm. 104–​5. Song Cycle. Copyright © 1951 Mobart Music Publications.
Copyright © renewed. All Rights Reserved. Used by permission of European
American Music Distributors Company, sole U.S. and Canadian agent for
Mobart Music Publications.

of the piece—​although as will be discussed, (037) starts to appear on the

surface in “Begegnung,” the song that precedes “Traum.” Before that (and,
probably—​it is unclear—​only before that), the A-​type hexachord is described
as the “ ‘tonic’ ” (353, scare quotes original).
Babbitt posits a process that manifests in the progression between the
hexachords. He notes that the opening and closing hexachords can be
transformed into each other’s set classes if either discrete trichord is trans-
posed by a tritone. This is both true and significant: the fact that these se-
ries, although distinct in terms of their total content and interior segmental
trichords, are composed of the same discrete trichords—​valuable currency
in a piece with trichordal arrays—​has profound implications for the piece’s
Figure 5.10. Linear hexachords in the arrays for Du. “A,” “B,” “C,” and “E” refer to the familiar “types” of all-​combinatorial
hexachords.
Figure 5.11. Trichords in the arrays for Du
 Poetic Form and Psychological Portraiture 147

array structure. This relationship is prefigured, Babbitt explains, by the con-

stant tritone transpositions within the work’s arrays. While the statement
about the relationship between the opening and closing hexachords is ac-
curate, I am less inclined to attribute it to these local transpositions. Any
trichordal array constructed from A-​type, B-​type, or C-​type hexachords
necessarily contains continual tritone transpositions, since those
hexachords transpose into their complement under T6, but none of Babbitt’s
other trichordal array pieces contain multiple distinct series traceable to
that fact.
The presence of two distinct series is an unusual and problematic depar-
ture from Babbitt’s typical practice. In an aesthetic so thoroughly oriented
around hierarchical development from a unitary source, the bifurcated
structure of Du dissonates markedly. There is a disunity at the heart of the
work—​a tension between opposing forces. This disunity mirrors the sepa-
ration between Ich and Du, with the opening, A-​type series representing the
terror of Ich’s waking life, and the C-​type series representing Du and motion
away from Ich. It will also represent Ich’s dream, as that too is distant from
Ich’s conscious reality.
The cycle begins with Ich and the A-​type series, although from the very be-
ginning, the C-​type hexachord and certain subsets of that hexachord already
suggest development away from the initial series. As the cycle develops, no-
tably in “Allmacht” and “Traum,” the focus moves away from a representation
of Ich, a shift that corresponds with the transition toward material devel-
oped from the second, C-​type series. The concluding song, “Schwermut,”
achieves a measure of synthesis: it both presents the second series and, as
“Wiedersehen” did with the first series, intermixes it with the contrasting (in
this case, A-​type) hexachord.
The five trichordal set classes used in Du’s arrays, and that nearly satu-
rate the piece’s surface, present a nuanced semiotic system, relating in var-
ious ways to the A-​type and C-​type series and, thus, to the poetic content
represented by those hexachords.20 The trichord (014) is both a subset and
a generator of the A-​type hexachord but is not a subset of the C-​type hexa-
chord. Conversely, (037) is a subset and generator of the C-​type hexachord
but is not a subset of the A-​type hexachord. The trichords (013) and (025)
are each subsets of both hexachords, but (013) can generate the A-​type

20 Babbitt (1976b) 2003, 353 refers to the “disparate generative capacities” of the work’s trichords—​

this is a special feature of both of Du’s series, a fact that may explain their use in this work.
148 Thinking In and About Music

hexachord but not the C-​type, and (025) can generate the C-​type hexa-
chord but not the A-​type. Thus, (014) and (037) are strongly representative
of the A-​type and C-​type, respectively, and (013) and (025), as subsets of
both hexachords but generators of only one, are more weakly representa-
tive of the hexachords that they can generate. The set class (015) is neutral.
It is a subset of both hexachords but a generator of neither. However, the
two orderings with which (015) appears in Du’s arrays—​<+5, −4> and <−1,
−4>—​are clearly referential, in that <+5, −4> is the ordering that appears
in the opening, A-​type series and <−1, −4> appears in the closing, C-​type
series. The distinction of ordering means, for instance, that the beginning
of “Allmacht,” the fifth song (shown in Figure 5.15), can be considered
forward-​looking and representative of the C-​type series, even though its
(015) trichords and B-​type hexachords are familiar by this point in the cycle.
By way of this semiotic system, the trichords of the piece’s arrays and sur-
face refer not to a single generative series, but in varying degrees to one series
or the other—​or ambiguously to both. In this way, surface reference to the
series, rather than being a unifying technique relating surface elements to
each other by way of a single hierarchical source as in Babbitt’s other works,
contributes to the sense of destabilization and disorientation so vivid in
Stramm’s poetry.
Du begins with “Wiedersehen.” The ambiguities inherent in the title—​is
this an expression of farewell (Auf Wiedersehen!), a reunion, or, literally, a
seeing again?—​are played out in the poem. Ich sees Du, remarking on her
stride (“Schreiten”—​note the rhyme with schreien, to scream; the final song
makes this near-​homonymity explicit). He looks at her, although she appar-
ently fails to acknowledge him (“In Schauen stirbt der Blick” [“In looking,
the glance dies”]). Then, in the pivotal action of the brief poem, she turns
away: “Du /​Wendest /​Fort!” (“You /​Turn /​Away!”). Having lost her, Ich
concludes the poem in aphoristic mystery: “Den Raum umwirbt die Zeit!”
(“Space is wooed by time!”).
“Wiedersehen” is an account of Ich’s experience, and its most salient hex-
achord is the A-​type hexachord in the vocal part. But the C-​type hexachord
lurks in the two lower registers of the piano. The basic scenario is focused on
Ich, but Du is also present, if at a distance. The surface organization of the
song is organized such that each aggregate presents a derived series com-
posed of four iterations of a single trichord ordering (see Mead 1994, 108–​
10)—​drawn from the orderings presented in the A-​type series. Figure 5.12
presents the text of the poem and the trichords used to set the various words.
 Poetic Form and Psychological Portraiture 149

Figure 5.12. Trichords used in “Wiedersehen.” “Array trichords” are the linear
trichords of the array. “Surface trichords” are the trichords into which the
surface is partitioned using derived series.

Of the eight aggregates in the song, the first four present, one aggregate at
a time, the four trichord orderings contained within the A-​type series. The
fifth through eighth aggregates reverse this pattern, retrograding the same
sequence of surface trichords. Although the surface of the entire cycle is al-
most saturated with references to array trichords, the successive presentation
of neatly derived series in “Wiedersehen” is unique.
The critical event of “Wiedersehen” is when Du turns away: “Du /​Wendest
/​Fort!” The visual arrangement of the poem emphasizes this moment, with
every word in its own line. This is the only action in the poem, and perhaps
the moment to which the title of the poem refers: if one interprets the word
“Wiedersehen” as an expression of farewell, this is what would justify it.
Appearing within the first poem set in Babbitt’s cycle, it gains additional sig-
nificance. It is the first time we understand the position of Ich in relation to
Du. Until this point, we have been watching him observe her, but we learn
here that he is observing her, now and probably always, at a distance—​and
the distance is increasing. Moreover, this is the moment in which we first
hear Du referred to directly, by the pronoun that is also the title of the cycle.
Babbitt’s setting of the lines and their surroundings, shown in Figure 5.13,
stands out from the rest of “Wiedersehen,” particularly dynamically: this
passage presents the only 𝆑 or 𝆑𝆑 dynamics in the song and indeed among
the only 𝆑𝆑 indications in the cycle. The surface of this passage is arranged to
refer powerfully to the second, C-​type hexachord of the cycle: the aggregate
setting “Du /​Wendest /​Fort!” is derived using (025). That is, it is set using
the sole trichord contained within the opening, A-​type hexachord that can
150 Thinking In and About Music

Figure 5.13. Du, “Wiedersehen,” mm. 6–​9. Song Cycle. Copyright © 1951
Mobart Music Publications. Copyright © renewed. All Rights Reserved. Used by
permission of European American Music Distributors Company, sole U.S. and
Canadian agent for Mobart Music Publications.

generate the C-​type hexachord. This ability of the (025) trichord is actually
realized in this aggregate: as shown in Figure 5.13, the aggregate setting “Du
/​Wendest /​Fort!” is partitioned into C-​type hexachords, creating a startling
burst of diatonicism.
The setting of “Du /​Wendest /​Fort!” demonstrates the forces latent in the
eventually fissive structure of the cycle. It establishes the connection between
the C-​type hexachord, its trichordal subsets, and the world that is outside
of—​that has turned away from—​Ich. The remainder of the cycle will even-
tually realize the implications of this moment. However, in its immediate
context, it is still firmly contained within the world of the A-​type hexachord.
 Poetic Form and Psychological Portraiture 151

Indeed, in the measures just following, as shown in Figure 5.13, Ich’s materials
return. After the C-​type hexachords of “Du /​Wendest /​Fort!” an aggregate
of A-​type hexachords generated by (013) trichords appears. This harmonic
change not only starkly juxtaposes the generative hexachords of the cycle but
also coincides with a change in focus. In the last line of the poem, “Den Raum
umwirbt die Zeit!” (“Space is wooed by time!”), we have returned to the oc-
cult ruminations of Ich’s mind.
By the end of “Wiedersehen,” a number of important things about the cycle
and its construction have been revealed. The vocal part has introduced us to
the first, A-​type series, whose segmental trichords will generate the arrays
of the first interlude and the next three songs. Although those segmental
trichords have not yet been realized as arrays, all four of them have been used
to realize derived series, on the surface of each aggregate, that are much like the
derived series that will form the lines of the trichordal arrays. It has also been
determined that, although the most prominent hexachord in “Wiedersehen”
is the A-​type hexachord in the voice, the discrete trichords of the vocal part
can be rearranged into C-​type hexachords, as indeed they are in the lower two
registers of the piano. This alternative hexachord is, then, established by the
most striking moment in the song—​“Du /​Wendest /​Fort!”—​as associated
with Du and with a motion away from Ich. And just as Du turns away from Ich,
the cycle as a whole will eventually turn away from the A-​type series.
“Wiedersehen” is followed by one of two interludes in the piece; the other
occurs, symmetrically, just before the last song in the cycle (the first interlude
is shown in Chapter 7, in Figure 7.6). These interludes have a rhetorical func-
tion: they set off the opening and concluding songs, which present the piece’s
two series, from the interior songs that are derived from those series. Both
interludes, as well as the central song “Verzweifelt,” have an unusual array
structure: three simultaneous lines arranged in (43) tetrachordal aggregates.
Each line is trichordally derived using trichords and hexachords familiar
from the rest of the cycle, with a resulting mismatch between trichord and
aggregate boundaries (see Mead 1994, 111). The change in array structure
may be taken to heighten the rhetorical break the interludes accomplish.21

21 The interludes’ tetrachordally combinatorial yet trichordally derived structure might also

be attributed to more practical concerns: as a means to simultaneously preserve the piano’s three
registers, to create aggregates without the voice part, and to use the trichordal derivation that also
characterizes the interior of the cycle. Nonetheless, the piano could have simply absorbed the voice’s
line in the counterpoint, expanding to four registers. This is basically what happens in the interlude of
A Solo Requiem, as noted in Dubiel 1991, 107.
152 Thinking In and About Music

The first interlude’s lines are the first trichordally derived lines in the cycle,
quickly presenting counterpoint based on all four trichords of the A-​type se-
ries. These are arranged into B-​type and E-​type hexachords, which, in com-
bination with the A-​type and C-​type hexachords presented in the first song,
fill out the cycle’s complete vocabulary of linear hexachords. (Although the
array is constructed using tetrachordal aggregates, the discrete hexachords
of each line are still all-​combinatorial hexachords.) The result is that the first
interlude might be described as introductory, preparing the trichordal and
hexachordal content of the songs that follow.
“Wankelmut” presents Ich at his most hysterical. The poem opens with Ich
seeking Du, expressing a desire to catch and keep her. Du becomes manifold,
infinitely so, in time (“immer” [“always”]) and in space (“Allwege” [“every-
where”]). Ich describes his increasing confusion by compounding iterations
of the word wirr (“confused”). The first iteration he speaks is the homophone
with the pronoun wir (“we”); one can imagine Ich seeking wir, perhaps hope-
fully describing his state in a universe in which he is surrounded by Du, and
finding himself instead wirr, confused. At this point in the cycle, confusion,
and not the unity denoted by wir, reigns. His monologue returns to Du, either
as a result of, or in spite of, his confusion (“Durch /​Die Wirrnis /​Du /​Dich”
[“Through /​The confusion /​You /​Yourself ”]), but the poem ends—​perhaps
surprisingly, as the exclamation point suggests—​with Ich. The full shape is
thus something of a palindrome, bracketed by Ich. “Wankelmut,” more than
any other song in Du, is a presentation of Ich’s frenzied mental state.
The array structure of “Wankelmut” is heavily focused on materials already
associated with Ich. As can be seen in Figure 5.11, the song begins and ends
with arrays built from (014), the unique trichord that is a segmental trichord
of the A-​type series and is not even a subset of the C-​type hexachord. Just as
the poem is bracketed by Ich, the song is bracketed by the trichord most as-
sociated with Ich. The other trichords with which the arrays of “Wankelmut”
are constructed are (015) and (013). That is, of the four trichords available in
the A-​type series, the only one not present is (025), the one trichord in that
series that can generate the C-​type hexachord. Both of the final arrays in the
song, in the moments in which Ich’s hysteria reaches its highest pitch, are
built using A-​type linear hexachords, in the only section of the cycle to use
A-​type linear hexachords exclusively.
The four arrays are roughly coordinated with significant moments in the
poem. The first array, built from (014), extends through the first three lines of
the poem, coextensive with the opening passage in which Ich refers to himself.
 Poetic Form and Psychological Portraiture 153

The array built from (015) begins at “Und fasse Du” (“And grasp You”), as the
subject changes to Du, and continues through “Und Du und Du und Du.”
That is, (014), the trichord most associated with the A-​type hexachord, is as-
sociated with lines discussing Ich, while (015)—​which is a subset of the C-​type
hexachord and, unlike (013) and (014), is not a generator of the A-​type hexa-
chord, and thus is relatively close to the C-​type series—​is associated with lines
discussing Du and Ich’s losing himself in Du. The array built from (013) begins
with the line “Viel tausend Du” (“Many thousand You”). This is the first line in
which we grasp Ich’s vision of a manifold, universal Du. The previous line, “Und
Du und Du und Du,” in retrospect is also referring to a plural, wankelmütig
Du, but upon first encounter its repetitions might just be read as emphatic.
Therefore, the line “Viel tausend Du” is the first clear indication of Ich’s mad vi-
sion and thus is a clear return to Ich’s wild mental state—​a fact that may justify
the use of (013), which is not only a generator of the A-​type hexachord but is
in fact combined into A-​type linear hexachords in this passage. The beginning
of the song’s final array, built from (014), does not come at such a clear textual
juncture, beginning with “Wirrer” and thus arriving in the middle of the pas-
sage that had begun two lines earlier, with “Wirr.” Nonetheless, this final array
encompasses most of the final section of the poem, focused on Ich’s confusion.
The title of the third song, “Begegnung” (“Meeting” or “Encounter”), is
surely ironic. As in “Wiedersehen,” Du is within sight of Ich. He observes
her intently. But there is no indication that they interact, or that Du acknow-
ledges Ich. In the first line, “Dein Gehen lächelt in mich über” (“Your walking
smiles over into me”), it is not Du who smiles over into Ich, but merely her
walking. As we have learned in “Wiedersehen,” this motion is away from
Ich, a fact that may contribute to Ich’s feelings about it: it tears his heart. Ich
comments extensively on her movement and at one point grabs at her blindly,
a foolish attempt that earns the scorn of a hostile universe (“Die Sonne lacht!”
[“The sun laughs!”]). But Ich is lamed by his “dull hesitation” (“Blödes Zagen
lahmet fort”), remaining bereft.
“Begegnung” is a turning point in the cycle. Having failed to approach Du
during this encounter, Ich will not try again. In fact, his desperate grasping
is the cycle’s last action of any type. The succeeding poems present a static
image (“Verzweifelt”), a discourse on Du’s powers (“Allmacht”), a dream
(“Traum”), and a final meditation (“Schwermut”), but the cycle’s portrayal of
Ich’s experience and his attempts to connect with Du has ended.
The song’s transitional nature is reflected in its array structure. Like
“Wankelmut,” “Begegnung” uses three out of four trichords from the first
154 Thinking In and About Music

series, but whereas “Wankelmut” focuses on (014), the distinguishing tri-

chord of the A-​type hexachord, “Begegnung” omits it, using only three
trichords that are subsets of the C-​type hexachord. The song begins with
two arrays that each simultaneously present linear E-​ type and C-​ type
hexachords. The C-​type hexachords suggest Du’s presence, although the fact
that the C-​type hexachords are never heard alone indicates that, even though
Du is nearby, she is not at arm’s length (as Ich’s failed attempt to grasp her
illustrates). The second half of the song is focused on the B-​type hexachord,
the hexachord that Babbitt described as transitioning between the A-​type
and C-​type hexachords that begin and end the cycle ([1976b] 2003, 353).22
The surface of “Begegnung” presents the four trichords of the first series in
a variety of combinations, generally not in a pattern that obviously interacts
with text setting. However, one detail is notable. As the final song derived
from the A-​type series, “Begegnung” is also the final song whose surface
details consistently reference that series. It therefore presents the cycle’s final
(014) trichords. As shown in Figure 5.14, these occur with the text “Blödes
Zahmet lahmet fort.” The (014) trichords, the trichords most associated with
Ich, are interleaved with (025), the first-​series trichord most associated with
the second series, with Du, and with motion away from Ich. This captures the
moment described in the text: Ich is present but is unable to approach Du,
finalizing the separation previously described by “Du /​Wendest /​Fort!” and
set using the same (025) trichord.
“Verzweifelt” stands at the center of Du. The poem presents stone and glass
as opposing images. Stone and glass are fixed, cold, and hard—​apt metaphors
for Du’s antagonistic metaphysics. The stone is identified as Ich; the ra-
diant glass, far off (“Weit”), is Du. Yet as hard as stone and glass are, neither
survives the poem intact: the stone is smashed (“schmettert”), the glass is
splintered (“grant”). The state Ich is trapped in is represented by the poem’s
third line: “Die Zeiten stehn” (“The times stand still”). The “times”—​plural,
perhaps evoking the chaotic multiplicity of “Wankelmut”—​are still. The
stone eternally shatters; the glass eternally splinters. At the heart of Du, Ich
and Du—​whose pronominal identities were never concrete—​are depicted as
disintegrating.
Coming after the completion of the songs developing the first series and
before the songs developing the second, “Verzweifelt” is the still point around

22 The B-​
type hexachord can be considered equidistant from both the A-​type and C-​type
hexachords, as discussed in Babbitt 1987, 48–​49.
 Poetic Form and Psychological Portraiture 155

Figure 5.14. Du, “Begegnung,” mm. 45–​46. M = <−4, +1> (014). P = <−3, +5>
(025). Song Cycle. Copyright © 1951 Mobart Music Publications. Copyright
© renewed. All Rights Reserved. Used by permission of European American
Music Distributors Company, sole U.S. and Canadian agent for Mobart Music
Publications.

which the rest of the cycle turns. Just as the interludes use tetrachordal
aggregates to set off the opening and closing songs, which present series, from
the interior songs derived from those series, “Verzweifelt” uses tetrachordal
aggregates to delineate the two halves of the cycle. Its array is built entirely
from (013) and (025), the two discrete trichords shared by both series, and
its surface groupings also highlight those two trichords. Although (013) and
(025) have elsewhere in the cycle respectively represented the A-​type and C-​
type hexachords, on account of their ability to generate those hexachords,
their presentation here—​together, in equal measure, and without other in-
terfering trichords—​suggests a pause in the piece’s progression. “Verzweifelt”
explores neither Ich’s consciousness nor represents Du or Du’s motion away
from Ich; it simply presents the two characters as frozen, immobile. The si-
multaneous presence of (013) and (025) facilitates this static image.
The array construction of “Verzweifelt” is asymmetrical. Although this
itself is not unusual, its role as the cycle’s centerpiece has implications for
the whole. Morris and Alegant (1988, 95–​98) propose an RM symmetry
extending across the entire cycle.23 This symmetry holds, as far as the

23 The M transformation exchanges the A-​type and C-​type hexachords while preserving all other

hexachords used in the cycle. This symmetry only affects harmonic content, not serial order; the two
distinct series in Du cannot be transformed into each other by M.
156 Thinking In and About Music

harmonic content of the cycle’s trichords and hexachords is concerned,

with three exceptions: the two interludes and this song. While the voice and
piano exchange their trichords, the song’s two arrays are built from distinct
hexachords: the first from B-​type hexachords, the second from A-​and C-​type
hexachords simultaneously. (As in the interludes, the lines of “Verzweifelt”
present all-​combinatorial hexachords in addition to the tetrachords neces-
sary for its (43) aggregates.) The boundary between the arrays falls at the line
“Die Zeiten stehn.” The deployment of the hexachords follows a disambigu-
ation in the poem. The opening lines introduce the images of the stone and
the glass but do not immediately associate them with Ich and Du; accord-
ingly, a hexachord not closely associated with either character is appropriate.
The second half of the poem clarifies the link between the images and the
characters, and therefore the trichordal materials of the first half of the song
are rearranged to form A-​type and C-​type hexachords simultaneously—​the
first time they have appeared simultaneously since the opening of the cycle—​
as a representation of the dual images. Despite this balancing of the A-​type
and C-​type hexachords, however, the vocal part—​clearly the most promi-
nent part in this song cycle—​indicates the direction the cycle is heading. As
Babbitt notes ([1976b] 2003, 353), over the course of “Verzweifelt” the voice
progresses from the B-​type hexachord to the C-​type hexachord that will in-
creasingly become the cycle’s focus.
“Allmacht” launches the second half of the cycle, a fact witnessed in its first
three notes, which present (015) ordered <−1, −4>, the ordering in which
that trichord appears in the second series. Although we have not yet heard
the second series, the contrast with previous (015)s, which are consistently
ordered <+5, −4>, is readily audible. The poem celebrates Du’s power, al-
ternating lines that present challenges with lines that discuss Du’s ability to
overcome those challenges. While Ich is probably still the narrator of this
poem—​presumably the extraordinary powers (“Du siegst Gott!” [“You are
victorious God!”]) ascribed to Du are a function of his image of her—​his
character is remarkably absent. There are no first-​person pronouns, a situ-
ation nearly unique in the cycle thus far, nor is there any suggestion of Ich’s
presence. (There are also no first-​person pronouns in “Wiedersehen,” but
“Wiedersehen” clearly depicts a sighting of Du from the perspective of Ich.)
Ich’s absence is reflected in the transition to trichords derived from the
second, C-​type series. Notably, however, the second half of the song presents
A-​type hexachords, first in the voice part and then in the lower registers of the
piano. These hexachords are introduced with the text “Wahn Verzweiflung”
 Poetic Form and Psychological Portraiture 157

(“Madness despair”). This corresponds to a shift in the text. Throughout,

odd-​numbered lines of the poem have presented a series of challenges, but
there is a shift at this point from more general problems to conditions that
might be described as Ich’s reaction to those concerns. For instance, two lines
before “Wahn Verzweiflung” is the line “Falsch und Tücke” (“falsehood and
malice”). In the context of the cycle, one may well understand the delusion
and despair to be Ich’s, but the falsity and malice are presumably conditions
afflicting him (and perhaps the rest of the universe), not descriptions of him.
Accordingly, the use of (013) trichords and A-​type hexachords to set “Wahn
Verzweiflung” and “Tod und Elend” (“death and misery”) is appropriate. The
last two pairs of lines return to more general concerns. A-​type hexachords
continue in this final section of “Allmacht,” but less prominently, in the lower
registers of the piano.
As indicated in Figure 5.15, “Allmacht” includes, for the first time in the
cycle, the (037) trichords characteristic of the C-​type hexachord. Although in
“Allmacht” they are not yet the linear trichords of the array, harmonic triads
appear with some frequency on the surface. The second syllable of “Fürchten,”
with its   𝆑 C-​major triad, is the first (037) simultaneity in the cycle—​a jarring
indication that the harmonic world of the cycle is changing.24
It would be imprecise to describe Du as “free verse,” for many of its poems
slip into passages of regularity. Nonetheless, “Allmacht” is the only poem in
all of Stramm’s volume that is consistently metric in foot, line, and stanza: it
uses trochaic dimeter arranged into non-​rhyming quatrains, such that the
last line of each quatrain is catalectic and marked with an exclamation point.
The effect is that the recitation of Du’s traits carries the energy of a hyp-
notic mantra. The placement of the poem in Babbitt’s cycle is significant: it
follows “Verzweift,” which has virtually no metric organization, and precedes
“Traum,” which is also completely trochaic (although not in regular lines
or stanzas), and “Schwermut,” which is trochaic until its fateful final three
lines. Their metric organization unites these three songs as the second half of
Du—​coinciding with and supporting their shared serial organization. In all
three settings (and particularly “Traum”), the trochees are brought out musi-
cally, most often by granting the stressed syllable a longer duration than the
following unstressed syllable. In the setting of “Allmacht,” Babbitt’s setting is
unusually rigid from the perspective of lineation, perhaps reflecting the fact

24 Babbitt ([1976b] 2003, 353) also draws attention to the setting of “Fürchten,” although on dif-

ferent grounds: he notes its minimal pitch-​class overlap with the opening series.
158 Thinking In and About Music

Figure 5.15. Du, “Allmacht,” mm. 57–​61, with (037)s indicated. Song Cycle.
Copyright © 1951 Mobart Music Publications. Copyright © renewed. All Rights
Reserved. Used by permission of European American Music Distributors
Company, sole U.S. and Canadian agent for Mobart Music Publications.

that the poem’s line organization is uniquely metric. Each line is set by one
aggregate, and thus one vocal trichord sets each line. Accordingly, each line,
with the four syllables of a trochaic dimeter, has one repeated note, except
for the catalectic final line of each quatrain. Except for the last, these cata-
lectic lines each end with a rising contour and increasing or stable dynamics,
sharply setting their exclamation point. But the last line, “Du siegst Gott!”
(“You are victorious God!”), in perhaps the most dramatic moment in the
cycle, rises on “siegst” to   𝆑𝆑 B♭5, the highest note in the vocal part. In awe,
Ich observes Du’s divine might. Following this outburst, the voice descends
and both the voice and piano are softened, transitioning into a gentler realm.
From its title, or even just its first few notes, it is clear: “Traum” stands
apart from the rest of Du. The dream—​presumably Ich’s—​is filled with
 Poetic Form and Psychological Portraiture 159

natural imagery: bushes, stars, blossoms, scents, sprays, and showers. It is

not a pleasant dream, particularly toward the end: the image of winds that
“schnellen prellen schwellen” (“dart bounce swell”) are more suggestive
of a tempest than of a gentle breeze, and the final lines, with tearing sheets
(“Tücher reißen”) and falling into the night (“Fallen schrickt in tiefe Nacht”),
seem to describe waking from a nightmare with a start. But Ich’s horrible re-
ality has been, at least temporarily, escaped. Indeed, the Ich-​Du relationship
seems completely absent from this poem. Uniquely among the seven poems
Babbitt selected for Du, there are no pronouns in “Traum” whatsoever.
“Traum” is the song in Du most thoroughly associated with the second se-
ries and its C-​type hexachord. In its opening, shown in Figure 5.16, its array
is derived from members of (037), the major or minor triad—​the unique tri-
chord in the second series that is not even a subset of the first series. The triads
are compounded, at the beginning, into diatonic, C-​type hexachords. There
are a number of (037) simultaneities in “Allmacht,” as discussed previously,
but they appear briefly and within a context generally focused on other sonor-
ities. The lingering on (037) in “Traum,” and its isolated presentation at the
beginning of the song, marks a sharp contrast with the remainder of the cycle.
The dream of “Traum” draws Ich out of the world that has tormented
him. Accordingly, the song presents the apotheosis of material that has been

Figure 5.16. Du, “Traum,” mm. 78–​80, with array components and poetic
scansion indicated. Song Cycle. Copyright © 1951 Mobart Music Publications.
Copyright © renewed. All Rights Reserved. Used by permission of European
American Music Distributors Company, sole U.S. and Canadian agent for
Mobart Music Publications.
160 Thinking In and About Music

associated throughout the cycle with states outside of Ich’s mind. Most of the
situations set using the C-​type hexachord or its materials have been linked to
Du or Du’s motion away from Ich. But in “Traum,” the insistent use of mate-
rial characteristic of the second series indicates that the series and its corre-
sponding set classes are linked to a class of things, a class encompassing the
world generally apart from Ich’s waking terror, of which Du is only an example.
“Traum” continues the trochaic meter of “Allmacht”—​indeed, even more
insistently, as it contains no catalexis or other variation until its final word.
Every trochee is set so that the stressed syllable is sustained for longer than
the unstressed one. The result—​along with the triadic harmonic language
that begins the song and aspects of the song’s contour and dynamics—​is a
detached, wistful affect. This supports the song’s sense of distance from the
remainder of the cycle, and from the weight of Ich’s longing.
The vocal part of the concluding song, “Schwermut,” presents the second
series. In this sense, it represents the clarification, the final disambiguation,
of the second half of the cycle (see also Mead 1994, 110). As with the first
song, the discrete (013) and (025) trichords that form every line are arranged
such that C-​type hexachords constantly coexist with A-​type hexachords,
which persist in the lower two registers of the piano. This suggests that, as
much as the cycle has represented a progression toward the second series,
this final song also contains a touch of synthesis.
Synthesis can also be found at the end of the cycle’s text. “Sterben wächst /​
Das Kommen /​Schreit! /​Tief /​Stummen /​Wir.” (“Dying grows /​The coming
/​Screams! /​Deeply /​Grow mute /​We.”) The universe’s hostility has not
abated. But Ich, it seems, has learned his place in it and in respect to Du.
“Schwermut” is melancholy, as its title suggests, but contains only a little of
the hysteria that had haunted some of the earlier poems. The final word of
the poem, and thus of the cycle, is “wir.” This is the only first-​person plural
pronoun in the entire cycle—​in fact, it is the only “wir” in all thirty-​one
poems in Stramm’s Du. In context, the pronoun can hardly be taken as indi-
cating unity or reconciliation. But its final three lines are, at last, a precise
description, from Ich, of his relationship with Du. Moreover, the final lines
recall “Wiedersehen,” the first song in the cycle: “Tief /​Stummen /​Wir” is
a metric echo of “Du /​Wendest /​Fort!”; these are the only two sentences
with a monosyllable /​trochee /​monosyllable pattern, each word on its own
line, in Stramm’s Du. The connection between these passages is reinforced by
Babbitt’s music, for the aggregate underlying “Tief /​Stummen /​Wir” is de-
rived from the same (025) mosaic as the aggregate underlying “Du /​Wendest /​
 Poetic Form and Psychological Portraiture 161

Fort!” (now realized contrapuntally, rather than temporally). As a result

of this connection, we can interpret the cycle’s final lines as presenting the
ultimate conclusion of Du’s turning away. Ich is left alone, in melancholic
silence. At this moment, the trochaic momentum of the second half of the
cycle finally breaks, and with descending vocal contour and a suddenly soft-
ened dynamic, Du ends in sober contemplation.
The use in Du of dual, incommensurable series is a departure from Babbitt’s
regular practice, even a divergence from principles he articulated about cog-
nitive efficiency and the function of a series. Recall Babbitt’s comments about
polytonality: it is “self-​contradictory” (Babbitt [1949] 2003, 3), apparently
because a piece cannot be perceived as a single whole if its components are
separately referable to different sources. Trichordal derivation is a technique
that enables links between elaborations (the derived series) and their source
(the underlying series), thus fostering the kind of cognitive subsumption
Babbitt felt necessary for musical cognition. But in response to the text of Du,
Babbitt turns the possibilities of trichordal derivation toward realizing the
poem’s dichotomy. A technique once intended to foster unity is used here, in-
stead, to create a vividly, precisely, and poetically articulated disunity.

Two Sonnets

Babbitt’s next two texted works were initially conceived at around the same
time: Two Sonnets was composed in 1955, the same year Babbitt abandoned
his first attempt at setting Vision and Prayer. These works are among his most
thoroughgoing attempts to project poetic form. Indeed, Babbitt (2005) him-
self claimed—​by way of “confession”—​that his interest in these texts was
primarily on account of their sounds.25 Both texts are religious meditations

25 It is unclear in the cited statement, which appears in a casually worded response within an inter-

view, whether Babbitt is referring to Two Sonnets and Vision and Prayer alone. He typically showed
no inclination to talk about poetic meaning in any of his settings or those of any other composer, but
this may reflect discursive preference rather than genuine lack of interest in poetic meaning. A quip
relayed by Peter Quantrill (2007) illuminates Babbitt’s public posture toward poetic meaning: “A stu-
dent recently asked Milton Babbitt what he made of the plot of Moses und Aron. ‘Oh I don’t know, I’m
not really a plot person,’ he replied. ‘Boy meets Girl, Moses meets Aron . . .’ ” But the evidence of his
compositions suggests a different attitude, as do the archival materials for his texted works, which
typically include photocopies of excerpts from scholarly interpretations of the poems he plans to
set (his preparatory study for setting text is mentioned in Babbitt 1979b, 45). His personal library
includes a number of volumes of literary analysis, suggesting a curiosity in poetic and literary in-
terpretation even apart from the texts he set (he was particularly interested in scholarship on James
Joyce).
162 Thinking In and About Music

rather than narratives or character studies; accordingly, there are few

aspects of their meanings that lend themselves to techniques of psycholog-
ical portraiture. At the least, I have found little in either, beyond the occa-
sional madrigalism, that is suggestive of a musical representation of poetic
meaning. But both settings are rich with musical analogues for patterns of
poetic form and sound.
Two Sonnets sets a pair of poems by Gerard Manley Hopkins that, until
the final lines of the second sonnet, remain apocalyptically grim. Fittingly,
Babbitt orchestrates the work with a low quartet: baritone, cello, viola, and
clarinet. The instrumental trio is strictly accompanimental: there are no
interludes, and the ensemble remains in a background role throughout. Its
most notable contributions to text-​music relations are changes of instrumen-
tation, texture, and contrapuntal technique that articulate key moments in
the text.
The sonnets are reprinted with annotations in Figure 5.17. From the per-
spective of rhyme scheme, these are straightforward Petrarchan sonnets—​
although as discussed in Chapter 3, “That Nature is a Heraclitean Fire and of
the Comfort of the Resurrection” ends with a series of couplets and shorter,
indented burden lines appended to the sonnet form (a variation known as
the “caudate sonnet”). But the rhyme scheme may be the only conventional

Figure 5.17a. “Spelt from Sybil’s Leaves.” A, B, C, and D indicate both the poetic
end-​rhyme scheme and the musical technique used to set each line.
 Poetic Form and Psychological Portraiture 163

Figure 5.17b. “That Nature is a Heraclitean Fire and of the Comfort of the
Resurrection.” Letters indicate both the poetic end-​rhyme scheme and the
musical technique used to set each line.
164 Thinking In and About Music

aspect of these poems. Both sonnets, for instance, use Hopkins’s character-
istic “sprung rhythm.” In this technique, there are a fixed number of stressed
syllables but a variable number of total syllables. “Spelt from Sybil’s Leaves”
uses an unusually lengthy line of eight stressed syllables, divided by a cae-
sura into two hemistichs each containing four stressed syllables. (Hopkins
notates caesuras and many stress marks to clarify this pattern.) “That Nature
is a Heraclitean Fire” uses, except for its burden lines, a slightly shorter
line of six stresses also bisected by a notated caesura. Stressed syllables are
separated by zero to four syllables, forming a range of different feet. It is un-
common for more than three consecutive feet to have the same number of
syllables, even while the total number of stresses per line remains fixed.
An interplay between metric regularity and flexibility thus characterizes
these poems. In an interview following a performance of Two Sonnets,
Babbitt lauds this aspect of Hopkins’s approach:

Who can resist Hopkins? Hopkins himself said so much about his concep-
tion of the rhythm and the metric and his so-​called sprung rhythm, which
just aches to be set to music, the notion of a duple division of the poetic
[line], and then within that [line], any number of syllables are possible,
but the two [hemistichs] must always be of course of the same, well, if you
wish, measurable length, and you can do this in music, you can have a duple
division and you can have these two [hemistichs] which are of the same
measurable, playable length. His whole conception of sprung rhythm was
so . . . plastic and, in every sense of the word, musical. (Babbitt 2005)26

Babbitt’s setting reflects this plastic variability within a fixed framework. In

“Spelt from Sybil’s Leaves,” each hemistich spans two measures. The only
exceptions are when a line begins with an anacrusis; in those cases, Babbitt
begins the hemistich just before its allotted two-​measure span, such that the
first stressed syllable falls on a downbeat. The four stressed syllables of each
hemistich typically fall on beats one and three of  measures, although there
is some syncopation. In “That Nature is a Heraclitean Fire,” each hemistich
spans three measures, with most stressed syllables falling on downbeats.
Between these metrically regular stressed syllables, variable surface rhythms
set the uneven patterns of weak syllables. Such correspondence between

26 Babbitt uses the word “foot” in the bracketed spots in this quotation. I have replaced it with more

standard terms.
Figure 5.18. Vocal part of Two Sonnets, “Spelt from Sybil’s Leaves,” mm. 2–​17
166 Thinking In and About Music

poetic and musical meter is highly traditional—​a nineteenth-​century Lied

can be expected to work much the same way.27 Tempo and metric changes
smooth over the notational differences between the sonnets: the duration be-
tween stressed syllables in the setting of “Spelt from Sybil’s Leaves” is typically
2 seconds; in “That Nature is a Heraclitean Fire,” it is typically 1.6 seconds.
Another immediately striking aspect of these sonnets is that they are rife
with sonic parallelism far beyond the end rhymes: parallelisms of phonemes,
which are constantly repeated or reiterated under variation, and of stress
patterns. The variable length of feet allows for the creation of numerous,
complex associations and echoes, which Hopkins (and Babbitt) takes plen-
tiful advantage of. The opening quatrain of “Spelt from Sybil’s Leaves” exem-
plifies these traits, although the first line includes the only exception to the
poem’s metric pattern: its second half is missing one stressed syllable, an ab-
sence noted by Hopkins’s ellipsis. Babbitt’s setting of this quatrain is exam-
ined in Figures 5.18 and 5.19.
Babbitt projects all aspects of poetic form discussed here. The stressed
syllables throughout the first sonnet typically fall on beats one and three,
and thus each hemistich spans two measures and each line four—​excepting
anacruses, as can be seen in both halves of the third line (these are boxed in
Figure 5.18). But the  bars in this passage deviate from the overall regularity.
The impetus seems to be the rhyme scheme: the four lines projected make
an ABBA quatrain, and the brief elongation of lines 2 and 3 associates those
two rhyming lines. Heightening the connection, the ends of those lines are
associated not only metrically but also both rhythmically and, more closely,
intervallically. A major seventh leap ends each line, creating madrigalisms for
“night” and “height.” The rhyming first and fourth lines are also linked musi-
cally. As indicated in the example, a number of pitch classes in both lines fall
in corresponding metric positions, and there are rhythmic correspondences
between their third measures.
Figure 5.19 presents a focused look at the first two lines of the same sonnet,
with attention to moments in which Babbitt uses musical parallelisms to
bring out phonemic parallelisms. The five trochees are all set on repeated
notes. The very similar tetrasyllables “attunable” and “voluminous” are set
with identical contours and similar rhythms, both rising to and extending
their accented, assonant second syllables. And the dactyls that end the

27 I am grateful to Loretta Terrigno for discussions on this matter.

Figure 5.19. Phonemic parallelism in Two Sonnets, “Spelt from Sybil’s Leaves,” mm. 2–​9. The dashed arcs at the bottom of the example
connect near rhymes; the solid arc connects identities.
168 Thinking In and About Music

second line (“womb-​of-​all,” “home-​of-​all,” “hearse-​of-​all”) are set with par-

allel rhythms and repeated notes on the weak syllables.
One feature of the metric regularity of the setting is that it highlights
phonemic repetition among stressed syllables. Some instances of these
repetitions, such as the alliterative “earnest” and “earthless,” are obvious
enough, but Hopkins presents a remarkable range of slant rhymes, near
rhymes, and so forth among the stressed syllables of these sonnets. The chain
“evening,” “strains,” “time’s,” and “vast” that begins the second line provides
an example. Each accented syllable shares a vowel or a near-​rhyming vowel
pair with the words that surround it, but because of the use of diphthongs,
no two words precisely rhyme. These connections are notated using the
International Phonetic Alphabet at the bottom of Figure 5.19, but for readers
unfamiliar with the notation, I recommend simply reading the words aloud
and listening for degrees of similarity. For instance, the vowels of “strains”
and “time’s” are both diphthongs ending with /​ɪ/​. By setting each accented
syllable on strong beats, Babbitt facilitates their aural association—​one that
might be easy to miss otherwise.
Babbitt also projects large-​scale poetic form: the overall rhyme scheme
and the grouping of lines into couplets, quatrains, sestets, and octaves. As
detailed in Figure 5.17, each end rhyme in each sonnet is set with a different
serial technique—​typically, full statements of the series or trichordal arrays
developed from segmental trichords of that series. The series itself is notable
for a number of reasons. It is trichordally combinatorial, which facilitates
sections in which each member of the ensemble simultaneously presents a
series form. Additionally, each hexachord of the series is divided into two
equivalent trichords—​(013) and (015) respectively—​presented in two dis-
tinct orderings. This enables sections in which the full ensemble develops
one trichord class (for instance, (013) in the setting of the second line), but
the voice uses a different ordering from the instrumental ensemble, distin-
guishing its role as the piece’s soloist. Figure 5.21 demonstrates this technique.

Figure 5.20. Series forms used in “A” lines in “Spelt from Sybil’s Leaves”
 Poetic Form and Psychological Portraiture 169

The lines marked “A” in the rhyme scheme for “Spelt from Sybil’s Leaves”
are set using complete series forms, all of which share hexachordal content;
see Figure 5.20. (Since the piece uses B-​type hexachords, series related by
T6 have invariant hexachords.) The same series forms appear, in different
instruments, in the quatrain-​initiating lines (1 and 5); a different group of se-
ries forms are used in the quatrain-​ending lines (4 and 8). Quatrains are thus
articulated musically through serial means. Moreover, there are four total
lines with this rhyme, and Babbitt sets these lines such that each member
of the ensemble presents each ordering possibility once. This defines the
sonnet’s octave (its first eight lines) as a musical unit.28
The lines marked “B” in the first sonnet are set using trichordal arrays de-
rived from (013). These are ordinary, eight-​aggregate trichordal arrays, cy-
cling through the various possible combinations of trichords. Sustaining
the harmonic rhythm of the “A” lines, however, each poetic line continues
to span four aggregates, and thus each array underlies a full couplet. In so
doing, Babbitt provides a musical analogue for the rhyming couplet: the lines
are grouped into a larger framework.
The two principal parts of a sonnet are the octave and the sestet: its first
eight and concluding six lines. Babbitt uses distinct strategies to set off the
octave and sestet of “Spelt from Sibyl’s Leaves.”29 The statements of complete
series forms and arrays built from (013) are exclusive to the octave. In the
sestet, Babbitt’s scheme serves to distinguish the “C” and “D” lines from each
other (and from “A” and “B” lines), to support their grouping into three pairs,
and to define the sestet as a whole. The lines marked “C” use a combination
of orderings of (014) and (013), while those marked “D” use (015); see Figure
5.21, which depicts the arrays for lines 9 and 10. Yet as can be seen in the
figure, each consecutive pair of lines presents a complete trichordal array,
cycling through the partitions of four trichords. Despite their differing mate-
rial, this defines the pair as a complete musical unit. Furthermore, each “C”
line reduces the accompanying ensemble to a duet (clarinet-​cello in line 9),
and within each duet, one instrument is featured, presenting two lines of the

28 A further, more abstract means also defines the series forms shown in Figure 5.20 as a unit. If B♭

is taken as 0 and the series labels are interpreted as transformations (with, e.g., S0 = T0)—​that is, if
each series form is labeled by the transformation relating it to the work’s primary series, here called
St—​then Figure 5.20 would be an accurate and complete multiplication table. Moreover, those eight
transformations (T0, T6, RT0, RT6, T1I, T7I, RT1I, and RT7I) form a mathematical group.
29 The octave and sestet are also separated by a sustained perfect fifth in the accompaniment.

A similar sustained fifth precedes the octave, in m. 1 of the piece, which suggests the sonority has an
initiating rhetorical function in the work.
170 Thinking In and About Music

Figure 5.21. Array for lines 9 and 10 of “Spelt from Sybil’s Leaves”

array rather than one (clarinet in line 9). “D” lines, in contrast, use the full
trio. The full sestet is musically defined by the pattern of rotating ensembles
accompanying “C” lines—​over the course of the sestet, each of the three
duets is presented once, and each accompanying instrument is featured once.
The second sonnet falls broadly into two parts: a Petrarchan sonnet
form spanning the first fourteen lines, which describes how “Nature is a
Heraclitean Fire,” and a series of “codas” composed of couplets and burden
lines that describe the “Comfort of the Resurrection”; the pivot happens with
the proclamation “Enough!” in the middle of line 15. The techniques Babbitt
uses to delineate the components of the first fourteen lines are broadly similar
to those used in “Spelt from Sybil’s Leaves,” so for that portion of the sonnet,
I simply refer the reader to the annotations at the bottom of Figure 5.17b.
But the codas call forth a dramatic change in musical technique befit-
ting the change in subject matter. Indeed, the setting of the final lines can be
considered a musical coda for the whole of Two Sonnets. Both “Spelt from
Sibyl’s Leaves” and the first fourteen lines of “That Nature is a Heraclitean
Fire” depict the same general theme: the “enormous dark” of man’s worldly
state. It is only in the concluding codas that the poem moves toward the sal-
vation of the Resurrection. Simultaneously, the perspective changes to first-​
person sanctification: “I am all at once what Christ is.”30 By juxtaposing and
ordering the sonnets as he did, Babbitt lets the final couplets of “That Nature

30 For perceptive commentary on this transition, see Walliser 1977, 37.

 Poetic Form and Psychological Portraiture 171

is a Heraclitean Fire” lift the gloom that hovers over both poems until that
point. Musically the setting up to then uses nearly consistent (34) aggregates;
this partition is disturbed only by the rare instance of swapping. In the con-
cluding couplets, Babbitt switches to a partition novel to Two Sonnets, (623),
in which the voice contributes a full hexachord and each accompanying in-
strument contributes only a dyad to each aggregate. (One of these arrays is
depicted in Chapter 3, in Figure 3.9.)
A text-​setting motivation may account for the change of strategy. The ir-
regular partitioning facilitates a change of texture, permitting the vocalist to
emerge, in prophetic incantation, above a nearly static accompaniment. Text
setting may also justify the profusion of trichordal content noted in Figure
5.17b. In comprehensively summarizing the piece’s trichordal material, array

Figure 5.22. Two Sonnets, “That Nature is a Heraclitean Fire,” mm. 129–​33.
Score in C.
172 Thinking In and About Music

construction suggests that the Resurrection is the fruition of all that has
come before.
In Figure 5.17b, the burden lines are numbered as continuations of the
lines that precede them, rather than as independent lines of their own.
Babbitt’s music supports this construal. In setting the first three such lines,
Babbitt simply sustains or repeats the final notes used to set the preceding
line. These moments thus heighten the sense of incantation. The setting of
the final line, shown in Figure 5.22, mostly exemplifies this, but in apparent
search for a poignant finishing touch, Babbitt deviates slightly from simple
repetition. Two pitches have been reserved so that the voice may end Two
Sonnets with a leap to high F4. It is as though the protagonist of the song
is straining to believe that he, “poor potsherd,” has indeed been elevated,
through Christ, into “immortal diamond.”

Vision and Prayer

Vision and Prayer is a landmark in Babbitt’s output: his first work to combine
synthesizer and a live performer. The poem, by Dylan Thomas, contains twelve
stanzas. The first six are shaped like diamonds, with lines gradually adding and
then dropping syllables; the last six, inversely, are shaped like wings. Figure
5.23 gives two representative stanzas from each half of the work.
Thomas’s poem is complex and obscure,31 but some images come through
clearly. The vision with which the poem opens is of a birth, apparently the
birth of Christ, but perhaps, symbolically, the birth of religious awareness.
The observer of the birth is a lost, damned soul, one who encounters religious
mystery through observation of the birth but is simultaneously distanced
from it, “in the next room.”32 By the end of the sixth stanza, he reports his
death; he is apparently absorbed into the universe of the vision. The next five
stanzas are the prayer, a reverie on the fate of the lost, to which the observer
still belongs, although “not wholly.” Stanza 12 is an epilogue. The observer is
“found,” apparently having reclaimed his identity in religious ecstasy, and is

31 Babbitt, too, found it to be obscure, noting as much in an uncompleted essay called “The Effect

of the Text,” included in the Milton Babbitt Collection, Box 12, Folder 3. Attempts at interpreting the
poem can be found in Tindall 1962, 239–​48 and Holbrook (1972) 2013, 183–​90. Intriguingly, Tindall
1962, 239 associates Thomas’s diamond shapes with the “immortal diamond” of “That Nature is a
Heraclitean Fire.”
32 The gender of the observer is not revealed in the poem, but its confessional tone leads me to

identify it with the poet.

 Poetic Form and Psychological Portraiture 173

Figure 5.23. Stanzas 2 and 12 of Vision and Prayer, exemplifying the diamond
and wing shapes characteristic of the first and second halves of the poem
respectively

simultaneously “lost” in that ecstasy, in the “blinding” sun. In the published

literature, the poem’s shapes are the aspect that has garnered the most critical
commentary. William Tindall’s (1962, 240) speculation seems most appro-
priate to Babbitt’s setting: he finds that the diamonds and the wings, with
their pattern of lines successively growing and shrinking, reflect the poem’s
concern with process. The diamonds especially seem to embody the poem’s
journey in miniature, from birth to dissolution.
Babbitt’s deployment of vocal techniques follows a symmetrical scheme
reminiscent of Thomas’s shaped stanzas—​perhaps particularly the growth
and decay pattern of the diamonds. Stanza 1 is set with rhythmic but
unpitched speech. Stanza 2 is entirely in Sprechstimme. The remaining
stanzas until the last are sung. The final stanza, then, is a return to speech.
Its first half is Sprechstimme; its second half reprises the rhythmic speech of
the opening. As Mimmi Fulmer (2016, v) notes, the move from speech to
song across the first three stanzas represents a kind of birth—​the birth of
articulate voice—​reflecting the birth depicted in the opening stanzas. The re-
turn to speech in Stanza 12, the poem’s epilogue, can be taken as a retreat
174 Thinking In and About Music

Figure 5.24. Pitch-​vowel correspondence scheme in the Sprechstimme passages

of Vision and Prayer (Stanza 2 and the first half of Stanza 12)

into interiority, following the expressive outburst of the prayer. The most
prominent word of the stanza is “I,” as Tindall (1962, 247) observes. Only the
second and twelfth stanzas begin with “I,” a fact brought out by Babbitt’s use
of Sprechstimme for both Stanza 2 and the first half of Stanza 12. (Moreover,
as a result of the pitch-​vowel correspondence scheme shown in Figure 5.24,
Stanzas 2 and 12 also begin with the same pitch in the voice.) The switch
to speech comes directly with the central line of Stanza 12. David Holbrook
(1972, 1) observes that the first lines of Vision and Prayer, “Who /​Are you,”
are balanced, perhaps answered, by the central lines of Stanza 12, “I /​Am
found.” This is precisely the moment at which Babbitt returns to speech, with
rhythms that approximate a diminution of the opening lines.
Vision and Prayer demonstrates Babbitt’s continued development of
techniques for articulating poetic form. Some techniques encountered in
earlier settings reappear—​for instance, several stanzas feature a correspond-
ence between serial units, such as trichords, and poetic lines—​but I will focus
on two new to the work. The first concerns a method for the projection of vo-
calic parallelism. During the Sprechstimme passages, the voice’s pitches follow
a pitch-​vowel correspondence scheme; see Figure 5.24.33 For instance, when-
ever the soprano sings a syllable containing the diphthong /​aɪ/​—​syllables like
“sky,” “I,” “bright,” and so forth—​she sings it on the pitch D♭5.
This scheme is organized such that phonologically similar vowels are
grouped together. In general, front vowels are given higher pitches than back
vowels. The diphthongs break up the pattern somewhat—​as indeed they

33 The existence of this scheme is noted in Fulmer 2016, viii.

 Poetic Form and Psychological Portraiture 175

Figure 5.25. Opening of the vocal part of Stanza 2 of Vision and Prayer.
Low-​register identities and near rhymes brought out by the pitch-​vowel
correspondence scheme are indicated with solid and dashed slurs, respectively.

must, given that diphthongs do not fall in any single place on the continuum
from front to back vowels—​but they tend to be grouped near pure vowels they
contain. The result is that this scheme fosters association not only between
words with identical vowels but also between vowels that are merely similar.
Figure 5.25, which presents the beginning of the second stanza, illustrates
the effect of this technique. At this point, the poem primarily uses central-​to-​
front vowels, but it occasionally lurches all the way back—​to “stone,” “bone,”
“wall,” and “moan.” Babbitt’s scheme places the central-​to-​front vowels in a
narrow band of the upper register and the back vowels on E or E♭4. The near-​
rhymes of the frontwards vowels are thus proximate in pitch space, as are the
near-​rhymes of, for instance, “wall” and “moan.” The abrupt shifts between
front and back vowels that characterize this part of the poem are articulated
by leaps in pitch space. In short, this scheme represents a formalization of a
tendency already observed: the use of musical associations to bring out sonic
associations in poetry. More specifically, it draws attention both to phonemic
parallelism regardless of its place within a line and to near rhymes between
words with common or similar vowels but distinct consonants. As noted pre-
viously, Babbitt had a life-​long interest in such clever, oblique, partial rhymes,
and for rhymes other than commonplace end rhymes.
Notably, pitch in the vocal part of the Sprechstimme passages is solely deter-
mined by the pitch-​vowel correspondence scheme. Derived series appear in the
synthesizer, but the vocal part is entirely independent of twelve-​tone concerns.
The full ensemble is thus disunified to a degree it almost never is in Babbitt’s com-
positional practice: the voice does not contribute to the synthesizer’s aggregates
176 Thinking In and About Music

or to any other aspect of serial pitch structure. For these brief passages, at least,
Babbitt prioritizes phonemic parallelism above serial concerns.
Another technique systematized in Vision and Prayer concerns the ar-
ticulation of lines and sentences. In “The Widow’s Lament” and Du,
Babbitt frequently articulates the ends of poetic lines, clauses, or sentences
rhythmically—​that is, using rests or sustained notes—​although he is gen-
erally rather casual about the projection of lineation, often allowing line
breaks to pass without substantial musical articulation. In Two Sonnets, as
discussed, the metric regularity of the poems is articulated by setting lines
and hemistichs within fixed durations. Vision and Prayer, however, is not
metrically regular, and given its significance to the poem’s shaped stanzas,
lineation seems to have been too important an aspect to overlook or treat
casually. Accordingly, Babbitt systematizes the use of rests: rests in the vocal
part always, and only, appear at line endings (as noted in Fulmer 2016, v).
This technique leaves sustained notes free to indicate grammatical
articulations. Among other effects, the independence of these parameters
results in unusually clear settings of enjambment. Enjambment occurs when
a grammatical unit carries over from one line into the next; perceiving it, thus,
requires the independent observation of lineation and grammar. Figure 5.26,
the spoken setting of the final four lines of the poem, demonstrates how Babbitt
facilitates independent perception of these separate aspects of Thomas’s poem.
The lines “His lightning answers my” and “Now I am lost in the blinding”
each initiate a sentence completed in the following lines: “His lightning
answers my /​cry” and “Now I am lost in the blinding /​One.” The line endings,
as elsewhere in Vision and Prayer, are set with rests, but the lengths of the
rests interrupting these sentences are almost imperceptibly short. This allows

Figure 5.26. End of vocal part of Vision and Prayer

 Poetic Form and Psychological Portraiture 177

Figure 5.27. Babbitt’s (1964b) program note for Philomel

musical momentum to spill over into the next line. Long sustains on “cry”
and “One” then halt that momentum, indicating the end of each sentence.
Vision and Prayer and Two Sonnets demonstrate the apex of Babbitt’s
concern for poetic form. Excepting the occasional madrigalism, however,
Babbitt seems less concerned with finding musical correlates for the meaning
of these poems—​the expressive impact of these works is more like a general
sensation of awe rather than any detailed examination of character. Philomel
would see his focus shift back toward psychological portraiture.

Philomel

Perhaps Philomel justified this return because the wrenching text Babbitt and
Hollander wrote for the work practically demands it. Composed for soprano
(again the fearless Bethany Beardslee), electronic synthesis on tape, and recorded
soprano (also Beardslee), Philomel depicts the title character regaining her lost
identity, transformed. Philomel draws from Babbitt a set of musical analogues for
the character’s dawning awareness, expressivity, and agency. His program note
for the piece, which details its situation and plot, is reprinted in Figure 5.27.
Philomel is divided into three principal parts.34 In Part 1, Philomel reawakes as a
nightingale and struggles to regain language. Beginning with /​iy/​alone, Philomel

34 Hollander 1967 details the poetic structure and history of the Philomel’s text in greater depth

than is provided here.

178 Thinking In and About Music

can first only express the phonemes contained with the names “Philomel” and
“Tereus”—​her previous identity and that of her tormenter (Hollander 1967, 136).
Gradually, over Part I, she begins to recount what has happened to her, coming
to the realization that motivates the work as a whole: that “a new self could be
founded in sound.” Part II is labeled by Hollander an “echo song”—​an adapta-
tion of the poetic form the “echo verse.” The form involves a series of questions
and answers or statements and responses, in which the answers take the final
sounds of the questions and repeat them, typically to ironic or humorous ef-
fect. Philomel, the nightingale, is conversing with the other birds in the forest;
interplay between the live soprano and the recorded tape facilitates the antiph-
onal effect. Philomel sings out: “Oh Hawk in the high and widening sky /​What
need I finally do to fly /​And see with your unclouded eye?” The Hawk’s response
echoes her final consonant and vowel, delivering a harsh blow: “Die.”
Part III is a strophic aria. As Babbitt writes, her voice is now “fully
evolved”—​she has regained expressivity. Hollander’s summary of the total ac-
tion of the piece vividly captures Philomel’s overall transformation: “Where
in the first [part], she was gradually emerging from the background of forest
noises, in the second she was seeking to join a new realm of creature. Here at
the end, she reigns over a kingdom of sound” (1969, 136). The piece’s musical
journey reflects a similar transformation: at the beginning, Philomel finds
her voice only note by note, and throughout the first part, her music is dis-
tinct from, yet embedded within, the music of the forest. In Part II, Philomel
echoes the surrounding birds on equal footing. In Part III, reigning, Philomel
sings freely, independently, reborn in eloquence and determination.
Hollander (1967) includes the original text he sent to Babbitt, which differs
somewhat from the text that appears in the finished score. Comparison be-
tween the two, and examination of the annotated copy of Hollander’s text
housed in the Milton Babbitt Collection (Box 12, Folder 3), enables one to
observe Babbitt’s alterations to Hollander’s text. Hollander (1967, 136) claims
repairs were made because certain passages “proved too much for the most
heroic diction,” but several of Babbitt’s revisions seem to have sprung from
other impulses. Characteristically, Babbitt was especially interested in rhyme,
and he frequently underlined repeated phonemes and occasionally replaced
words with more heavily rhyming alternatives. For instance, “Trees filled with
mellowing /​feminine fame” was replaced with “Trees filled with mellowing
/​felonous fame,” and the annotations in Babbitt’s copy suggest he was inter-
ested in the parallelisms between the first syllables of “filled,” “mellowing,” and
“felonous” and between the second syllables of “mellowing” and “felonous.”
Babbitt also removes several lines from the ends to Parts I and III. These
Figure 5.28. Text of Philomel
Figure 5.28. (Continued)
Figure 5.28. (Continued)
182 Thinking In and About Music

passages, even following his revision, are a bit more expansive than the
sections that precede them. With evident dramatic sensibility, Babbitt tightens
them up. Figure 5.28 presents the text as it appears in the published score.
Philomel’s agency and expressivity are projected by her ability to commu-
nicate the piece’s core musical material. In short, Philomel learns how to sing
the piece’s series.35 At the piece’s outset, she expresses a unique series; it is
derived from, but apart from, the piece’s primary series, which appears in full
only in the electronics as the forest’s music. In the echo song, a cyclic pattern
of invariants demonstrates how Philomel’s utterances echo through her new
environment. She has found voice as a bird of the forest. But except for the
brief central section of Part II, she does not express any more than disjointed
fragments of the series. It is only in Part III that Philomel finds the expressive
agency to consistently articulate the piece’s primary series.
Figure 5.29 shows the opening of Philomel, the tape voice’s vocalise on /​iy/​
that precedes the entrance of the live voice at the end of m. 8. This passage has
received more attention than perhaps any other in Babbitt’s music.36 While
I will focus on pitch and the vocal part, a few comments on the electronics and
rhythmic structure are called for. The piece uses the time-​point system, as in-
dicated in Figure 5.29: the passage presents a monolinear time-​point presenta-
tion of I0 (this is noted in Swift 1976, 245). The time points divide the measure
into sixteenth notes. Therefore, every one of the many attacks that does not fall
on a sixteenth note can be understood as a subdivision of a time-​point interval.
Each such division forms a string of equal values that spans two attacks on
structural time points. The result is an accompaniment of irregular, conflicting
pulsations and long rising or falling gestures. The murmurings of a forest night.
Statements of the piece’s primary series are indicated in Figure 5.29.
Initially, the series is projected entirely by register: S4 is interpreted in m. 1
from top to bottom, extending downward from the tape voice’s E.37 S4
initiates a stack of transpositionally related series forms. As shown in Figure
5.30 (and as readily visible in Figure 5.29), these series are chosen such that E
successively moves, one order position at a time, into the series.

35 This has been noted by Boyd-​Hurrell 2017, 227 and Payette 2008, 300.
36 See Adamowicz 2011; Babbit (1976b) 2003, 344–​45; Boyd-​Hurrell 2017, 206–​38; Greitzer 2007,
105–​59; McClary 1989, 75; Mead 1994, 126–​27; 2004, 262–​68; Lewin (1991) 2006, 391–​95; Payette
2008, 294–​303; Swift 1976, 241–​46; and Taruskin 2010, 202–​4.
37 As mentioned in Chapter 3, Babbitt’s practice following Philomel would consistently require that

simultaneities drawing multiple order positions from a single series form be interpreted bottom-​up.
But here, the significance of S4 in the piece (its framing role in Parts II and III, for instance), the focal-
izing power of the voice, and the context of the rest of the passage strongly suggest a top-​down inter-
pretation. All published analyses of the passage concur.
 Poetic Form and Psychological Portraiture 183

Note by note, Philomel is finding her voice. Mead (2004, 263) describes
“[Philomel’s] voice first feeling its way into a musical space, then carving out
for itself a place to make music.” She can begin with just a single pitch. Having
found a second pitch, she immediately retreats to E. Gradually, over the
course of the next few measures, she keeps adding pitches, testing the limits
of her newfound vocality, before again retreating. Articulation indicates her

Figure 5.29. Philomel, mm. 1–​8 (until live voice entrance), with time points and
series forms indicated
184 Thinking In and About Music

Figure 5.29. (Continued)

mindset: when she leaves E5, she asserts a new phrase, but each of her first
four attempts end in retreat. It is only with the slide up to the concluding B♭
that she breaks free, heralding the arrival of the live voice.
Except for the focal E, Philomel does not repeat a pitch class throughout
this passage. In effect, she has defined a twelve-​tone series of her own (as
mentioned in Babbitt [1976b] 2003, 344 and Mead 2004, 263)—​one distinct
from the series present in the full ensemble. In the discussion that follows,
this will be called the “tape-​voice series,” reflecting its origins, although it will
eventually be taken up by the live voice and the electronics. One might say
this series is derived—​not through trichordal development, as in Babbitt’s
earlier pieces, but through the process of systematically extracting notes
from transpositionally related series that have a fixed note, the repeated E,
gradually changing order positions. As in trichordal derivation, the tech-
nique creates a new, distinct series through manipulation of the original, pri-
mary series. And as in earlier examples, the resulting distance between the
series—​the fact that the derived series stems from, yet is differentiated with,
the primary series—​lends itself to interpretation.38

38 Even though the derived, tape-​voice series predominates in the voice, the piece’s primary series

lurks close by. As Mead 1994, 126 notes, one consequence of Babbitt’s technique in mm. 1–​8 is that
the initial pcs of each of Philomel’s fragments, <45370t>, embed the first hexachord of primary series
form I4.
 Poetic Form and Psychological Portraiture 185

Figure 5.30. Series deployed in Philomel, mm. 1–​8, as shown in Figure 5.29.
Bold pcs are sung by the tape voice; others appear in the synthesizer alone.

That a complete twelve-​tone series can be derived from this procedure is

a diamond-​dust-​rare property of Philomel’s series. Given the significance of
this property to this piece, it is worth briefly lingering on how the series is
ordered to incorporate it. Figure 5.31a gives, as its top line, a generalized first
hexachord of a series—​starting on 0, for simplicity, with the remaining order
positions written as variables a . . . e. Each variable corresponds to a pitch class
in S0 or, equivalently, to an interval between the pc in that order position and 0.
Beneath S0 is a stack of transpositions generated by the inverses of the suc-
cessive intervals of the series (“−X” should be interpreted as “the mod 12 com-
plement of X”). Therefore, S−a transposes S0 by the inverse of a (that is, the
inverse of the first interval in the series); this has the effect of subtracting a from
every member of S0. The pitch class 0 is now in the second order position: a−a.
S−b works similarly: every member of S0 is transposed by −b, which is accom-
plished by subtracting b from every member of S0; this has the effect of making
0 the third order position. And so on through S−e: 0 successively shifts over,
position by position. Figure 5.31b fills in the variables with the series forms of
Philomel mm. 1–​8, effectively rewriting Figure 5.30 such that E = 0.39

39 Figure 5.31b corresponds to the upper-​left quadrant of the 12 x 12 matrix for Philomel’s series,

provided E = 0. Following Swift 1976, 243, many commentators have noted the relationship between
mm. 1–​8 of Philomel and its matrix (Lewin [1991] 2006, 391–​95; Taruskin 2010, 203–​4; Adamowicz
2011; Boyd-​Hurrell 2017, 213–​15). I find Mead’s (2004, 264) interpretation, in which the procedure
shown in Figure 5.31b represents Philomel finding her new voice note by note before returning to her
original starting point, more compelling, and think the connection to the matrix is no more than a
coincidence. Any transpositional stack of series in which the first pc successively moves inward will
necessarily correspond to the top of a 12 x 12 matrix, as Babbitt ([1976b] 2003, 344) remarks in refer-
ence to Swift’s observation. (Moreover, a matrix is not audible.)
186 Thinking In and About Music

Figure 5.31a. Generalized representation of the transpositional stack defining

Philomel’s tape-​voice series. Values corresponding to the tape-​voice series are in
boldface.

Figure 5.31b. Figure 5.31a interpreted with the values of Philomel’s series. If
E = 0, this corresponds to mm. 1–​8 of Philomel, with pcs in boldface sung by the
tape voice and others sounding in the tape.

The benefit of a generalized representation such as Figure 5.31a is that its

values can be interpreted as pitch classes or as pitch-​class intervals. This can
be confirmed by comparing variables in Figure 5.31a with numbers in the
same position in Figure 5.31b. For instance, “a−d” is the interval between the
fifth and second pcs of the series. The corresponding value in Figure 5.31b
is 7, the interval between 4 and e (that is, the interval from the fifth to the
second pc of Philomel’s series). The indication “b−c” in Figure 5.31a is the
interval between the fourth and third pcs of the series. The corresponding
values in Figure 5.31b is 4, which is the interval between 9 and 1 (from the
fourth to the third pc of Philomel’s series).
The transpositional stack of series in Figure 5.31a corresponds to the
tape-​voice pattern of Philomel, and the entries in boldface in Figure 5.31a are
those the tape voice sings in this pattern. The principle by which this pattern
realizes a twelve-​tone series is as follows: every boldface value in Figure 5.31a
except for the repeated 0s must be a different number. This has consequences
for the intervallic construction of the series. The boldface values in Figure
5.31a include every pitch-​class interval in the first pentachord of the series.
There is only one pentachordal set class that can be ordered such that all ten
of its intervals are distinct: (02347), with interval-​class vector [222220], has a
small number of orderings such that every interval besides 6 is represented.
 Poetic Form and Psychological Portraiture 187

The interval 6, then, can be picked up by the interval between the first and the
last pc of the hexachord.
These limitations—​an opening pentachord of (02347) and ic6 between
the first and last pc—​cannot be satisfied by any of the all-​combinatorial
hexachords (none of which contain both ic3 and ic6). Philomel, therefore,
is one of only a handful of pieces in Babbitt’s output whose series is not all-​
combinatorial (indeed, it is not hexachordally combinatorial in any respect).
But crucially for Part I of Philomel, the derived tape-​voice series is combi-
natorial. Its discrete hexachords are of the all-​combinatorial A-​type; in the
forms in which it appears in Part I-​1, these are {234567} and {89te01}. The
combinatoriality of the tape-​voice series has deep consequences for all of
Part I of Philomel: the entire section is organized around combinatorially de-
fined hexachordal areas of the derived, tape-​voice series.40 Even though the
primary series is not combinatorial, combinatoriality still forms the basis of
Part 1’s serial organization, much as in Babbitt’s other works.
When the live voice enters in m. 8, she launches yet another derived
twelve-​tone series, which will accordingly be called the “live-​voice series.”
This too is generated from a transpositional stack that moves a fixed pc one
order position at a time. Figure 5.32a gives the first instance of this series.
Undoing the retrograde, for the sake of comparison with the tape-​voice se-
ries, Figure 5.32b translates the live-​voice series onto the generalized tem-
plate provided in Figure 5.32a. Boldface notes continue to represent the
tape-​voice series; underlined notes now represent the live-​voice series. It can
be observed that the possibility for creating a twelve-​tone series using these
particular, underlined order positions is nearly a result of creating one with
the boldface order positions. Eleven out of twelve of the underlined values
are inverses of the boldface values (e.g., b−a replacing a−b, or d replacing
−d). The only additional values that need to be accounted for are e−a
(unique to the live-​voice series) and c−d (unique to the tape voice series). If
they are equivalent—​that is, if the interval between the sixth and second pc
in the series is the same as the interval between the third and the fourth—​
then a series that can create a derived twelve-​tone series using the boldface
order positions can also create one using the underlined order positions. In
Figure 5.31b, it can be observed that the corresponding intervals in the se-
ries for Philomel do indeed match.

40 Dubiel 1990a, 62–​75 discusses “Sounds and Words,” one of the only other pieces Babbitt wrote

without a combinatorial series, and similarly observes that combinatoriality is nonetheless signifi-
cant to the piece.
188 Thinking In and About Music

Figure 5.32a. Live voice series in Philomel, mm. 8–​13. Boldface pcs are sung;
others sound in the tape.

Figure 5.32b. Generalized representation of the transpositional stack defining

Philomel’s two derived series. The tape-​voice series is in boldface. The live-​voice
series is underlined.

Like the tape-​voice series, the live-​voice series is combinatorial; it also has
distinct harmonic content. Its discrete hexachords are of the C-​type. The
dissimilarity between the A-​type tape-​voice series, the C-​type live-​voice se-
ries, and the omnipresent, non-​combinatorial primary series characterizes
Part I-​1 of Philomel. The contrasting series provide a narrative impulse to
the work that tracks the psychological progression of its title character. The
monodrama is about Philomel regaining her voice and her identity, reborn as
a nightingale. As Philomel finds new life in the forest, she learns to sing the
forest’s music—​its series, the primary series of the monodrama. The distance
between Philomel’s music and the piece’s primary series eventually closes.
The first barrier to fall is the live-​voice series, which is unique to Part I-​1.
This matches a shift in the text: for the remainder of Part I, the tape voice and
live voice fuse into a single, although multiply represented character—​note
the tape voice no longer contradicts the live voice after the first lines of Part
I-​2. Indeed, although unmarked as such in the score or text, Part I-​1 might
be considered introductory, from both a musical and textual perspective.
Musically, it establishes the primary materials of the piece yet is a discrete,
autonomous section. Textually, the character Philomel seems bewildered in
Part I-​1. She gradually rediscovers the ability to articulate words, but all she
can do is remark on her observation of trees and tears. It is only in Part I-​2
that she is able to articulate the cause of her distress: Tereus (although his
 Poetic Form and Psychological Portraiture 189

presence haunts the phonemes of Part I-​1, even to the vocalise on /​iy/​that
launches the work). Musically, Part I-​2 initiates the material used throughout
the remainder of Part I. It is governed by the tape-​voice series alone (even
when the live soprano is singing), and the forms of that A-​type series are,
for the rest of Part I, consistently those in the hexachordal area formed
by {56789t} and {e01234}. These hexachords have a minimal number of
common tones (three each) with the A-​type hexachords of Part I-​1. The ef-
fect again is to demarcate Part I-​1 from what follows.
With one exception, in each of the remaining sections of Part I (from
Part I-​2 on), the pitch material is generated from four transpositional
stacks: statements of the tape-​voice series from the {56789te}/​{e01234}
hexachordal area that cycle through the four ordering possibilities (S, R, I,
RI). It is clear from this that even though the series I have designated “pri-
mary” has a good argument for primacy—​it is, after all, omnipresent, and
the tape-​voice series is generated through the selection of a few of its order
positions—​it is the tape-​voice series, and its combinatoriality, that provides
the organizational impetus for the section. The forms of the primary series
that appear are only those necessary to yield the combinatorial forms of the
tape-​voice series.
This has hermeneutic implications. Philomel has yet to fully recover her
agency, to find full expression in her new life in the forest. She achieves that
only with her adoption of the primary series in Part III. But in Part I, the
forest’s music is organized by the serial properties of Philomel’s tape-​voice
series. We, as listeners, are hearing the forest from her perspective, the per-
spective of one yet to achieve full voice.
The preeminence of the tape-​voice series generalizes the programmatic
implications of its initial deployment. In the opening measures of the piece,
the tape voice searches for notes, adding one pc at a time before retreating
to E. The tape-​voice series is thus imbued with Philomel’s rediscovery of her
voice. By organizing Part I around that series, Babbitt casts the whole of Part
I as a continuous process of expressive growth, a reflection that Philomel’s
voice continues to develop. This matches the growing phonemic vocabu-
lary of the text, as it expands toward and then beyond the names “Philomel”
and “Tereus.” By the final section of Part I, Philomel has begun to reclaim
her identity. No longer simply recounting the violence she has suffered, she
begins to recognize her own, new voice. She begins the section puzzled, un-
able initially to hear herself as herself; she can only ask, searchingly, “What
is that sound? A voice found?” She first notes her transformation into music
190 Thinking In and About Music

conditionally, “as if a new self could be founded on song.” But by the end
of the section, she can finally claim her new identity affirmatively: she is
“becoming [her] own song.” Musically this last section is somewhat more
expansive than those that precede it, enabling Philomel to ponder on her
recognition at some length. Babbitt facilitates this by shifting the last trans-
positional stack from the penultimate section into Part I-​5 before Part I-​5
presents a full complement of its own.
Part I ends with an interlude that includes something nearly unique in
Babbitt’s mature work: a thematic reprise. An intensified return to the material
of the first eight measures, now in the synthesizer alone, rounds off the section.
It is a farewell to Philomel’s tape-​voice series. No longer alien to the forest, re-
born in song, Philomel seeks guidance from the birds she flies among.
Fittingly enough, Part II of Philomel, the “echo song,” is a cacophony of
musical echoes. The section features seven questions or exclamations (by
Philomel) and seven responses (by various birds of the forest). Perhaps as a
result, the music is episodic: seven discrete sections, for each question and
answer, separated by brief interludes. The overall design is roughly symmet-
rical: an increase in contrapuntal complexity in sections 1–​3, from three up
to six voices, answered by a corresponding decrease in sections 5–​7. Section
4, the central section, is musically distinct from the rest of Part II.
Babbitt deviates from Hollander’s text in a significant respect (see
Hollander 1967, 138–​ 39). Hollander specifies that the questions and
exclamations come from the live soprano and the responses come from the
recorded voice on the tape. The characterization in Hollander’s original is
simple: the live singer is the character Philomel; the voices on the tape rep-
resent the birds around her. In Babbitt’s setting, the live singer does ask the
questions. But the tape answers only with the first, echoing word of the re-
sponse. The live singer then re-​enters, repeating the first word of the response
and continuing with its remainder. The live singer therefore is not simply the
character Philomel, as in a staged drama. Her role is more like a storyteller—​
often representing Philomel, but also slipping into other voices as the narra-
tive requires. Aspects of Part I similarly suggest a fluid, storyteller-​like role
for the live singer. In Part III, Philomel’s aria, there is less ambiguity: the live
soprano seems to be fully identified with the character Philomel. When the
tape voice enters, it is merely to support her.
Excepting section 4, the technique used throughout the section is that named
“mutual partitioning” in Babbitt’s brief technical comments about the piece.
As he describes, mutual partitioning enables series forms to “mutually express
 Poetic Form and Psychological Portraiture 191

another, by order-​preserving partitional extraction” ([1976b] 200, 344). Figure

5.33, from the first section, demonstrates the technique. S4—​the primary series
form of the piece reappearing here to launch Part II—​alternates throughout
the section with RIe. Both series forms include <45te>, <3870>, and <1692> in
order (although not as segmental adjacencies). In the setting of this passage, the
series are realized in linear temporal order, but the three invariant components
are partitioned into three-​part counterpoint.
The technique is echoic in several respects. First, as Babbitt suggests, the
two series forms are shown to echo (“mutually express”) one other through
their shared tetrachords. More concretely, the invariant subsets rotate in a
swirl of invertible counterpoint. Over the course of the section, each line will
state each tetrachord twice, once during the expression of each series, and as a
result the lines continually echo each other.41 Thus, the soprano’s music is con-
stantly reiterated by the electronics. Philomel’s voice resounds in the forest.
She contributes to it now as an equal: no longer deriving distinct material, her
line is equivalent to the electronic lines around her. But she is still questioning,
uncertain—​and not yet able to carry the primary series of the piece in full.
This tetrachordal partition of the first section is unusually simple and
efficient.42 The reason this capability holds between S4 and RIe is straight-
forward: the series forms are exchanged by the operation RI3, and the
tetrachords <45te>, <3870>, and <1692> are symmetrical under RI3 (as is
noted in Babbitt [1976b] 2003, 361n.9 and Adamowicz 2011). Rounding off
Part II, the final section returns to a tetrachordal partition with harmonic
content identical to the first section, although with new series forms (via the
inversional symmetry of the tetrachords). In the second, third, fifth, and sixth
sections of Part II, a variety of more complex mutual partitions are deployed.
For the second section (“Oh Hawk”), three series forms are repeated in cyclic
rotation, mutually partitioned into five shared subsets; see Figure 5.34. Three
subsets have invariant ordering among the series forms, like the tetrachords
in sections one and seven. Two dyads, {02} and {39}, do not. As in sections
one and seven, the series forms are realized contrapuntally so as to bring out
the various subsets: the five subsets are projected in the vocal line and four
registers in the electronics.

41 The result is a section length of six series forms: 2 different series forms * 3 contrapuntal

dispositions. The lengths of all sections in Part II except the fourth are the result of similar formulas.
42 Indeed, this partition is special enough that it might be considered a determinant of the second

hexachord of the series. The tape-​voice pattern of Part I limits the ordering of the first hexachord, and
achieving a (43) mutual partition between pairs of series forms, given a fixed first hexachord, substan-
tially limits the possibilities of the second hexachord.
192 Thinking In and About Music

Figure 5.33. Mutual partitioning exemplified in Philomel, Part II-​1

Figure 5.34. Series forms used Philomel, Part II-​2. The five shared subsets are
indicated with different typefaces.

The central section of Part II does not use the mutual partitioning tech-
nique. Indeed, the procedure used in this section may be unique in Babbitt’s
output. The voice, for the first time in the piece, sings complete series forms,
as does the tape; see Figure 5.35. Their respective series forms seem to have
been chosen to maximize pc intersection (and perhaps to avoid S4, whose full
linear statement is reserved for Part III). The voice and tape frequently align
in octave unisons, after each of which the two parts proceed to fill in just as
many pcs as carry them to the next alignment. The result is a different kind of
echo between the voice and the electronics: the resonance of octave unison.
At long last, in Part III Philomel achieves full expression. No longer rel-
egated to a series of her own, as in Part I, or to fragments of the series, as
in Part II, Philomel sings, over and over, full series forms, beginning and
ending Part III with S4, the principal series of the piece—​that which is intro-
duced in the tape in m. 1. Any disjunction between her and the electronics is
gone. “Reigning,” she has captured the forest’s music. It is hers, inflected with
her own pain. She has moved from derived series to the primary series, and
from inarticulateness and distance to coherence, presence, and a reclaimed
identity—​she is proud, angry, and fully transformed.
Figure 5.36 presents the opening of Part III, the first linear statement of
S4 by the live soprano. As an aside, note Babbitt’s continued sensitivity to
poetic form. The four initial rhyming trochees are all set in parallel ways,
 Poetic Form and Psychological Portraiture 193

Figure 5.35. Serial counterpoint in Philomel, Part II-​4. Boldface notes appear
simultaneously, in octave unison. The vocal line includes the live soprano and
the tape voice; the latter is underlined.

with dynamics distinguishing the stressed and unstressed syllables, and

all six stressed syllables are emphasized either dynamically, metrically, or
durationally.
The overall form of Part III is driven by a steadily growing contrapuntal com-
plexity: each successive stanza features arrays built from an increasing number
of lines. From the perspective of Babbitt’s compositional development, Part III
includes innovations in partitioning and aggregate construction that bridge
Composition for Tenor and Six Instruments and the all-​partition array, a tech-
nique Babbitt would adopt only a year after the completion of Philomel. Figure
5.37, the array underlying the final stanza of Philomel, gives one intriguing in-
stance. The aggregates move from a relatively uneven partition (only two lines
sounding) to a perfectly even one (all lines present two pcs).43 This progression
foreshadows in miniature the opening of the all-​partition array in Relata I (see
Figure 3.2). But a full investigation of Part III is unnecessary for the purposes
of this chapter, because Babbitt’s psychological portraiture of Philomel’s re-
birth as a nightingale is completed with her arrival on S4. In Hollander’s (1967,
136) words, this moment marks “the fully realized singing of the fully formed
nightingale.” All that remains is her strophic aria reflecting, in full-​throated
anger, on her torment and her transformation. Her “pain” is “fused into music.”
Her “suffering,” “redeemed in song.”
Philomel is unlike any other piece Babbitt wrote, as the composer readily
admits (Babbitt 2000). The derivation pattern of Part I, a representation of
Philomel’s reaching out with her new voice, is unique to the piece, and the
non-​combinatorial series that results is nearly so. The mutual partitioning
pattern of Part II, which forges echoes between the live soprano and the

43 This figure was completed with assistance from archival materials in the Milton Babbitt

Collection, Box 12, Folder 3. Three components are omitted from the published score.
194 Thinking In and About Music

Figure 5.36. Philomel, beginning of Part III, vocal part. Trochees are bracketed;
stressed syllables are marked with accent symbols.

Figure 5.37. Array from the opening of stanza 5 in Part III of Philomel.
Bracketed components are serially expected but missing in the published score.

tape and thereby represents the forest’s echoes of Philomel’s questions, is

also motivated by the work’s particular dramatic setting. It was not used
again. Even the use of manipulated recording—​the tape voice—​is unique
in Babbitt’s output, the sounds in his other electronic works being gener-
ated entirely on the synthesizer. But this foray into musique concrète allows
Philomel’s fractured identity to be represented by sonic and even spatial
multiplicity, as her voice is distributed between the live singer and the quad-
raphonic tape—​an effect that must have been especially remarkable at the
work’s premiere, when both the live and taped voices were those of Bethany
Beardslee. In all of these elements, the particular dramatic situation of the
text inspired substantial alterations in Babbitt’s approach.
 Poetic Form and Psychological Portraiture 195

Conclusion

Psychological portraiture in Philomel, as in “Widow’s Lament” and Du, is de-

pendent at least in part on serial derivation, which offers for metaphorical
construal the distinction between an underlying series and elaborations or
manipulations of it. Much of the projection of poetic form in all five works
discussed in this chapter is also a function of the serial techniques Babbitt
favored at the time. The harmonic changes that result from the transition
between trichordal arrays constructed from different set classes enable the
composer to articulate significant moments in poetic form, such as sentence
endings, line endings, or stanza breaks. Babbitt’s transition to all-​partition
arrays would close off these means of projecting psychological portraiture
and poetic form, for all lines in every block of his all-​partition arrays are
members of the same series class. Perhaps as a result, text setting in works
after Philomel is less linked to serial concerns. It becomes, on the whole, a
more local affair than it had been, concerned primarily with memorable and
precise articulations of words and sentences and the framing of significant
moments. From a pitch-​class perspective, the arrays exist in one stream, the
text in another—​running in parallel but not mutually projecting each other
as they had in earlier works. While Babbitt had previously adjusted his serial
practice in order to set his texts, in the vocal works following Philomel, pitch-​
class structure is generally indistinguishable from his instrumental pieces.44
Looking back on his output as a whole, the moment when Philomel finally
manages to sing her work’s principal series form marks the end both of her
transformation and the end of a chapter in Babbitt’s career.

44 Rhythm is another matter, one deserving of further inquiry. In several early and late texted

pieces—​“The Widow’s Lament,” Du, Two Sonnets, “The Virginal Book,” “Now Evening after Evening,”
“Pantun,” “The Autobiography of the Eye”—​there seems, as far as I have been able to determine, to
be no serial rhythmic structure. (For the latter three, sketch materials provide support for this: see
Milton Babbitt Collection Box 11, Folder 2; Box 12, Folder 1; and Box 2, Folder 2.) This contrasts with
almost all of Babbitt’s instrumental work and may reflect concern for text setting.
 Afterword
“Anything Vital is Problematical”

We are left with more questions than answers. The most pressing, the ques-
tion of whether Babbitt’s music realizes the organicist, constructional, and
cognitive program of his writings, turns out not to be readily answerable: the
topics addressed in his writings play a part in his music, but they are ev-
idently only a portion of the compositional matters that concerned him.
Listeners to Babbitt’s music would do well to consider the subjects raised
in his writings, but they should not feel limited by them. The composer evi-
dently was not.
In short, one should study Babbitt’s writings not because his theories are
a guide to his music as a whole, but because they propose mindsets, along
with a certain set of techniques, that have the potential to illuminate certain
aspects of his music. The remaining aspects, some of which seem to conflict
with the implications of his writings and some of which are merely orthog-
onal to them, require other mindsets and other techniques. His writings pre-
sent an incomplete picture of his music. He did not claim otherwise: in his
few extended comments about his own music, he never promised a compre-
hensive account—​indeed, he never claimed to give a comprehensive account
of any piece in any of his analytical writing.
One might ask, then, why Babbitt chose to speak about those elements
he did and elide or obscure those he did not. Two possible answers come to
mind. First, like that of many twentieth-​century artists, Babbitt’s career can
be read as an attempt to piece together a new practice by radically focusing
on certain elements of the inherited tradition. In Babbitt’s case, this meant a
thoroughgoing serial re-​imagination of Schenkerian organicism. Given that
this aspect of his practice connects him to his predecessors, such as Schenker
and Schoenberg, it is no surprise that he focuses on it in his writing and
270 Afterword

speech.1 A second answer lies in Babbitt’s concern for theoretical language

and methodology. In line with his philosophical commitments, Babbitt lim-
ited himself to statements that are empirically verifiable. Organicism, once
recast axiomatically or in the language of cognitive science, is relatively ame-
nable to explication in that language. But there is no reason to infer, as many
have, that Babbitt’s meta-​theoretical views limit his music in the same way
they limit his discourse. He certainly never claims they do.
For as a practical, working musician, Babbitt chose to supplement (or, at
least, articulate) his systematic constructional principles with a variety of
compositional means, ranging from the rhetorical (how does one signal the
end of a piece?), to the expressive, to the technically clever, to the witty. If the
results fail to look logical or unified—​if they cannot be analyzed in light of any
single ideology—​it is because they are not. To me, this is evidence of richness,
if of a very different kind of “richness” than that Babbitt intended by the term.
In 1950, at the dawn of his career, Babbitt posed a remarkable criticism
of René Leibowitz. Leibowitz, Babbitt ([1950a] 2003, 12) claims, failed to
recognize “the extremely problematical nature of the twelve-​tone system.”
Given that Babbitt also claimed the twelve-​tone system “cedes nothing to any
musical system of the past or present” (Babbitt [1960] 2003, 67), these are
astonishing words. But, he continues, “Anything vital is problematical; the
nonproblematical is static” (Babbitt [1950a] 2003, 12). The term “problem-
atical,” that is, as the ensuing discussion clarifies, is not used in the sense of
containing flaws, but of posing questions—​of occasioning thought. Leibowitz
had underestimated just how much reconsideration of basic principles the
twelve-​tone system requires. Nor, for that matter, was that reconsideration a
finite pursuit—​Babbitt does not imply that the twelve-​tone system would at
some point become adequately theorized and, thus, nonproblematical.
Babbitt’s music, too, resists satisfying, logically consistent, bounded theori-
zation. It will continue to occasion thought. His exegetes have the pleasure—​
and the responsibility—​to revel in the problematical issues it raises.

1 Babbitt justifies this choice propaedeutically:

To direct a listener’s attention to the unique aspects of a work, particularly when he prob-
ably knows the work little or not at all . . . is to emphasize that which will provide least
aid in initial comprehension, for—​to such a listener—​uniqueness is far less significantly
helpful than is communality. . . . Indications of the procedural sources, the technical
traditions . . . provide not only a point of entry but, eventually, the bases for determining
the depth, extent, and genuineness of the work’s originality. (Babbitt [1970] 2003, 237)
This remark is reiterated in Babbitt 1987, 3–​4.
 7
Babbitt’s Gestural Dialectics

Introduction: The First Measure of Post-​Partitions

Figure 7.1 presents all the series forms used in Post-​Partitions, a 1966 com-
position for solo piano.1 From this level of background structure, the
piece appears fixed, cold, and impersonal; a mechanical cycling through
permutations of series forms. Each pair of lines is isolated within a narrow,
fourteen-​pitch register, and the six total pairs are projected in maximal dis-
persion across the lowest eighty-​four pitches on the instrument.
A zoomed-​in look at the partitioning of the first aggregate, shown in
Figure 7.2, does little to disturb this impression. Each line contributes but
a single pitch. At this stage, it hardly seems justified to call them lines—​the
various pitches amount to no more than isolated events, and no linear conti-
nuity connects one to another. From this point of view, it appears the aggre-
gate will not amount to much more than pointillistic scattering, each pitch
distinct from every other both contrapuntally and in register.
The realization of this aggregate, however, shown in Figure 7.3, presents
a vivid, visceral gesture: a jagged ascent followed by a 𝆑𝆑𝆑𝆑𝆑 crash onto the
lowest note on the piano. The separate pitches contributed by each series
form are threaded together to form a coherent energetic profile. Rather than
being perceived as abstract pointillism, the gesture grabs the listener bodily.
Tracking the piano’s uneven rise and violent descent, a listener is carried into
the work on a whirlwind of explosive motion.
The full profile of the measure resembles a condensed form of the “pro-
totypical four-​part gesture” discussed by Patrick McCreless (2006, 12); his
schematic is reproduced in Figure 7.4. Something like the “rebound” at the
end of the gesture occurs in the next measure: its most salient events are
in a middle register. Although McCreless identifies the gesture in Mario
Davidovsky’s Electronic Study No. 1 (1961), as McCreless describes, “the ges-
ture is quintessentially one for the Romantic piano” (2006, 14). And indeed,

1 The full array for Post-​Partitions can be found in Mead 1994, 274–​77; see also 171–​88.

Thinking In and About Music. Zachary Bernstein, Oxford University Press. © Oxford University Press 2021.
DOI: 10.1093/​oso/​9780190949235.003.0007
232 Thinking In and About Music

Figure 7.1. Series forms in Post-​Partitions

Figure 7.2. Partitioning of the first aggregate of Post-​Partitions

despite its irregular rhythms and unforgivingly atonal harmonic language,

an alter Duft of virtuosic Romantic pianism wafts through Post-​Partitions. If
Babbitt’s compositional impetus was to stitch the raw material of Figure 7.2
into a dramatic flourish, the familiar gestural profile of Figure 7.4 provided a
convenient vehicle.
The first measure of Post-​Partitions is in some respects an extreme ex-
ample. It is, for instance, the only measure in the piece to draw exactly one
pitch from each line. And some parts of Post-​Partitions, unlike m. 1, are in-
deed too disjunct to support deeply felt gestures. But as the piece continues,
there are numerous points at which I experience a sense of embodied energy.
Indeed, the first measure of Post-​Partitions is representative of a general phe-
nomenon in Babbitt’s music. The serial background, composed of numerous
 Babbitt’s Gestural Dialectics 233

Figure 7.3. Post-​Partitions, m. 1

series forms projected in narrow registral bands, is worked into a surface that
crackles with gestural motion.
A number of authors have reported such experiences. Indeed, energetic,
gestural language is reasonably common in published analyses of Babbitt’s
music. It features prominently in numerous writings by Dubiel (esp. 1992)
and Mead (esp. 1999; also 1994), as well as in excellent one-​off pieces by
Benjamin Boretz (2006), Leong and McNutt (2005), and Greg Sandow
([1982] 2004). Boretz’s (2006, 11) description of Babbitt’s music as a “sam-
pling of actions” is particularly apt. The characteristic energetic behavior of
Babbitt’s music involves a concatenation of local, unpredictable gestures—​a
“sampling,” that is, rather than a purposive sequence leading toward a goal.2
Extending these authors’ work, the purpose of this chapter is to explore
the sense of gesture Babbitt’s music can impart in light of the odd, even hos-
tile situation in which these gestures are formed. Generally, Babbitt’s arrays
will be posited as fields from which dynamic realizations may emerge or
against which dynamic realizations may struggle. The competing concerns of
twelve-​tone structure and gestural sensation can be envisioned as a dialectic: a

2 This fact, too, suggests the pertinence of Levinson 1997.

234 Thinking In and About Music

frequency

rebound
time

CRASH

1 2 3 4

dynamic
curve

Figure 7.4. A “prototypical four-​part gesture,” reproduced from McCreless

2006, 12

process of opposition from which complete, multi-​dimensional, and multi-​

faceted compositions emerge. In addition to highlighting the music’s gestural
qualities, attention to this dialectic is revealing about a variety of topics, in-
cluding the evolution of Babbitt’s contrapuntal and rhythmic techniques, his
text setting, and certain rhetorical deviations from serial expectations.

Sources of Embodied Energetics: Kinesthetic Empathy

and Musical Forces

Musical gesture has been a major topic of research in recent years. Attempting
to summarize the state of the field would be a futile task. But I will adopt, pre-
liminarily, the definition of Robert Hatten (2004, 95): gesture is “the signifi-
cant energetic shaping of sound through time.”
As a prelude to the following discussion, Hatten’s definition of gesture is
worth lingering on. Gestures are “significant.” This is not a claim about their
 Babbitt’s Gestural Dialectics 235

importance or salience but about their role in communication. Gestures

carry meaning. They transmit information about bodily response and may
signify aspects of rhetoric, thematic character, emotional reaction, or topic.
Musical gestures are “energetic.” They tap into human sensations of exer-
tion: motion, tension, inertia, pressure, weight, attraction, repulsion, and
so forth. The result is the “shaping of sound.” “Shaping” is a curious word
to consider in combination with “energy,” much less “sound.” Energy is not
an entity with a shape. (What is the shape of tension?) Sound, too, defies
the finite dimensionality of objects that may be shaped. But the words in
combination, the notion of “energetic shaping of sound,” capture an im-
portant aspect of musical experience: a compelling energetic profile can
mold separate bits of musical data into a Gestalt with a defined character.
It can transmute a sequence into a memorable chunk. And, finally, gestures
are “in time.” For energy to resonate with human bodies, it has to operate
in a manner analogous to energy exerted by humans or energy affecting
humans—​that is, temporally. This requirement excludes some energetic
ways of talking about music and simultaneously puts useful limits on the
topic. For instance, organicist rhetoric often posits an energy motivating
the generation of pieces from initial kernels. But such accounts are spe-
cifically not in literal, real-​world time. They are abstract conceptions, for-
malistic means of theorizing musical ontology in which energetic language
functions mystically or in the manner of “prop oriented make-​believe” (as
discussed in Chapter 1). This language is useful as a heuristic, but it is not
intended to communicate any connection to the energy experienced by
human bodies.
An idea common to many theories of gesture is that a source of per-
ceived energy is “kinesthetic empathy” with a performer (Mead 1999, 3ff.).3
Namely, this theory proposes that listeners physically respond to aspects of
a performed piece as if they themselves were performing it. Listeners em-
pathize with the kinesthetic motions required for the piece’s realization.
This notion is deepened and greatly elaborated by Arnie Cox (2016, 2011),
who calls his expanded version of the concept the “mimetic hypothesis,”
buttressing the theory with a good deal of cognitive science. As Cox (2016,
37) explains, “Many or most musical sounds are evidence of human motor
actions that produce them”—​a fact that facilitates the transfiguration of the

3 A related term for this phenomenon is “virtual participation.” See Margulis 2014, 140–​58 for val-

uable discussion.
236 Thinking In and About Music

passive experience of musical hearing into the imaginary mimetic reenact-

ment of musical production. Mimetic translation encompasses a variety of
responses, going beyond direct imitation of a performer’s motions. For in-
stance, it can involve such reactions as abdominal clenching in response to
dynamic intensity.
As a result of kinesthetic empathy, a listener can feel the rise and sudden
crash in the opening of Post-​Partitions energetically: the registral trajectory
corresponds to a pianist’s right-​and-​left motions on the keyboard. Those
with a background in piano performance may feel this most directly—​as
Cox (2016, 51–​52) discusses, familiarity can breed responsiveness—​but even
those without such a background can still empathize with the exertions the
performance of this measure requires. In general, a listener need not be fa-
miliar with the mechanics of sonic production in order to experience kin-
esthetic empathy—​nor need the sound even be produced by an acoustic
instrument. Through “cross-​modal mimetic motor imagery,” a listener un-
familiar with the production of a particular sound may nonetheless mimet-
ically experience that sound as energetic by the translation of it into a more
familiar mechanism (52).4
Many of the gestures in Babbitt’s music, like that seen in Post-​Partitions,
are organized around directed registral trajectories, playing into familiar
sensations of ascent and descent or expansion and contraction. Others
are based around intensity of amplitude, with swells or fades inspiring
sensations of tension or release. Septet but Equal features sensations in-
spired by both register and dynamic intensity. Figure 7.5 shows the
opening tutti and the cello solo that follows (as detailed by Mead 1997,
114–​17, the work progresses through an exhaustive list of instrumental
combinations).
Within the frenzy of the opening ensemble, each instrument realizes one
or several striking registral trajectories. Note in particular the counterpoint
of the jagged descent in the first clarinet and the rocketing, then retreating,
violin; both gestures are loosely imitated, under rhythmic variation, in the
piano in m. 2. The bass clarinet more calmly descends, sustaining a low F

4 An appealing related suggestion for listening to Babbitt’s music is suggested by Greg Sandow

([1982] 2005, 255), who recommends listeners engage with his music by imagining themselves
singing along. As Cox (2016, esp. 28–​32) explains, “subvocal” participation is a key component of
mimetic engagement.
 Babbitt’s Gestural Dialectics 237

before tagging the lowest note in its range and bouncing back into a high
register. The other instruments’ gestural profiles are more discontinuous and
local. The viola, for instance, is static, perhaps non-​gestural, in mm. 1–​2, be-
fore suddenly initiating a sequence of leaps in m. 3.
The dynamic profile of the opening tutti is much simpler than its coun-
terpoint is. As is virtually universal in Babbitt’s late music, each member of
the ensemble plays at the same dynamic value at any given time (a result of
Babbitt’s deployment of the time-​point system in his later years). The first

Figure 7.5. Septet but Equal, mm. 1–​9. Score in C.

238 Thinking In and About Music

Figure 7.5. (Continued)

few measures of the opening tutti realize a series of three periodic swells
between 𝆏𝆏 and 𝆑𝆑, in which 𝆏𝆏 spans one beat and 𝆑𝆑 spans two. From a
 Babbitt’s Gestural Dialectics 239

mimetic perspective, these swells can be experienced “amodally,” in Cox’s

(2016, 45–​46) terms—​that is, not by the imitation of any performative ex-
ertion but “with a clenching and relaxing in the gut that matches something
of the energy dynamics of the music.” The swells unify an otherwise highly
differentiated ensemble with energetics that transcend the septet’s many
contrapuntal lines.
A dramatic slackening of energy occurs between mm. 2 and 5, as the
opening tutti recedes into the cello solo. As with the ensemble or registral
lists discussed by Dubiel (1992, 2008), the juncture between the tutti
and the cello solo is smoothed over and mediated, serving to place the
cello solo as the end result of a process of textural thinning rather than
merely as the next item on the list of ensembles. The tutti is first culled of
sounds most distant from the cello: the upper registers of the clarinets,
violin, and piano have ceased activity by m. 2, and those instruments are
silent altogether by the downbeat of m. 3. The bass clarinet lingers a bit
longer. Timbrally distinct from but registrally proximate to the cello, it
carries the impulse from the clarinets into the lower strings. The piano’s
last burble, in m. 3, similarly carries the piano’s energy down into what is,
roughly, the cello’s register. The remainder of the tutti is a duet between the
cello and its closest relative, the viola. The viola is mostly inactive in mm.
1–​2, reserving its material for this concluding burst of activity in m. 3,
during which the cello reiterates its final trichord. (Each member of the
ensemble—​counting the two hands of the piano separately—​completes an
aggregate in each section.) This brief duet between the cello and the in-
strument most like it gives way, smoothly, to the cello solo. Following the
cello solo, this progression is reversed: the next section is a trio between
cello, viola, and bass clarinet.
The cello begins with a series of short, skittish motions. These fragmen-
tary gestures contrast with the continuous, confident swells of the full en-
semble. In mm. 6–​7, the cello gets stuck in a middle register, trying out
variations of a three-​note figure. An initial, quiet leap up at the end of m. 7
is followed by a breakthrough at the end of m. 8, a high 𝆑 peak whose per-
formative difficulty communicates physical strain. Having reached this mo-
mentary climax, the cello retreats, relaxing downward with a series of drops
in register and dynamic, interrupted only by a momentary fillip toward the
end of m. 9.
While mimesis may account for a good deal of gestural sensation, some
authors have theorized aspects that may touch on it only tangentially. In
240 Thinking In and About Music

particular, one concept developed by Matthew L. BaileyShea (2012) and

BaileyShea and Seth Monahan (2018), following on work by Steve Larson
(2012), is the notion of “forces,” imaginary analogues for energy affecting
human bodies. One such force is called “generalized gravity,” the principle
that “musical contours, like our bodies, are pulled downward toward a more
stable state” (BaileyShea and Monahan 2018, 33). More intensity is required
to rise than fall; energy is built up in ascent and discharged in descent. The
concept is familiar from tonality—​the descent of Schenker’s Urlinie can be
considered as codifying a similar notion (Schachter 1996, 330–​31)—​but no-
tably for BaileyShea and Monahan (2018, 33), it is in effect “whether or not we
can predict or define the conditions for ultimate stability” (emphasis original).
It is in this sense that their gravity is “generalized.” Gravity pulls down, even
when no tonic is in sight.
When considering such forces, a perspective distinct from that of mimesis
occurs. Rather than imagined energy within music, exterior forces are imag-
ined as acting upon music.5 The forces felt are not those of bodily energy itself
but of the energies that typically affect bodies. The difference in perspective
means that mimesis and musical forces are often both operative; a listener
can sympathize with a performer and simultaneously posit an energetic con-
text within which that performer acts. Other physical forces—​inertia, mag-
netism, momentum, and so forth—​may also be called to mind by musical
experiences.6
Returning to Post-​Partitions, generalized gravity illuminates some of the
remarkable effect of its opening measure. Given gravitation, a fall at the end
of the measure seems a natural consequence of the opening rise, but the
depth of the drop and the increasing dynamics contribute to the sense that
the music crashes rather than releasing energy in a stabilizing descent. The
resonant perfect fifth supporting A0 further contributes to this effect.
The presence of the lowest note on the piano is a crucial determinant of
the gesture’s quality: the coincidence of the loudest dynamic in the piece and
A0 suggest that the crash is, from the perspective of dynamic and register, as
extreme as possible. From a serial perspective, the availability of this extreme
is hardly to be taken for granted. It requires that the lowest register of the first
block of the array presents one of the four series forms that begin with pitch
class A. Given the many mutually constraining aspects of array construction,

5 BaileyShea 2012 includes insightful commentary on the differences between these perspectives.
6 Larson 2012 explores this topic in detail in reference to tonal music.
 Babbitt’s Gestural Dialectics 241

this can conceivably be taken as a premise of the piece, or at least a decision

with long-​term consequences. For instance, assuming the series as prior—​as
Mead (1994, 171–​72) notes, it is borrowed from Partitions—​and also taking
the piece’s distribution of hexachordal areas as given (see Mead 1994, 172–​
73), R8 is the only series form within the hexachordal area assigned to the
lowest register that begins with A. Moreover, as can be seen in Figure 7.1,
the series forms for the piece are arranged in maximally diverse orderings
projected in three distinct ways (as is typical for Babbitt’s all-​partition arrays).
Within each block, quartets of registrally adjacent lines express the four pos-
sible serial orderings (S, I, R, RI); within each consecutive pair of blocks, each
line-​pair expresses the four possible orderings; and within each half of the
piece, each line expresses the four possible orderings. Accordingly, the pres-
ence of R8 in the lowest line at the beginning of the piece—​in combination
with these other constraints—​has ramifications for the choice of other series
forms across the array.
Moreover, the gesture justifies a registral projection that makes A0 avail-
able. Given the six registers of the array, Post-​Partitions uses as much of the
piano as possible, assuming that the registers are conjunct and evenly dis-
tributed (as is normal for Babbitt’s practice). The registers are fourteen notes
wide and thus occupy eighty-​four of the piano’s eighty-​eight pitches. This
distribution resembles that which Babbitt would use for many of his later
piano pieces, but unlike those later examples, the eighty-​four notes used are
specifically the lowest eighty-​four notes on the piano. Later pieces will, in-
stead, use the middle eighty-​four notes (cf. Figure 7.14). A0 would not be
available with such a balanced distribution.
Yet another aspect of m. 1 of Post-​Partitions worth noting for its gestural
consequences is the partitioning of its aggregate. Gesture here is partially
driven by dynamics, but it is most vividly a result of register, and the registral
flexibility of the measure is a result of its partitioning. The seemingly poin-
tillistic partitioning of the opening aggregate, with but a single pitch class
in each voice, turns out to be a resource, as those various pitch classes are
brought together to form a coherent gesture. As detailed in Chapter 3, the
development of Babbitt’s compositional practice in the mid-​to-​late 1950s
and early 1960s is, to a significant extent, a story of increasing partitional
variety. The final result, of which Post-​Partitions is an early example, is the
all-​partition array. The use of this array—​in contrast to the regular (34)
aggregates of the trichordal arrays Babbitt uses most often in the late-​1940s
and ’50s—​enables a greatly expanded field of gestural possibility. Each
242 Thinking In and About Music

aggregate presents a unique registral and, thus, potentially gestural profile.

Whether or not this was a deliberate motivation for the composer, the re-
sult is a profound uptick in his music’s gestural interest and vitality. And the
most flexible partition of all—​that in which each line contributes one pc—​is
placed at the head of the array for Post-​Partitions.
Over the course of Post-​Partitions, my ability to bodily entrain to the
music waxes and wanes. At times, the piece is simply too disjunct. Its con-
stant huge leaps sometimes exceed my ability to trace musical lines, and
the music recedes into abstraction. As BaileyShea and Monahan (2018,
46) describe such experiences, rather than imagining “disparate points as
instances of a single object in motion . . . the effect is more like pitches
teleporting from point to point.” But more often than not, an attempt to
seek out embodied experiences where I can find them proves fruitful. My
body responds with a varying and unpredictable degree of embodied en-
gagement that comes in different forms. Sometimes those experiences
are mimetic; sometimes they suggest exterior forces. They are always
unsettling—​in the sense that they never allow me to relax into any ex-
tended patterns of motion familiar from daily life or other music. I find this
to be part of the piece’s allure.

The Gestural Function of Virtuosity

The physical thrill of virtuosity is a particular locus of gestural empathy—​

in Babbitt’s music and more generally.7 Cox’s mimetic hypothesis holds that
listeners can participate in a performer’s exertions, so it stands to reason
that the extraordinary exertions of virtuosic performance can produce ex-
ceptional reactions in listeners. It is worth noting, in this regard, how vir-
tuosic the overwhelming majority of Babbitt’s music is, with most pieces
approaching the fastest rhythmic values realizable by human performers.
There is, needless to say, no aspect of Babbitt’s serial procedures that can rea-
sonably be said to demand virtuosity: the tempi of most of his pieces are not
determined serially, and when they are, it is only tempo relations that are so
determined, not absolute tempo values. Similarly, the dispersal of most of his
pieces across all or nearly all of their instruments’ ranges is not demanded

7 A more extended sense of virtuosity in Babbitt’s music—​including compositional, and not merely

performative, display—​is examined by Leong and McNutt (2005, 2019).

 Babbitt’s Gestural Dialectics 243

by any aspect of serialism. His serial procedures do call for multiple contra-
puntal lines, but their distribution could easily be more compact. The leaps
that characterize the surface of his music seem to be intended at least in part
for their inherent thrill, their virtuosity, their kineticism, the disjunct and
distinctive gestures they create. The acrobatics are, in a sense, the point—​or
at least part of the point.
To my ears, an aspect of Babbitt’s background other than his interest in
serialism shines through in his music’s virtuosity: his youth as a jazz clari-
netist.8 An at-​tempo performance of one of his more dynamic pieces—​say,
Post-​Partitions, Sextets, Tableaux, Reflections, My Ends Are My Beginnings,
Dual, Around the Horn, or Tutte le Corde—​gives the impression of a hot
jazz number, performers sweating, out of breath, with off-​balance starting
and stopping at different speeds, constant flying leaps in every direction,
spanning the whole of the instruments, just managing to keep their musical
vision within the limits of their physicality.9 Yet just as Dixieland jazz is dis-
ciplined by its chord changes, Babbitt’s music is disciplined by his arrays;
the gestures in each are conditioned by the competing priority of harmonic
context.
Babbitt frequently acknowledges the difficulties of his music, but typically
only from a conceptual or linguistic standpoint (e.g., Babbitt 1987, 166ff.). As
to performance difficulties, he only mentions “ensemble intricacies” and the
lack of direct repetition (Babbitt [1970] 2003, 256–​57; 1995, 84). He does not
acknowledge the simple fact that so much of his music is fast. But it is notable
that in his comments on electronics and cognition, a principal source of in-
quiry is in the speeds he could obtain before relationships slipped into inau-
dibility (see, e.g., Babbitt [1991b] 2003, 450). While electronic music may not
be virtuosic in a literal sense, the speed of his purely electronic music thrills
nonetheless, perhaps as a result of a cross-​modal mapping whereby listeners
imaginatively empathize with what it would take to perform the piece.
Yet virtuosity is not simply a function of velocity—​of “notes per square
inch,” as Babbitt (1987, 168) might say. It asks performers to defy bodily lim-
itations and thus can be especially compelling when it requires techniques
that seem unnatural. Large, fast leaps on the piano require a quick motion
of the wrist and elbow, and thus the logic of momentum dictates that the

8 On his time as a jazz musician, see Babbitt 1995, 56–​60.

9 J. K. Randall notes, “what turns [Babbitt] on is stuff that goes [snapping] . . . at a mile a minute”
(Hilferty 2011). Babbitt (1995, 80) acknowledges popular music is an influence on his composition
but declines to specify how.
244 Thinking In and About Music

note following a leap should be loud. When a large leap is followed by a soft
dynamic, a quick motion must be followed by a last-​minute halt. This is chal-
lenging to perform and sounds that way: the tension and control required
is communicated to listeners. There is, moreover, a sense of risk involved, a
sensation that translates into a nervous energy.
Babbitt exploits this possibility in the first interlude of Du, as shown in
Figure 7.6. At the end of “Wiedersehen,” the relatively restrained opening

Figure 7.6. Du, mm. 9–​14. Song Cycle. Copyright © 1951 Mobart Music
Publications. Copyright © renewed. All Rights Reserved. Used by permission
of European American Music Distributors Company, sole U.S. and Canadian
agent for Mobart Music Publications.
 Babbitt’s Gestural Dialectics 245

song, and continuing into the interlude, the pianist begins to jump around
quite frantically. In m. 9 and particularly m. 13, the pianist is asked to make
a series of acrobatic leaps. But all of these leaps are paired with soft-​to-​
moderate dynamics. The unnatural restraint this entails results in a growing
tension, a sensation that builds throughout the interlude before spilling over
into “Wankelmut,” which begins with a doubling of the tempo and loud dy-
namics. “Wankelmut” presents the cycle’s protagonist at his most frenzied,
and the physical energy the pianist and singer expend contributes to the de-
piction of his mania.

Liminal Periodicity and the Threat of Discontinuity

The leaps of Figure 7.6—​to take only one instance; further examples
could be found in nearly every minute of Babbitt’s output—​exemplify a
signal threat to the perception of gesture: discontinuity. As Hatten (2004,
97) explains, gesture requires “the subsuming of multiple actions toward
a single goal.” If actions cannot be so subsumed, they will ordinarily not
be felt as contributing to a gesture.10 Discontinuity can happen along any
one of a number of musical axes: registral leaps, differences in intensity,
textural fragmentation, changes of timbre or playing technique, and more.
All can attenuate the perception of gesture. Almost all of Babbitt’s music is
highly discontinuous, to the point that the gestural thread is constantly in
danger of being lost.
Discontinuity is a particular threat to embodiment in the domain of
rhythm. Regular rhythmic periodicity facilitates entrainment and, thus, em-
bodiment. Indeed, Justin London (2006) has found that periodic rhythms
corresponding to regular physical motions such as walking or running are
required for gestural sensation. This claim is too strong: as other authors have
observed (see particularly Kozak 2015), aperiodic music may also inspire a
sense of gesture, as can periodicities faster or slower than those within the
range London cites as pertaining to walking or running.11 But it is only an
overstatement of an important truth: periodicity is a crucial factor in many
embodied musical experiences. Metric regularity in the dance music of many

10 BaileyShea and Monahan 2018, 36ff. similarly note the significance of repetition to embodi-

ment, and the threat to embodiment attending music, such as Babbitt’s, with little direct repetition.
11 Heile 2019 critiques London 2006 on similar grounds.
246 Thinking In and About Music

cultures provides clear evidence of the link between rhythmic periodicity

and bodily motion. And even in gestures organized around registral trajec-
tories, such as those discussed previously, entrainment to a periodic pulse
can be an invaluable aid in hearing continuous impulses that can be tracked
bodily.
A subtle but clear trend toward periodicity can be traced across Babbitt’s
career. With a few exceptions, there is relatively little periodicity in his early
published music (from Composition for Four Instruments through the 1950s),
but periodic rhythms gradually reappear in his music in the 1960s and ’70s.
This may indicate a growing appreciation, on the part of the composer, with
the gestural potential of periodic rhythm, although there is no way to be
sure—​he never publicly opined on embodied cognition.12 At any rate, my
own experience is of heightened gestural sensation in his later music that is
guided by the periodic rhythms I encounter, however fleeting they are.
Figure 7.7 reproduces two passages from About Time. This piece was
written during the brief period in the early 1980s when Babbitt turned
away from the time-​point system and began writing pieces in which regular
quarter notes are consistently divided into even rhythms. In these “pulse
pieces,” not every division of the quarter note need be filled in. In the first
beat of m. 1 of About Time, for instance, the first sixteenth note of the right
hand and the first eighth note of the left hand are omitted. But in many cases,
enough of a string of equal note values is present that one can hear fleeting
bursts of periodicity.
The result is that in About Time, a listener can momentarily entrain onto
repeated short durations. This effect is critical to the piece’s gestural expres-
sion: it imparts a feeling of movement at various different speeds that is con-
stantly starting and stopping. This variability is consistently framed, held in
check, by the pulse. The pulse can be sensed as the beginning and ending
of the strings of equal note values, and in many passages is articulated fre-
quently enough to constitute a layer of metric impulse. In Figure 7.7b, for
instance, every beat is struck.
The quality of motion engendered by the strings of repeated note values
can vary; Figures 7.7a and 7.7b give a sense of the range of possible gestural

12 Surely a significant reason for this must be generational: embodied cognition only became a

significant area of inquiry long after Babbitt’s views were established (with, e.g., Johnson 1987). It did
not enter music theory until near the end of his life, buoyed by feminist scholarship that also did not
emerge until decades after Babbitt’s ideas had formed (see Cusick 1994).
 Babbitt’s Gestural Dialectics 247

Figure 7.7a. About Time, mm. 1–​5

Figure 7.7b. About Time, mm. 30–​33

effects. In Figure 7.7a, the piece’s opening, the two hands share pitch-​class
material. For the first two measures, they present precisely the same se-
quence of pitch classes, somewhat misaligned. Beginning in m. 3, the imi-
tation starts to break apart, but the hands still present the same unordered
pitch-​class sets in nearly every beat of mm. 3–​5. The fragmentary repeated
248 Thinking In and About Music

note values articulating the equivalent pc strings, delivered in a discontin-

uous sequence of bursts, provides the sensation of the two hands chasing—​
and sometimes lurching—​after one another. Figure 7.7b is more fluid. There
is relative continuity in both hands, a sense of sustained motion in parallel.
Sometimes the hands present aligned subdivisions; more often they differ.
The parallel impulses are united at the outset of m. 30 and diverge toward
the end of the measure before occasionally, and momentarily, re-​converging.
The passage hurries along, but smoothly. A soft, ruminative passage in mm.
31–​32 contributes to this effect, providing temporary relief from the leaps
into the outer register.
Following his brief dalliance with evenly divided quarter notes, Babbitt
returned to the time-​ point system in the mid-​ 1980s, settling into a
rhythmic practice he would use relatively consistently thenceforth. In these
later pieces, the time-​point system recedes into the middleground, and
the durations between time points—​much like the quarter notes of About
Time—​are divided evenly. None but the Lonely Flute and String Quartet
No. 6 use this iteration of the time-​point system, as discussed in Chapters 1
and 2. This updated technique may be considered a way to realize the
periodicities developed in the pulse pieces—​and the serial cross-​references
they signify—​within the time-​point system. Whatever the composer’s
motivations, the result of the subdivisons is that the surface, like that of
About Time, is filled with small-​scale periodicity that can facilitate gestural
sensation. In addition, there are numerous larger-​scale periodicities cor-
responding to repeated time points. In virtually all of Babbitt’s later time-​
point pieces, there are a great number of these time-​point repetitions, such
that single time-​point aggregates can be stretched out for a minute or more
of fast and densely realized music. These periodicities too have gestural
effects.
Figure 7.8 provides a score to the opening of Babbitt’s String Quartet No.
6 with time points indicated. Between these time points come numerous
small-​scale repeated note values, creating microbursts of fast periodic move-
ment much like those of the pulse pieces—​and sometimes even slightly larger
successions, such as the quarter notes of m. 16. Time points are articulated using
dynamics, as shown in Figure 7.9. Note that the dynamics assigned to each
time point are applied to the whole ensemble, so that, as in Septet but Equal,
every member of the ensemble plays at the same dynamic level simultaneously.
 Babbitt’s Gestural Dialectics 249

Changes in dynamics thus result in ensemble gestures of intensification and

release.
Arrows in Figure 7.8 point to time-​p oint repetitions; each occurs
after a multiple of twelve sixteenth notes. As can be seen throughout
the example, these repetitions happen with extraordinary frequency. In

Figure 7.8. String Quartet No. 6, mm. 1–​16, with time points marked. Arrows
point to repeated time points.
250 Thinking In and About Music

Figure 7.8. (Continued)

the sixteen measures shown, there are only seven distinct time points
articulated. For much of the second half of the excerpt, only three time
points are repeatedly reiterated: {1, 5, 9}, a time-​p oint “augmented
triad.” Given the three ic4s in {1, 5, 9}, not only is there periodicity
at the length of the twelve-​sixteenth-​note modulus, there is also peri-
odicity at the quarter-​note span connecting the separate time points,
in addition to the continuous microbursts of the subdivisions. String
Quartet No. 6 often settles into such “augmented triads” (this is noted
in Dubiel forthcoming).
 Babbitt’s Gestural Dialectics 251

Figure 7.8. (Continued)

Figure 7.9. First aggregate of the time-​point array for String Quartet No. 6

The frequency of the time-​point repetitions has effects both gestural and
systematic. From a gestural perspective, the period corresponding to the
time-​point modulus—​2.14 seconds—​is repeatedly experienced. To my
ears, there are a few too many confounding events for this to fully amount
to meter—​the repeated spans do not align, are frequently interrupted, and
often do not continue for very long.13 The effect might instead be described

13 Moreover, they are not consistently coordinated into multiple layers of periodicity, a require-

ment many theorists have posited for the perception of meter (see Mirka 2009, 4–​5 and 13–​17; the
idea dates to the eighteenth century).
252 Thinking In and About Music

as having a “liminal periodicity.”14 The piece slips in and out of regu-

larity, dancing with the possibility of regularity but shifting often enough
to never settle into it. There are enough even spans caused by time-​point
repetitions to give the impression of a slow, overarching pulse—​more like a
hypermetric pulse connecting downbeats than a tactus, to use Lerdahl and
Jackendoff ’s (1983, 21) term.15 The divisions spanning time-​point intervals
also often suggest brief bursts of meter. Of the music shown in Figure 7.8,
the quarter notes in m. 16 may be the most salient example. But these bursts
are always temporary, for few consecutive time-​point intervals are spanned
by the same rhythmic values. As with the larger pulses spanning time-​point
repetitions, the faster pulses of the divisions create a sense of liminal pe-
riodicity: the music sometimes offers tantalizing glimpses of regularity
but recedes before any deeply felt metric sense can be established. Perhaps
the most consistently metric passages in the quartet are those in which
the time points are grouped into “augmented triads.” In such passages, a
pulsation on the length of a quarter-​note takes hold, and even something
approaching  meter can be felt—​the quarter-​note durations grouped into
the three-​quarter-​note modulus. But all such passages, too, are temporary,
eventually fading as time points are added or replaced. The numerous levels
of periodicity in Babbitt’s later music all contribute to entrainment, sepa-
rately and occasionally together, but none amount to predictable, steady
pulsation. Listeners are caught in a liminal state—​approaching meter but
never attaining it.
From a systematic perspective, the result of the many time-​ point
repetitions is that progression through the piece’s time-​point array is excep-
tionally slow. The aggregate underlying the passage, shown in Figure 7.9,
spans a full thirty-​two measures, and the rate of progression does not sub-
stantially accelerate later in the piece. As mentioned in Chapter 6, over the
course of the quartet, only thirty-​one time-​point aggregates appear, rend-
ering the piece’s time-​ point structure under-​ complete by twenty-​ seven
aggregates. The quartet achieves the gestural effects accomplished by peri-
odic rhythm at the expense of array completion.

14 I am grateful to Joshua Banks Mailman for suggesting this term.

15 For Lerdahl and Jackendoff, the tactus is the most salient metric level in a tonal piece, the level
at which conductors beat or listeners tap their foot. The time-​point modulus in String Quartet No. 6
is near the limit of the slowest possible beats, according to London 2006, 128. It is slower than most
musical beats, but it can be felt.
 Babbitt’s Gestural Dialectics 253

Serial Anomalies and Text Setting

The periodicity resulting from time-​point repetitions is not the only aspect
Babbitt seems to favor over fidelity to serial expectations. Indeed, in nearly
every one of his pieces, there are several, apparently deliberate, violations of
serial structure:

I have never written a piece within the past twenty years that I didn’t get
letters about in which people say, “I’m sorry but there are 4,892 mistakes in
your piece.” . . . Maybe one or two of them are right; there are misprints or
miscopying or something. But mainly they incorrectly assume if the piece
starts a certain way, it has to go that way, that’s all. Part of the notion of the
piece may have been that certain things change their dependencies, their
contingencies, and their relationships in the course of the piece. Things
change! (Babbitt 1987, 35)

The mention of changing relationships, the hint of multiple compositional

concerns, and the whiff of spontaneity found in this quotation challenge the
vision of orderly outgrowth from an initial impulse that animates Babbitt’s
writing on serialism.16 For the “things” that “change” are not a matter of one
twelve-​tone scheme replacing another—​indeed, the vast majority of serial
anomalies affect only a few, typically three or fewer, notes at a time, after
which the previously established twelve-​tone plan continues apace.17 Most
often, the apparent motivations for these anomalies touch on twelve-​tone
concerns tangentially, if at all. And frequently, the result of serial anomalies
is a more compelling—​longer or smoother—​registral trajectory.18 In such

16 Babbitt’s apparently intentional serial anomalies bear comparison with the practice of other

composers. As Straus 1999 argues, the vast majority of serial anomalies in the music of Stravinsky are
errors that should be corrected. Some Schoenberg scholars (e.g., Cone 1972 and Haimo 1984) have
come to broadly similar conclusions about serial anomalies in Schoenberg’s music. Others (e.g., Hall
1975, Glofcheskie 1976, and Boss 2014, 259–​60 and 270) disagree, finding, as Hall 1975, 182 puts
it—​and as I find is the case with Babbitt—​that “some of the deviations apparently result from the
precedence of other musical factors over serial procedures.”
17 Perhaps for this reason, Dubiel (1992, 118) also questions “whether these instances are best

described as . . . developmental change,” as Babbitt seems to suggest.

18 Dubiel 1992, 118–​19, notes several serial anomalies and similarly ascribes them to registral

effects. Lake 1986, 103–​6 discusses anomalies in Babbitt from a different perspective.
254 Thinking In and About Music

cases, gestural considerations apparently motivate breaking from serial

expectations.19
A number of established norms come undone about two-​thirds of the way
through It Takes Twelve to Tango, beginning in m. 50; see Figure 7.10. This
brief piano piece uses a slightly narrower registral distribution than most of
Babbitt’s other piano works, but it still tends to engage most of the six octaves
available in most of its measures. In m. 50, however, there is a sudden fo-
cusing in on the lowest register and then a gradual climb back to the high

Figure 7.10. It Takes Twelve to Tango, mm. 50–​60. Extended bar lines indicate
aggregate boundaries.

19 Despite the fact that many serial anomalies seem intentional, there are certainly misprints in his

music, as Babbitt admits. Many of the folders in the Milton Babbitt Collection contain lists of errata
or lists of questions about possible errata prepared by the composer or by performers, copyists, or
analysts. The lists of errata prepared by others were carefully checked, and sometimes augmented, by
the composer, who added notations as to how or whether to fix suggested misprints. Some of these
lists were prepared following the publication of the score; these scores were not subsequently revised,
so the errors unfortunately remain in print.
 Babbitt’s Gestural Dialectics 255

Figure 7.11. Array for It Takes Twelve to Tango, mm. 50–​60. Parentheses
indicate an insertion; brackets indicate serially expected notes that are absent.

register that unfolds over the next seven measures. This elongated gesture
is achieved in part by the choice and disposition of the piece’s all-​partition
array, the relevant portion of which is shown in Figure 7.11: the array’s line-​
pairs are projected so that the unique aggregate limited to a single line-​pair is
placed in the lowest register, and the aggregates that follow successively rise.
This registral climb is extended by two unusual facts about the passage. The
first, and most notable and audible, is rhythmic: most of the piece establishes
a harmonic rhythm of one measure per aggregate, but the aggregates in
mm. 50–​55 and 57–​58 unfold over two measures each, nearly doubling the
timespan over which the rise takes place (see Bernstein 2011 for more on
this aspect of the work). Second is the registral dislocation of {F, B♭} in m. 56,
which appears an octave lower than predicted by the array. This anomaly
makes the gesture smoother by helping to bridge the gap between the lowest
and highest registers.
The late song “Now Evening after Evening” begins with a striking use of
gesture to project text setting. In general, embodied cognition lends a pow-
erful aid to text setting. Energetics can facilitate a listener experiencing the
actions, images, or sensations described in text that is set musically. If a mu-
sical stimulus inspires a bodily sensation, other stimuli that inspire similar
sensations may be brought to mind.
The texted works discussed in prior chapters provide numerous examples
of energetic correspondence with textual meaning. The sensation of Philomel
256 Thinking In and About Music

testing out her new voice in the opening measures of her eponymous compo-
sition is realized by the tape voice’s leaps (see Figure 5.29). She must abandon
her initial safe, stable E5 for a lower register, and leap back up at the end of
each of her explorations. The energetic strain she, along with her mimetically
empathetic listeners, must undergo matches her tentative attempts to redis-
cover her expressivity.
In “The Widow’s Lament in Springtime,” the widow’s sadness is represented
by descending contours—​one is given the impression of slumping shoulders
or a downcast gaze. The opening six notes of the vocal part begin the piece
with this gesture, and each of the three trichords setting “Thirty-​five years
/​I lived with my husband” similarly move downward (Figure 5.5). While
the external beauties of springtime lift her music, and perhaps her spirits,
the piece ends with a vivid depiction of the widow “sinking” into the marsh
(Figure 5.7). The ending of Two Sonnets features an opposite trajectory
(Figure 5.22). The protagonist reaches up to F4, 𝆏𝆏, in searching prayer.
“Now Evening after Evening,” a setting of a poem by Derek Walcott,
provides a further, particularly beautiful example. The poem depicts the sea-
side, with gentle waves rolling in from a calm Atlantic (“the surf demands
attention”). The lulling atmosphere is marked by the hypnotic repetitions
within the opening line, “Now evening after evening after evening”; see
Figure 7.12. This line is set so that pitch repetitions match the repetitions
of the text, creating a wave-​like melodic pattern—​reflecting the calm, au-
tumnal, lamenting mood of the poem, as well as more directly depicting the
motion of the ocean’s waves. As the music proceeds, this contour is general-
ized, with the voice expressing more complex shapes but continuing to evoke
gentle undulations—​typically moving at least two intervals in the same di-
rection before turning, tracing a series of gradual ascents and descents. The
song does not continue in this vein throughout—​a greater emotional pitch in
its second half results in slightly more discontinuity—​but the gestural profile
of the beginning sets the tone for what follows.20
Figure 7.13 gives the vocal portion of the array underlying Figure 7.12 (the
accompanying piano’s lines are omitted).21 As can be seen in both figures,
A appears an octave lower than serially expected. The registral dislocation

20 I am not aware of the composer commenting on this, but the line “poetry transforms reader into

poet,” which appears later in the song, is a marvelous encapsulation of how thinking about art can
become thinking in art.
21 The full array for the passage is T of the first block of Reflections, shown in Peel and Cramer
3
1988, 186.
 Babbitt’s Gestural Dialectics 257

Figure 7.12. “Now Evening after Evening,” vocal part, mm. 1–​11

Figure 7.13. Vocal portion of the array for mm. 1–​11 of “Now Evening after
Evening” (the lines in the piano are omitted). Parentheses indicate an insertion;
brackets indicate a serially expected note that is absent.

seems to be for the sake of maintaining or heightening the pervasive wave-​

like contour of the melody. Had the pitch class appeared in the register ex-
pected, the contour would be less continuous and ultimately less appropriate
for the text. At this point, an aesthetic generalization can be hazarded. In
anomalous moments such as those seen in It Takes Twelve to Tango and “Now
Evening after Evening,” the competing criteria of serialism and gestural sen-
sation seem to be resolved in favor of gesture.

Rhetorical Closing Gestures in Three Late Piano Pieces

In most of Babbitt’s late piano pieces, the piano presents six registers of four-
teen pitches each, spanning the central eighty-​four notes on the piano. Every
258 Thinking In and About Music

register is further divided into its upper and lower twelve notes (with sub-
stantial overlap between the lines within each register). There are a number
of reasons the scheme, shown in Figure 7.14, seems to have been attractive
to Babbitt. The total number of separate spans fits his compositional pro-
clivities: twelve registral lines can present a twelve-​part array (as in Envoi
[1990] and Preludes, Interludes, and Postlude), a pair of six-​line arrays (About
Time), a trio of four-​line arrays (Tutte le Corde and Canonical Form), or a
sextet of separate line-​pairs (Emblems, The Old Order Changeth, and Allegro
Penseroso). The registers of the line-​pairs are as wide as possible, given the
range of the piano and a desire to equalize the span of each register, and the
overall distribution is balanced, unduly favoring neither the upper nor lower
register (cf. Post-​Partitions, as discussed previously). However, the scheme
does leave the lowest and highest two pitches on the instrument undefined.
Nonetheless, in a number of these pieces, Babbitt reaches those outer
pitches at the end, treating motion to extremes as a rhetorical gesture sig-
nifying closure.22 Hatten (2004, 95) defines a “rhetorical gesture” as one
that “disrupt[s]‌or deflect[s] the ongoing musical discourse, contributing
to a contrasting dramatic trajectory.” The addition of these excluded notes
disrupts the flow of discourse, altering it from motion within fixed limits to
motion that exceeds those limits. These anomalous transgressions of registral

Figure 7.14. Registral distribution of most of Babbitt’s late piano music

22 Several of Ligeti’s études use the limits of the piano to similar rhetorical effect. “Vertige” (1990),

for instance, ends just as a descending line reaches A0 and an ascending one reaches C8.
 Babbitt’s Gestural Dialectics 259

structure—​as much of a transgression as there can be, given the generous

gamut already allotted—​disrupt prevailing norms, creating a dramatic tra-
jectory contrasting enough that they contribute to a sensation of closure.23
(What had been is no longer.)
In each case, these highest and lowest notes double expected notes in adjacent
registers. This doubling solves a technical issue: given that serial counterpoint
throughout each piece is presented within strictly defined registral bound-
aries, these outermost pitches have no meaning in the piece’s operative syntax.
Linking them to array-​correct pitches gives them contrapuntal significance. The
doubling also simply draws attention to these unusual sounds: whether or not
a listener recognizes that C8 had been excluded from About Time, for instance,
the concluding C7–​C8 octave is a notably unusual, marked event in the piece
(Figure 7.16b). The effect of the rhetorical gesture is thus heightened.
As Dubiel (1992, esp. 117–​18) has observed, Canonical Form ends with
a long descent—​or rather, a focusing-​in on the bottom of the piano, as the
upper registers are successively dropped. The final section of the piece, shown
in Figure 7.15, features only the array set using the lowest two registers of
the six-​register scheme given in Figure 7.14. In this section, the descent in-
deed exceeds that scheme, pushing downwards to, first, B♭0, repeated sev-
eral times, and then, isolated and   𝆑𝆑, A0, supported by octave doubling and,
a beat later, by a resonant fifth. In sepulchral cannonade, the basement of
Canonical Form is broken open.
That low A—​ the bombastic culmination of the piece’s concluding
descent—​heralds but does not quite deliver the end of the piece. In its final
measures, the piece returns to a more moderate register. Indeed, the D♭3 re-
peated in the piece’s last two measures is the second-​highest pitch available,
given the isolated use of the lowest array. Canonical Form thus realizes an
elongated version of McCreless’s prototypical four-​part gesture (Figure 7.4).
It begins with a gradual ascent over the first few minutes of the piece, a high
point as the uppermost register is activated and developed during the piece’s
broad middle, a long descent as the upper registers disappear, a deep crash
facilitated by the anomalous expansion of the registral gamut to low B♭ and A,
and a concluding rebound, with an ending locally signaled by a characteristic
barrage of repeated notes.24

23 That rhetorical considerations may justify adjustments to musical syntax resonates with earlier

conceptions of musical rhetoric, notably Christoph Bernhard’s Tractatus compositionis augmentatus

(translated in Bernhard 1973).
24 Dubiel 1992, 91–​120 elaborates on this overall registral trajectory at length.
260 Thinking In and About Music

Figure 7.15. Canonical Form, mm. 304–​17. Anomalous extremes are boxed.

About Time also breaks its registral scheme at the end, but to different ef-
fect. The piece uses a superarray with two six-​line arrays projected, respectively,
using the three upper and three lower line-​pairs shown in Figure 7.14. The piece
presents the blocks of these arrays either together or, sometimes, separately, in
passages that use only the upper or lower half of the piano.25 The last section
of the piece is an upper-​register solo. Therefore, the concluding use of the two
lowest notes, shown in Figure 7.16a, appears in the penultimate section, with
B♭0 appearing in m. 323 and A0 coming in m. 326. The upper extrema appear

25 The piece’s registral progression is shown in Mead 1994, 213.

Figure 7.16a. About Time, mm. 323–​26. Anomalous extremes are boxed.

Figure 7.16b. About Time, mm. 341–​48. Anomalous extremes are boxed.
262 Thinking In and About Music

only later, at the end of the piece as a whole, as shown in Figure 7.16b. B7 appears
in m. 340. C8, the final extreme, is the last sound heard in the piece.
Note that all four extreme notes in About Time are 𝆏𝆏, and the first three
are relatively enmeshed in their surroundings. The effect is subtle, a sharp
contrast to the   𝆑𝆑 of Canonical Form’s concluding octaves. While there is the
overall sensation of expanding outward, the only one of these last notes that
seems to complete a gesture in a local sense is the last. From m. 343 on, the
upper registral limit seems stuck around E7. That pitch is repeated a number
of times, and the two flurries in m. 346 can be read as first trying and failing
to break through that barrier—​although F7 gets its nose under the tent—​and
then succeeding, reaching G7 𝆑𝆑, a success celebrated by immediate repeti-
tion and, then, by the two 𝆑𝆑𝆑 G7s in mm. 347–​48. Having broken through,
C8 at the end falls out, the furthest possible extension of the ascent. The lock
having been broken, the door swings open, the dynamic dropping to 𝆏𝆏 as the
music rises into the mist.
The conceptual metaphor demonstrated at the end of Canonical Form and,
even more vividly, in About Time is, in Candace Brower’s (1997) terms, that
of a “blockage” overcome. The registral boundaries of the distribution shown
in Figure 7.14 define a “container,” a fixed span with a ceiling and a floor. The
container’s limiting effect may be only dimly felt during the course of the
piece—​with so much on the table, only a glutton would desire more—​but the
concluding expansion is just enough to signal that the container is broken.
The long descent ending Canonical Form (only the final stage of which is
shown in Figure 7.15) can be taken as building up a “force of propulsion” that
overcomes that blockage (44). The strident repetitions of the low octaves give
evidence of the energy expended in their attainment, and the concluding re-
bound may be read as a reassertion of prior norms. About Time is much qui-
eter, but less ambiguous. The swirling repetitions, which break through E7
and then continue past G7 to the ultimate C8, similarly build up propulsion,
and there is no rebound. The container is cracked, delivering a rhetorical ges-
ture that conclusively ends the work.
Tutte le Corde uses a comparable technical procedure—​it, too, reaches the
highest and lowest notes on the piano in its concluding measures—​but is a
contrasting case from a gestural and rhetorical perspective. Rather than the
dramatic presentations of Canonical Form and About Time, the four registral
extremes of Tutte le Corde seem to simply appear in the piece’s concluding
bars. The final measures of the piece, shown in Figure 7.17, contain the
excluded notes in quick succession. As in About Time and Canonical Form,
 Babbitt’s Gestural Dialectics 263

Figure 7.17. Tutte le Corde, mm. 297–​303, with anomalous extremes boxed and
(037) trichords indicated

the extreme notes in Tutte le Corde appear with non-​anomalous instances

of the same pitch class (although sometimes that array-​correct pitch is two
octaves away).26 Supported by neither an extensive registral trajectory, as in
Canonical Form, or even a local one, as in About Time, the extreme notes in
Tutte le Corde do not describe a clear gesture. Moreover, without the associ-
ated drama of a blockage overcome, the conceptual metaphor supporting the
gestures of the earlier pieces seems not to be in effect. A container this porous
is, well, not much of a container at all.

26 The {C, E♭, G} simultaneity in m. 298 is indicated 8va in the published score. Given the array and

Babbitt’s sketches for the work (Milton Babbitt Collection, Box 19, Folder 1), I believe this to be a mis-
print and have corrected it in Figure 7.17.
264 Thinking In and About Music

Rather than seeking to understand these extremes as rhetorical gestures—​as

having a role that extends beyond vivifying the local gestures of which they are a
part—​it may be helpful to consider them in light of the closing signals discussed
in the preceding chapter. There are several ways in which the end of Tutte le
Corde seems to embody motion to extremes, and the use of extreme register is
only one of these. Another is a harmonic oddity almost precisely coextensive
with the extreme pitches. From the high C through the note after the low B♭,
the entire passage presents a sequence of (037) trichords, as indicated in Figure
7.17. This outpouring of major and minor triads might be ascribed to tonal rem-
iniscence, but Tutte le Corde does not present any obvious programmatic mo-
tivation for nostalgia or intertextual quotation. It may be better to describe this
as another sort of motion to an extreme—​an extreme focus on a particular har-
mony, one included in the series but not a prior focus in the piece—​precisely co-
incident with the registral extremes also employed in the passage. Babbitt (1987,
97) alludes to a similar narrative strategy in Partitions, whose final section,
uniquely in the piece, is generated entirely from the first trichord of its series,
partially in order to “gain homogeneity” as the piece closes. (More abstractly,
another kind of homogeneity in the final section is that all cross-​references
index a single array—​the lowest of the three present.)
In Canonical Form, About Time, and Tutte le Corde, as in many examples
discussed previously, various aspects of Babbitt’s practice seem to be in con-
flict. Consistent array projection is a virtue from the perspective of serial
clarity; it enables the confident assignment of heard notes to their proper
contrapuntal line. But from a gestural perspective, the static fixity that results
poses a challenge. Coherent gestural threads can be wrought out of multiple
lines, as many examples in this chapter demonstrate. But for certain special
rhetorical effects or in search of compelling registral trajectories, Babbitt
is willing to deviate from serial expectations. Similarly, a time-​point array,
despite the cognitive benefits it may provide in terms of interdimensional
redundancy, trends toward pointillism and a lack of periodicity—​overall,
a degree of discontinuity inconducive to entrainment.27 Extending the
time-​point array mitigates this problem, even at the cost of leaving it
under-​complete.
Given that Babbitt is willing to cast aside aspects of his serial plans for other
effects, an analyst may understandably want to follow suit: to put aside the
arrays and focus on the effects. But the arrays define an interpretive context

27 Aspects of this paragraph, and especially this sentence, are indebted to Dubiel forthcoming.
 Babbitt’s Gestural Dialectics 265

within which various extra-​systematic concerns such as gesture can be un-

derstood as standing in marked relief against a suppressed or dialectically
opposed background.28 The series may still be “continuously, thoroughly,
and utterly influential,” but its influence is sometimes only felt through its
absence. Indeed, as Dubiel (1992, 118) notes, the fact that a serial anomaly
is anomalous may “enhance the impression” of the “alternative criterion”
that justifies deviating from the piece’s norms. And in the many passages
discussed previously that do not result in anomalies but whose gestures are
forged despite a hostile environment, a comprehension of serial context may
heighten appreciation of the gestures that do appear.29 There is more to ad-
mire in Babbitt’s gestures than the kinesthetic empathy they foster. Also re-
markable is the fact that they exist in music rich with musical virtues of a very
different kind.

Conclusion: Babbitt’s Gestural Music

and Formalist Discourse

In light of the preceding, I feel safe in asserting that formalist analysis of

Babbitt’s music—​analysis concerned with identifying serial procedures and
relationships—​remains necessary. But it is also incomplete. Gestural consid-
erations, along with issues of text setting, temporality, and rhetoric discussed
in previous chapters, not only affect the perception of Babbitt’s music, but
they also seem to have mattered enough to the composer that he was willing
to deviate from serial expectations in order to achieve certain effects. And
this suggests that a significant reappraisal of the composer is overdue.
I hope that a touch of polemics may be forgiven. As cited previously, other
analysts have commented on the gestural qualities of Babbitt’s music. But
the dominant discourse on Babbitt—​perhaps not dominant among pro-
fessional theorists, but overwhelmingly so among historical musicologists,
journalists, and in popular accounts—​is to describe Babbitt as the consum-
mate formalist. His music is “mathematical,” “scientific,” “logical,” “cerebral,”
“intellectual.” In these authors’ telling, Babbitt’s is not a music concerned
with embodied sensation, and certainly not a music that prioritizes gesture
or rhetoric to the extent that it would sacrifice serial fidelity.

28 For a related study of suppressed structure, see Rothstein 1991.

29 See Dubiel 1997, 45 for related discussion of Babbitt working “against the grain.”
266 Thinking In and About Music

These reactions may be fostered by inferences from Babbitt’s prose rather

than by his music.30 Consider Susan McClary’s (1989) comments on the
apprehension of Philomel. Philomel is a work programmatically devoted
to embodiment, to heartbreaking effect. The visceral empathy the piece
inspires in its listeners is central to its impact. McClary (1989, 75) notes
that her students “have trouble listening passively to Philomel as yet an-
other instance of serial and electronic manipulation: they have difficulty
achieving the kind of objective intellectual attitude that would permit
them to focus on considerations of sterile compositional technique.” She is
plainly sympathetic to the subjective response of her students, even offering
admiring words about the music’s “theatrical and emotional power.” But
consider the assumptions she makes as she describes the mindset listeners,
“discourage[d]‌” by Babbitt’s writings, should adopt. One should listen “pas-
sively,” “objectively,” “intellectually.” A listener’s focus should be on “compo-
sitional technique,” on “serial . . . manipulation.” Any personal, embodied
listening is apparently something to be avoided in order to grasp the in-
tended purpose of the piece.31
Listeners should not be so discouraged. I experience embodied sensation
in most of Babbitt’s pieces, and I find nothing in Babbitt’s prose to suggest
that one should hear otherwise, were this even possible. Babbitt’s writing
focuses on formalist aspects, and he indeed argues that discourse on music
be limited to verifiable observations. But he nowhere suggests that such for-
malistic limitations apply to his composition, nor does he suggest that the
listening experience itself should be so circumscribed.32
The matter is not that of denying formalist values in Babbitt’s music. The
hierarchical logic of nested exhaustive lists, the analeptic and proleptic cross-​
references, the inter-​dimensional parallelism afforded by the time-​point
system—​all these are also present and, at least in abstract conception, are dis-
tinct from embodiment. The question, rather, is properly understanding how
Babbitt’s compositional designs and the formalist interrelationships they
manifest function within the total compositions that use them.

30 Taruskin 2009, 274–​77 features a particularly lurid example of this conflation.

31 As noted in Heile 2004, 166, McClary may be erroneously conflating Babbitt with some of his
Darmstadt contemporaries.
32 Heile 2004, 173 astutely observes that “new-​musicological critics of modernism seem to have

taken the university composers’ legitimizing discourse at face value, as if it really was an adequate
representation of modernist aesthetic values.”
 Babbitt’s Gestural Dialectics 267

Different pieces suggest different stances. Sometimes, and most neutrally,

Babbitt’s arrays might be thought of as an unrealized set of potentialities.33 It
is in the process of realization that expressive values, such as embodiment,
come into play. In many other passages or works, such as those focused on
in this chapter, a listener might imaginatively conceive of compositions as
arising from a dialectical tension between competing criteria. In these situ-
ations, the assorted expressive values of the realized surface can be imag-
ined as being in tension with the formalist values of the array, as though the
composer is trying to wrest expression out of the array, stitching together
fragments that seem irreconcilably scattered. Finally, there are a few cases in
which compositional designs seem to expressly facilitate embodied experi-
ence. The opening of Philomel, whose serial technique enables the title char-
acter to rediscover her voice, note by note, is a striking instance.
In short, it is a mistake to limit one’s interpretation of entire compositions
to those values that might be read in unrealized serial counterpoint. Analysts
have grown comfortable navigating the syntactic-​semantic divide in tonal
music: the grammatical rules of tonal syntax are not taken as interfering with
a piece’s expression. As Boretz (1992, 256) once quipped: “Is Tchaikovsky’s
Sixth Symphony about B minor or is it about loneliness?” The answer, of
course, is that the whole question is misbegotten: the symphony is not exclu-
sively defined by its grammatical principles nor by its expressive ends. Serial
music can be similarly multifaceted.
An evocative passage by Stephen Peles on Schoenberg’s “Tot,” Op. 48, No.
2 (1933) bears quotation. Peles describes the relation between the abstract
equilibrium of the twelve-​tone aggregate as it exists in the realm of pitch
classes and the messier, yet richer, world of actual compositions—​a world
that is necessarily one of pitches, not pitch classes:

Whatever else it might have been to Schoenberg, to compose with his

method was to give a body to something that in its native form has none, to
make sounds of things that themselves are not sounds, and to move from
an abstract and inherently stable world to one with potential for instability.
(Peles 2004, 83)

33 Stephen Peles (personal communication) has suggested that an array can be thought of as com-

parable to a figured bass. Like a figured bass, an array is a partial specification of a composition that
guarantees certain syntactical considerations are met. A related point is made in Morris 1987, 4.
268 Thinking In and About Music

For Schoenberg, Peles supposes, the distance between an ideally balanced

abstraction and the instability of its realization was a mark of tragedy, a re-
flection of an imperfect world. For Babbitt, the distance between the ide-
alized, symmetrical, abstract counterpoint of his arrays and his resulting
compositions seems to have been an aesthetic challenge: a goad to creative
solutions, a puzzle, even a field of play. The realization of an array “gives a
body to something that in its native form had none,” and gives his listeners’
bodies sensations of movement in a context that seems doomed to—​and
at times approaches—​the condition Cox (2011) describes as mimetic “dis-​
invitation”; that is, resistance to “mimetic engagement.”
Beyond appreciation of this fact, two additional reasons compel investi-
gation of gesture in Babbitt’s music. First, gestural analysis enables—​even
requires—​analysts to consider all aspects of music together, in the multidi-
mensionality that is necessarily how any work presents itself for audition.
Babbitt’s serial procedures assign different functions to the various musical
dimensions—​serial functions (pitch and rhythm) or articulative functions
(e.g., dynamics articulating time-​ point counterpoint)—​ so it is natural
and typical for serial analysis to consider the dimensions separately. But
embodied cognition is affected simultaneously by all musical dimensions;
therefore, an analyst of gesture must consider musical events in their totality.
Second, the array realization and its attendant “instability” is a primary com-
ponent of what distinguishes one of his works from the next. The organicist,
constructional, and cognitive principles discussed in Chapters 1 and 2 are al-
most wholly invariant across Babbitt’s output, and specific serial techniques
are often repeated from one work to the next, particularly in the second half
of his career.34 But the gestural and rhetorical content of his works is dis-
tinct, to the degree that pieces with equivalent arrays can sound utterly dis-
similar (compare Reflections and “Now Evening after Evening”). As a result,
attention to gestural and rhetorical aspects encourages analysts to seek out
the “distinctions between particularity and systematic generality” (Babbitt
[1972] 2003, 286–​87) and thereby to “expose more broadly and deeply the
singularity of a musical work” (Babbitt [1979a] 2003, 375).

34 As Mead 1994 details, many works share equivalent all-​partition arrays. Indeed, I see scant evi-

dence of the “perpetual race to the patent office” mentality Taruskin (2010, 153) ascribes to Babbitt.
Babbitt would adjust his techniques when motivated by some desideratum or another, but he seems
not to have felt compelled to make each piece technically innovative.
 Glossary

Aggregate—​The complete chromatic gamut—​the twelve pitch classes or the twelve time
points—​or an instance of such.
Array—​A contrapuntal serial construct, typically one in which lines comprising series
forms or derived series are sorted into aggregates.
Array, all-​partition—​An array in which every possible partition type is represented, for
however many lines happen to be in that array.
Array, trichordal—​An array with four lines in which each line contributes one trichord
per aggregate. The trichords for these arrays are drawn from trichord orderings in the
underlying series.
Block—​A section of an array in which each line completes a series form. In all-​partition
arrays, block boundaries can be somewhat fuzzy, as lines begin or end their series forms
slightly before or after the beginning or end of a block.
Combinatoriality—​The ability for a series form to be combined with another series form
to create aggregates.
Discrete segments—​Non-​overlapping segments of a series.
Dyad—​A set of two pitch classes, or a set class of such two-​note collections.
Hexachord—​A set of six pitch classes, or a set class of such six-​note collections.
Hexachordal area—​A group of series forms with invariant hexachords.
Hexachords, all-​combinatorial—​Hexachords that can be transposed or inverted into
themselves or their complements, with the result that any series form using them can
be combined with an S, R, RI, or I form to create aggregates. The six all-​combinatorial
hexachords are known by letter names, following a convention introduced in Martino
1961: A = (012345), B = (023457), C = (024579), D = (012678), E = (014589), and
F = (02468t). Reflecting their association with familiar scales, the A-​type is also referred
to as the “chromatic” hexachord and the C-​type as the “diatonic” hexachord.
Interval class (ic)—​A class comprising complementary pitch-​class intervals. They are
named after the smaller of the two intervals and abbreviated “icX.” For example, ic3
contains pitch-​class intervals 3 and 9.
Interval-​class vector—​A six-​digit tabulation of the multiplicity of each ic within a set, or-
dered from ic1 on the left to ic6 on the right.
Invariance—​The preservation of invariants under some transformation.
Invariant—​An object or property of a series or collection that is preserved under some
transformation. Often, the term refers to a set of pitch classes that is shared between
two or more series forms.
272 Glossary

Inversion (I)—​Written In, this transformation inverts pitch classes or time points using
index number n. In (x) = n − x mod 12.
M transformation—​“M” multiplies each pitch-​class or time-​point number by 5. It is syn-
onymous with what some authors refer to as “M5,” the “circle-​of-​fifths” transformation,
or the “circle-​of-​fourths” transformation.
Modulus—​The span defining equivalence for pitch classes or time points. For pitch
classes, the modulus is an octave; for time points, it is a span of twelve time-​point unit
intervals.
Mosaic—​An unordered set of discrete unordered sets partitioning the aggregate.
Order position—​A position within a series; for example, the first order position is the
beginning of the series.
Partition—​A division of a set of objects into discrete parts. In this volume, the term refers
consistently to the division of an aggregate into discrete parts. Partition types are in-
dicated using exponential notation. The notation (34) indicates an aggregate with four
components of three pitch classes (or time points) each. An aggregate with a single
component of twelve pitch classes (or time points) is indicated (121).
Pitch—​A specific frequency. Pitches throughout the volume will be designated with
Acoustical Society of America notation. A4 is the A above middle C.
Pitch class (pc)—​A class of pitches related by one or more octaves. A (in any octave)
is a pitch class. Pitch classes can be written with familiar letter names or numerically.
This volume uses “fixed Do” numbering: C = 0, C♯/​D♭ = 1, D = 2, D♯/​E♭ = 3, E = 4, F = 5,
F♯/​G♭ = 6, G = 7, G♯/​A♭ = 8, A = 9, A♯/​B♭ = t (for 10); and B = e (for 11).
Pitch-​class interval—​The interval between two pitch classes. The pc interval between x
and y is y − x mod 12. The same intervals may also be written by appending + or − to an
interval class, with the minus sign indicating mod 12 complementation. For example,
the pc interval between C and A may be written 9 or −3. Numbers with + and − are used
to represent pitch intervals in some other sources, but in the present volume, this nota-
tion is used exclusively for pitch-​class intervals.
Retrograde (R)—​To present something (such as a series) in reverse order. When a series
is retrograded, its intervals are reversed and inverted.
Retrograde inversion (RI)—​A form of a series in which the pitch classes are retrograded
and inverted. When a series is retrograde-​inverted, its intervals are reversed.
Secondary set—​An aggregate formed by the last hexachord of one series form and the
first hexachord of the series form that follows it.
Segment—​A connected portion of a series. Compare discrete segment.
Series—​A strict simple ordering of pitch classes or time points. This term is synonymous
with what many authors call a “row.” Series, and all other ordered collections, are indi-
cated with angle brackets, < and >. Note: Babbitt uses the term “set” for this concept.
 Glossary 273

Series, derived—​A twelve-​tone series composed of repeated transformations of a seg-

ment, often a trichord.
Series class—​A class of the (usually) forty-​eight series forms related by transposition,
inversion, and/​or retrogression.
Series form—​A member of a series class. The four different types of series forms are la-
beled S (not inverted or retrograded; corresponding to the “P” indication used by some
authors), I (inverted), R (retrograded), and RI (retrograded and inverted). Subscripts
indicate transposition: the subscript attached to an S or I form indicates its first pitch
class; the subscript attached to an R or RI form indicates its last pitch class.
Series, twelve-​tone—​A series composed of a single instance each of the twelve pitch
classes.
Series, underlying—​An “underlying” series is the basic referential series form of a piece.
Set—​A collection, typically of pitch classes, considered without specification of an
ordering. Sets are written between curly brackets. The set {4, 5, 8}, for example, contains
one or more instances of E, F, and G♯ or A♭, in any order. Note: Babbitt uses “set” for what
this volume calls a “series.”
Set class—A class of sets related by transposition or inversion. Set classes are named after
a “prime form,” a member of the class starting on 0, and written with parentheses. For
example, (014) is the set class containing {4, 5, 8}; this can be confirmed by transposing
{4, 5, 8} into {0, 1, 4}.
Subsets/​Supersets—​If a set class can be contained in another set class, it is called a subset
of that set class. If a set class contains a member of another set class, it is called a superset
of that set class.
Superarray—​A structure consisting of the superimposition of multiple arrays.
Tetrachord—​A collection of four pitch classes, or a set class of such four-​note
collections.
Time point—​A metrically defined moment, typically one with significance to a piece’s use
of the time-​point system.
Time-​point class—​A class of time points separated by multiples of the modulus—​most
often, by multiples of twelve sixteenth notes.
Time-​point system—​Babbitt’s principle technique for rhythmic serialism, used in most
of his works after 1960. Time-​point classes are treated analogously to pitch classes in
the twelve-​tone system.
Time-​point unit interval—​The interval between two consecutive time points. In many of
Babbitt’s pieces, the time-​point unit interval is the duration of a sixteenth note.
Transposition—​Written Tn, pitch-​class transposition transposes a group of pitch classes
by pitch-​class interval n. Tn (x) = n + x mod 12.
Trichord—​ A collection of three pitch classes, or a set class of such three-​
note
collections.
274 Glossary

Trichord ordering—​An ordering of a trichord that may be transposed, inverted, and/​or

retrograded. Trichord orderings in Babbitt’s music are extracted from underlying se-
ries and form the basis of trichordal arrays. They are notated with pitch-​class intervals
corresponding to their presentation in the underlying series. For instance, the sequence
<B, E♭, C> instantiates the trichord ordering <+4, −3>. Other forms of that ordering in-
clude the interval sequences <−4, +3>, <+3, −4>, and <−3, +4>, respectively the inver-
sion, retrograde, and retrograde inversion of the original ordering. Trichord orderings
may be transposed to begin on any pitch class.
Twelve-​tone system—​A system of relations in which twelve-​tone series and/​or aggre-
gates form the basic materials.
Urlinie—​In Schenkerian theory, the “fundamental line” forming the melodic background
of a tonal composition.
Ursatz—​In Schenkerian theory, the “fundamental structure” of a tonal composition,
consisting of an Urlinie and a bass arpeggiation.
Weighted aggregate—​When an aggregate has been transformed by the independent in-
version of its components such that the result no longer exhausts the chromatic gamut,
the result is called a weighted aggregate.
 References

Abbreviations
CEMB = The Collected Essays of Milton Babbitt
JMT = Journal of Music Theory
MTO = Music Theory Online
MTS = Music Theory Spectrum
PNM = Perspectives of New Music
TP = Theory and Practice

Abrams, M. H. 1953. The Mirror and the Lamp: Romantic Theory and the Critical Tradition.
London: Oxford University Press.
Adamowicz, Emily. 2011. “Subjectivity and Structure in Milton Babbitt’s Philomel.”
MTO 17/​2.
Adler, Jeremy. 1980. “The Arrangement of the Poems in Stramm’s Du/​Liebesgedichte.”
German Life and Letters 33/​2: 124–​34.
Agawu, V. Kofi. 1987. “Concepts of Closure and Chopin’s Opus 28.” MTS 9: 1–​17.
Alegant, Brian. 1993. “The Seventy-​Seven Partitions of the Aggregate: Analytical and
Theoretical Implications.” PhD diss., The University of Rochester, Eastman School
of Music.
Alegant, Brian, and Don McLean. 2007. “On the Nature of Structural Framing.”
Nineteenth-​Century Music Review 4/​1: 3–​29.
Andreatta, Moreno. 2012. “Mathématiques/​Musique et Philosophie dans la tradition
américaine: la filiation Babbitt/​Lewin.” In Á la lumière des mathématiques et à l’ombre
de la philosophie, ed. Moreno Andreatta, François Nicolas, and Charles Alunni, 51–​74.
Paris: Delatour France.
Anson-​Cartwright, Mark. 2007. “Concepts of Closure in Tonal Music: A Critical Study.”
TP 32: 1–​17.
Arndt, Matthew. 2016. Review of Robert P. Morgan, Becoming Heinrich Schenker: Music
Theory and Ideology. JMT 60/​1: 89–​95.
Arnold, Stephen, and Graham Hair. 1976. “An Introduction and a Study: String Quartet
no. 3.” PNM 14/​2–​15/​1: 155–​86.
Ayer, A. J., ed. 1959. Logical Positivism. New York: The Free Press.
Babbitt, Milton. (1946) 1992. “The Function of Set Structure in the Twelve-​Tone System.”
PhD diss., Princeton University.
Babbitt, Milton. (1949) 2003. “The String Quartets of Bartók.” CEMB, 1–​9.
Babbitt, Milton. (1950a) 2003. Review of Schoenberg et son école and Qu’est ce que la mu-
sique de douze sons? By René Leibowitz. CEMB, 10–​15.
Babbitt, Milton. (1950b) 2003. Review of Polyphonie—​ Revue musicale trimestrielle;
Quatrième cahier: Le Système dodécaphonique. CEMB, 16–​21.
Babbitt, Milton. (1952) 2003. Review of Felix Salzer, Structural Hearing. CEMB, 22–​30.
276 References

Babbitt, Milton. (1955) 2003. “Some Aspects of Twelve-​Tone Composition.” CEMB, 38–​47.
Babbitt, Milton. (1958) 2003. “The Composer as Specialist.” CEMB, 48–​54.
Babbitt, Milton. (1960) 2003. “Twelve-​Tone Invariants as Compositional Determinants.”
CEMB, 55–​69.
Babbitt, Milton. (1961a) 2003. “Past and Present Concepts of the Nature and Limits of
Music.” CEMB, 78–​85.
Babbitt, Milton. (1961b) 2003. “Set Structure as a Compositional Determinant.” CEMB,
86–​108.
Babbitt, Milton. (1962) 2003. “Twelve-​Tone Rhythmic Structure and the Electronic
Medium.” CEMB, 109–​40.
Babbitt, Milton. (1963) 2003. “Reply to George Perle.” CEMB, 141–​46.
Babbitt, Milton. (1964a) 2003. “Remarks on the Recent Stravinsky.” CEMB, 147–​71.
Babbitt, Milton. 1964b. Program note to Philomel. G. Schirmer, Inc. http://​ www.
musicsalesclassical.com/​composer/​work/​25741#.
Babbitt, Milton. (1964c) 2003. “The Synthesis, Perception, and Specification of Musical
Time.” CEMB, 172–​77.
Babbitt, Milton. (1965) 2003. “The Structure and Function of Musical Theory.” CEMB,
191–​201.
Babbitt, Milton. (1966) 2003. “Edgard Varèse: A Few Observations of His Music.” CEMB,
213–​21.
Babbitt, Milton. (1968) 2003. “Three Essays on Schoenberg.” CEMB, 222–​36.
Babbitt, Milton. (1970) 2003. “On Relata I.” CEMB, 237–​58.
Babbitt, Milton. (1972) 2003. “Contemporary Music Composition and Music Theory as
Contemporary Intellectual History.” CEMB, 270–​307.
Babbitt, Milton. (1974) 2003. “Since Schoenberg.” CEMB, 310–​34.
Babbitt, Milton. (1976a) 2003. “Celebrative Speech.” CEMB, 335–​40.
Babbitt, Milton. (1976b) 2003. “Responses: A First Approximation.” CEMB, 341–​66.
Babbitt, Milton. (1979a) 2003. Forward to David Epstein, Beyond Orpheus. CEMB,
371–​76.
Babbitt, Milton. 1979b. Interview in The Music Makers, ed. Deena Rosenberg and Bernard
Rosenberg, 40–​59. New York: Columbia University Press.
Babbitt, Milton. 1981. “The Music Theoretician’s Dilemmas.” Keynote address, Society for
Music Theory, Los Angeles.
Babbitt, Milton. 1983. Interview by Austin Clarkson, New York City, December 14, 1983.
http://​wolpe.org/​page10/​page10.html.
Babbitt, Milton. 1984a. Speaking of Music: Milton Babbitt. Interview by Charles
Amirkhanian. KPFA, November 15, 1984.
Babbitt, Milton. (1984b) 2003. “The More than the Sounds of Music.” CEMB, 383–​87.
Babbitt, Milton. (1985a) 2003. “ ‘All the Things They Are’: Comments on Kern.” CEMB,
395–​98.
Babbitt, Milton. (1985b) 2003. “I Remember Roger.” CEMB, 388–​94.
Babbitt, Milton. 1986a. “An Interview with Milton Babbitt.” Interview with Laura
Karpman. PNM 24/​2: 80–​87.
Babbitt, Milton. 1986b. “Conversations with American Composers: Ev Grimes Interviews
Milton Babbitt.” Music Educators Journal 72/​9: 52–​53 and 58–​66.
Babbitt, Milton. 1986c. Interview by Tim Page. New, Old, and Unexpected. WNYC, April
17, 1986.
Babbitt, Milton. 1987. Words about Music. Edited by Stephen Dembski and Joseph N.
Straus. Madison: University of Wisconsin Press.
 References 277

Babbitt, Milton. 1988. “An Interview with Milton Babbitt.” Interview by Tim Page.
Boulevard 3: 147–​57.
Babbitt, Milton. (1989) 2003. “On Having Been and Still Being an American Composer.”
CEMB, 428–​36.
Babbitt, Milton. 1991a. “A Composer’s View.” Harvard Library Bulletin 2/​1: 123–​32.
Babbitt, Milton. (1991b) 2003. “A Life of Learning.” CEMB, 437–​58.
Babbitt, Milton (1994) 2003. “Brave New Worlds.” CEMB, 459–​65.
Babbitt, Milton. 1995. Interview with William Duckworth. In William Duckworth,
Talking Music: Conversations with John Cage, Philip Glass, Laurie Anderson, and Five
Generations of American Experimental Composers, 52–​93. New York: Schirmer Books.
Babbitt, Milton. 1997. “A Response.” PNM 35/​2: 129–​36.
Babbitt, Milton. 1998a. “In Search of the Ideal Listener.” Interview by Charles Wuorinen.
In Musically Incorrect: Conversations about Music at the End of the Twentieth Century,
ed. Hayes Biggs and Susan Orzel, 23–​41. New York: C.F. Peters.
Babbitt, Milton. 1998b. New Music Series, Piano Concerto no. 2: Works and Process at the
Guggenheim. Interview by Robert Taub. Produced by Mary Sharp Cronson. Video
recording, VHS.
Babbitt, Milton. (1999) 2003. “ ‘My Vienna Triangle at Washington Square,’ Revisited and
Dilated.” CEMB, 466–​87.
Babbitt, Milton. 2000. “Milton Babbitt Talks about Philomel.” Interview by James Gross.
Perfect Sound Forever April 2000 edition.
Babbitt, Milton. 2001. “Milton Babbitt: A Discussion in Twelve Parts.” Interview with
Frank J. Oteri. New Music Box: New Music USA. https://​nmbx.newmusicusa.org/​
milton-​babbitt-​a-​discussion-​in-​12-​parts/​.
Babbitt, Milton. 2002. “Two Discussions with Milton Babbitt.” Interview by James Romig.
http://​www.jamesromig.com/​TEXT_​files/​BabbittDiscussions.pdf.
Babbitt, Milton. 2003. The Collected Essays of Milton Babbitt, ed. Stephen Peles with
Stephen Dembski, Andrew Mead, and Joseph N. Straus. Princeton: Princeton
University Press.
Babbitt, Milton. 2005. Celebrating Milton Babbitt: Works & Process at the Guggenheim.
Interview by Charles Wuorinen. Produced by Mary Sharp Cronson. New York City.
May 23, 2005. 4.75″ videodisc (NTSC), New York Public Library.
Bailey, Christopher. 2002. “Aspects of Sand: An Acousmatic Work with Listener Interface.”
PhD diss., Columbia University.
BaileyShea, Matthew L. 2012. “Musical Forces and Interpretation: Some Thoughts on a
Measure of Mahler.” MTO 18/​3.
BaileyShea, Matthew, and Seth Monahan. 2018. “ ‘Von Anderem Planeten’?: Gesture,
Embodiment, and Virtual Environments in the Orbit of (A)tonality.” In Music,
Analysis, and the Body: Experiments, Explorations, and Embodiments, ed. Nicholas W.
Reyland and Rebecca Thumpston, 31–​50. Leuven: Peeters.
Ball, Phillip. 2011. “Schoenberg, Serialism, and Cognition: Whose Fault if No One
Listens?” Interdisciplinary Science Reviews 36/​1: 24–​41.
Bauer, Marion. 1933. Twentieth Century Music: How It Developed, How to Listen to It.
New York: G. P. Putnam’s Sons.
Beardslee, Bethany, with Minna Zallman Proctor. 2017. I Sang the Unsingable: My Life in
Twentieth-​Century Music. Rochester: University of Rochester Press.
Bell, Eamonn Patrick. 2019. “The Computational Attitude in Music Theory.” PhD diss.,
Columbia University.
278 References

Bernhard, Christoph. 1973. “The Treatises of Christoph Bernhard,” trans. Walter Hilse.
Music Forum 3: 1–​196.
Bernstein, Zachary. 2011. “Duplicated Subdivisions in Babbitt’s It Takes Twelve to Tango.”
MTO 17/​2.
Bernstein, Zachary. 2013. “Hearing, Studying, and Remembering Milton Babbitt at the
CUNY Graduate Center.” TP 37–​38: 289–​97.
Bernstein, Zachary. 2017. “The Problem of Completeness in Milton Babbitt’s Music and
Thought.” MTS 38/​2: 241–​64.
Berry, David Carson. 2016. “Schenkerian Analysis and Anglo-​American Music Criticism
in the 1930s: A Quest for ‘Objectivity’ and a Path toward Disciplinary Music Theory.”
TP 41: 141–​205.
Boretz, Benjamin. (1969) 1995. Meta-​Variations: Studies in the Foundations of Musical
Thought. Red Hook, NY: Open Space.
Boretz, Benjamin. 1992. “(‘starting now, from here, . . .’): three consecutive occasions of
sociomusical reflection.” PNM 30/​1: 249–​61.
Boretz, Benjamin. 2006. “Where Have We Met Before? For a 90th Birthday Celebration of
Milton Babbitt.” PNM 44/​2: 6–​12.
Bortz, Yuri. 2005. “Selective American Perspectives on Issues of Twenty-​First-​Century
Musical Progress.” DMA diss., The Ohio State University.
Boss, Jack. 2014. Schoenberg’s Twelve-​Tone Music: Symmetry and the Musical Idea.
Cambridge: Cambridge University Press.
Boulez, Pierre. (1952) 1968. “Schoenberg Is Dead.” Reprinted in Notes of an Apprenticeship,
trans. Herbert Weinstock, 268–​75. New York: Alfred A. Knopf.
Boyd-​Hurrell, Sophie. 2017. “The ‘Speculative Ear’: Explorations in Adorno and Musical
Modernism.” PhD diss., Melbourne Conservatorium of Music.
Brackett, John. 2003. “The Philosophy of Science as a Philosophy of Music Theory.” PhD
diss., University of North Carolina at Chapel Hill.
Brody, Martin. 1993. “ ‘Music for the Masses’: Milton Babbitt’s Cold War Music Theory.”
The Musical Quarterly 77/​2: 161–​92.
Brower, Candace. 1997. “Pathway, Blockage, and Containment in Density 21.5.” TP
22: 35–​54.
Burkhart, Charles. 1978. “Schenker’s ‘Motivic Parallelisms.’ ” JMT 22/​2: 145–​75.
Burnham, Scott. 2016. “On the Last Measure of Schubert’s String Quintet.” In Music in
Time: Phenomenology, Perception, Performance, ed. Suzannah Clark and Alexander
Rehding, 152–​68. Cambridge, MA: Harvard University Press.
Carlin, Laurence. 2000. “On the Very Concept of Harmony in Leibniz.” Review of
Metaphysics 54/​1: 99–​125.
Carmichael, L., H. P. Hogan, and A. A. Walter. 1932. “An Experimental Study of the Effect
of Language on the Reproduction of Visually Perceived Form.” Journal of Experimental
Psychology 15/​1: 73–​86.
Carnap, Rudolf. (1932) 1996. “The Elimination of Metaphysics through the Logical
Analysis of Language.” Reprinted in Logical Empiricism at its Peak: Schlick, Carnap, and
Neurath, ed. Sahotra Sarkar and trans. Arthur Pap, 10–​32. New York: Garland.
Carnap, Rudolf. (1967) 2003. The Logical Structure of the World and Pseudoproblems in
Philosophy. Translated by Rolf A. George. Peru, IL: Open Court Publishing.
Cavell, Stanley. 2010. Little Did I Know: Excerpts from Memory. Stanford: Stanford
University Press.
 References 279

Chadabe, Joel. 1997. Electric Sound: The Past and Promise of Electronic Music. Upper
Saddle River, NJ: Prentice-​Hall.
Cherlin, Michael. 1988. “Hauptmann and Schenker: Two Adaptations of Hegelian
Dialectics.” TP 13: 115–​31.
Citkowitz, Israel. (1933) 1985. “The Role of Heinrich Schenker.” Reprinted in TP 10/​
1–​2: 17–​22.
Clarke, Suzannah. 2007. “The Politics of the Urlinie in Schenker’s Der Tonwille and Der
freie Satz.” Journal of the Royal Musical Association 132/​1: 141–​64.
Cohn, Richard. 1992. “The Autonomy of Motives in Schenkerian Accounts of Tonal
Music.” MTS 14/​2: 150–​70.
Colson, Daniel. 2015. “Alternate Pitch Structures in the Music of Milton Babbitt.”
Unpublished manuscript.
Cone, Edward T. 1967. “Beyond Analysis.” PNM 6/​1: 33–​51.
Cone, Edward T. 1972. “Editorial Responsibility and Schoenberg’s Troublesome
‘Misprints.’ ” PNM 11/​1: 65–​75.
Cone, Edward T. 1976. “Yet Once More, O Ye Laurels.” PNM 14/​2–​15/​1: 294–​306.
Cook, Nicholas. 1995. “Heinrich Schenker, Polemicist: A Reading of the Ninth Symphony
Monograph.” Music Analysis 14/​1: 89–​105.
Cox, Arnie. 2011. “Embodying Music: Principles of the Mimetic Hypothesis.” MTO 17/​2.
Cox, Arnie. 2016. Music and Embodied Cognition: Listening, Moving, Feeling, and
Thinking. Bloomington: Indiana University Press.
Cox, Gareth. 2004. “Blumengruß und Blumenglöckchen: Goethe’s Influence on Anton
Webern.” In Goethe: Musical Poet, Musical Catalyst, ed. Lorraine Byrne, 203–​22.
Dublin: Carysfort Press.
Cusick, Suzanne G. 1994. “Feminist Theory, Music Theory, and the Mind/​Body Problem.”
PNM 32/​1: 8–​27.
Derkert, Jacob. 2008. “Mathematics and Ideology in Modernist Music Theory.” In What
Kind of Theory is Music Theory?: Epistemological Exercises in Music Theory and Analysis,
ed. Per F. Broman and Nora A. Engebretsen, 217–​51. Stockholm: Acta Universitatis
Stockholmiensis.
Dubiel, Joseph. 1990a. Review Essay: Composition with Pitch-​ Classes: A Theory of
Compositional Design, by Robert Morris. Journal of Musicological Research 10: 47–​80.
Dubiel, Joseph. 1990b. “Three Essays on Milton Babbitt (Part One).” PNM 28/​2: 216–​61.
Dubiel, Joseph. 1991. “Three Essays on Milton Babbitt (Part Two).” PNM 29/​1: 90–​122.
Dubiel, Joseph. 1992. “Three Essays on Milton Babbitt (Part Three).” PNM 30/​1: 82–​131.
Dubiel, Joseph. 1997. “What’s the Use of the Twelve-​Tone System?” PNM 35/​2: 33–​51.
Dubiel, Joseph. 2005–​6. “Theory as Mindset and as Text: Some Observations.” PNM 43/​
2–​44/​1: 160–​76.
Dubiel, Joseph. 2008. “With Wanton Heed and Giddy Cunning: Milton Babbitt’s Allegro
Penseroso.” In Around Set Theory, ed. Moreno Andreatta, Jean-​Michel Bardez, and John
Rahn, 123–​41. Paris: Editions Delatour.
Dubiel, Joseph. 2012. “In Memory.” MTS 34/​1: 9–​10.
Dubiel, Joseph. Forthcoming. “Pulling a Fast More-​Than-​One: Milton Babbitt’s Time-​
Point Practice.”
Duerkson, Marva. 2008. “Schenker’s Organicism Revisited.” Intégral 22: 1–​58.
Eckardt, Jason. 1997. Review of Andrew Mead, An Introduction to the Music of Milton
Babbitt. Current Musicology 63: 84–​101.
280 References

Eitan, Zohar. 1997. Highpoints: A Study of Melodic Peaks. Philadelphia: University of

Pennsylvania Press.
Ewell, Philip A. 2020. “Music Theory and the White Racial Frame.” MTO 26/​2.
Forte, Allen. 1997. “Milton Babbitt’s ‘Three Theatrical Songs’ in Perspective.” PNM 35/​
2: 65–​84.
Fulmer, Mimmi. 2016. Milton Babbitt’s Vision and Prayer: An Introduction, Performer’s
Guide and Study Score. Edited by Mimmi Fulmer and David Dies, with additional notes
from Bethany Beardslee and Godfrey Winham. Verona, NJ: Subito Music.
Gamer, Carlton. 2012. “Milton at the Princeton Seminars: A Reminiscence.” PNM 49/​
2S: 361–​64.
Gieryn, Thomas F. 1983. “Boundary-​Work and the Demarcation of Science from Non-​
Science: Strains and Interests in Professional Ideologies of Scientists.” American
Sociological Review 48/​6: 781–​95.
Girard, Aaron Robert. 2007. “Music Theory in the American Academy.” PhD diss.,
Harvard University.
Girard, Aaron Robert. 2010. “Music as a (Science as a) Liberal Art at Princeton.” Zeitschrift
der Gesellschaft der Musiktheorie 7: 31–52.
Gleason, Scott. 2013. “Princeton Theory’s Problematics.” PhD diss., Columbia University.
Gleason, Scott. 2015. “Princeton Theory’s Problematic of Solipsism.” Theoria: Historical
Aspects of Music Theory 22: 131–​88.
Glofcheskie, John. 1976. “ ‘Wrong’ Notes in Schoenberg’s Op. 33a.” Studies in Music from
the University of Western Ontario 1: 88–​104.
Goethe, Johann Wolfgang von. 1962. Italian Journey. Translated by W. H. Auden and
Elizabeth Mayer. London: Penguin.
Goethe, Johann Wolfgang von. 2009. The Metamorphosis of Plants. Translated by Douglas
Miller. Reprinted with a new introduction and photography by Gordon L. Miller.
Cambridge, MA: MIT Press.
Goldenberg, Yosef. 2012. “The Interruption-​Fill and Corollary Procedures.” MTO 18/​4.
Goodman, Nelson. 1977. The Structure of Appearance. 3rd ed. Boston: Reidel.
Greitzer, Mary. 2007. “Tormented Voices.” PhD diss., Harvard University.
Guck, Marion. 1994. “Rehabilitating the Incorrigible.” In Theory, Analysis, and Meaning in
Music, ed. Anthony Pople, 57–​76. Cambridge: Cambridge University Press.
Haimo, Ethan. 1984. “Editing Schoenberg’s Twelve-​Tone Music.” Journal of the Arnold
Schoenberg Institute 8/​2: 141–​57.
Hall, Anne C. 1975. “A Comparison of Manuscript and Printed Scores of Schoenberg’s
Violin Concerto.” PNM 14/​1: 182–​96.
Handke, Robert. 1909. “Das Linearprinzip J. S. Bachs.” Bach-​Jahrburch 6: 1–​11.
Harker, Brian. 2008. “Milton Babbitt Encounters Academia (and Vice Versa).” American
Music 26/​3: 336–​77.
Harrison, Daniel. 1988. “Some Group Properties of Triple Counterpoint and Their
Influence on Compositions by J. S. Bach.” JMT 32/​1: 23–​49.
Hatten, Robert. 2004. Interpreting Musical Gestures, Topics, and Tropes: Mozart, Beethoven,
and Schubert. Bloomington: Indiana University Press.
Heile, Björn. 2004. “Darmstadt as Other: British and American Responses to Musical
Modernism.” twentieth-​century music 1/​2: 161–​78.
Heile, Björn. 2019. “New Music Theatre and Theories of Embodied Cognition.” In New
Music Theatre in Europe: Transformations between 1955–​1975, ed. Robert Adlington,
273–​93. London: Routledge.
 References 281

Hempel, Carl. 1958. “The Theoretician’s Dilemma: A Study in the Logic of Theory
Construction.” Minnesota Studies in the Philosophy of Science 2: 173–​226.
Hempel, Carl. 1965. Aspects of Scientific Explanation. New York: Free Press.
Herring, Christoph. 1961. “The Genesis of an Abstract Poem: A Note on August Stramm.”
Modern Language Notes 76/​1: 43–​48.
Hilferty, Robert (director). 2011. Babbitt: Portrait of a Serial Composer, ed. Laura
Karpman. NPR Video.
Hobsbaum, Philip. 1996. Metre, Rhythm and Verse Form. Abingdon: Routledge.
Holbrook, David. (1972) 2013. Dylan Thomas: The Code of Night. London: Bloomsbury.
Hollander, John. 1967. “Notes on the Text of Philomel.” PNM 6/​1: 134–​41.
Howland, Patricia L. 2010. “Formal Processes in Post-​Tonal Music: A Study of Selected
Works by Babbitt, Stockhausen, and Carter.” PhD diss., CUNY Graduate Center.
Hubbs, Nadine. 1991. “Schenker’s Organicism.” TP 16: 143–​62.
Hush, David. 1982–​ 83. “Asynordinate Twelve-​ Tone Structures: Milton Babbitt’s
Composition for Twelve Instruments, Part One.” PNM 21/​1–​2: 152–​208.
Iverson, Jennifer. 2019. Electronic Inspirations: Technologies of the Cold War Musical
Avant-​Garde. New York: Oxford University Press.
Johnson, Michael L. 1972. “Hopkins, Heraclitus, Cosmic Instress, and the Comfort of the
Resurrection.” Victorian Poetry 10/​3: 235–​42.
Johnson, Mark. 1987. The Body in the Mind: The Bodily Basis of Meaning, Imagination, and
Reason. Chicago: University of Chicago Press.
Keiler, Allan. 1989. “The Origins of Schenker’s Thought: How Man is Musical.” JMT 33/​
2: 273–​98.
Kermode, Frank. (1966) 2000. The Sense of an Ending: Studies in the Theory of Fiction.
New York: Oxford University Press.
Keuhn, Mikel. 1995. “The Phenomena of Phonemena: Structure and Realization in
Milton Babbitt’s Work for Soprano and Synthesized Tape.” PhD diss., The University of
Rochester, Eastman School of Music.
Korsyn, Kevin. 1993. “Schenker’s Organicism Reexamined.” Intégral 7: 82–​118.
Kozak, Mariusz. 2015. “Listeners’ Bodies in Music Analysis: Gestures, Motor
Intentionality, and Models.” MTO 21/​3.
Krieger, Murray. 1956. The New Apologists for Poetry. Minneapolis: University of
Minnesota Press.
Kurth, Richard. 2002. “The Art of Cadence in Schönberg’s Fourth Quartet: Metric
Dissonance or Dialectic?” Journal of the Arnold Schönberg Center 4: 245–​70.
Lake, William E. 1986. “The Architecture of a Superarray Composition: Milton Babbitt’s
String Quartet No. 5.” PNM 24/​2: 88–​111.
Lakoff, George, and Mark Johnson. 1980. Metaphors We Live By. Chicago: University of
Chicago Press.
Larson, Steve. 2012. Musical Forces: Motion, Metaphor, and Meaning in Music.
Bloomington: Indiana University Press.
Leibniz, Gottfried Wilhelm. 1898. “Principles of Nature and of Grace, Founded on
Reason.” In The Monadology and Other Philosophical Writings, trans. Robert Latta,
405–​24. Oxford: Clarendon Press.
Leong, Daphne. 2011. “Webs and Snares: Multiple References in Babbitt’s Homily and
Beaten Paths.” MTO 17/​2.
Leong, Daphne, and Elizabeth McNutt. 2005. “Virtuosity in Babbitt’s Lonely Flute.”
MTO 11/​1.
282 References

Leong, Daphne, and Elizabeth McNutt. 2019. “Script versus Structure: Virtuosity
in Babbitt’s Lonely Flute, with Reflections on Process.” In Performing Knowledge:
Twentieth-​Century Music in Analysis and Performance, by Daphne Leong et al., 263–​86.
New York: Oxford University Press.
Lerdahl, Fred, and Ray Jackendoff. 1983. A Generative Theory of Tonal Music. Cambridge,
MA: MIT Press.
Lester, Joel. 1989. Analytic Approaches to Twentieth-​ Century Music. New York: W.
W. Norton.
Levenson, Daniel. 2011. Modernism. New Haven: Yale University Press.
Levinson, Jerrold. 1997. Music in the Moment. Ithaca: Cornell University Press.
Lewin, David. (1986) 2006. “Music Theory, Phenomenology, and Modes of Perception.”
Reprinted in Studies in Music with Text, 75–​151. New York: Oxford University Press.
Lewin, David. (1987) 2011. Generalized Musical Intervals and Transformations.
New York: Oxford University Press.
Lewin, David. (1991) 2006. “Some Problems and Resources of Music Theory.” Reprinted
in Studies in Music with Text, 493–​514. New York: Oxford University Press.
Lewin, David. 2006. Studies in Music with Text. New York: Oxford University Press.
London, Justin. 2006. “Musical Rhythm: Motion, Pace, and Gesture.” In Music and
Gesture, ed. Anthony Gritten and Elaine King, 126–​41. Burlington: Ashgate.
Maggart, Alison. 2017a. “Referential Play in ‘Serious Music’: Allusions to the Past in
Several of Milton Babbitt’s Works from the Late 1980s.” PhD diss., University of
Southern California.
Maggart, Alison. 2017b. “Milton Babbitt’s Glosses on American Jewish Identity.” Current
Musicology 101: 53–​80.
Mailman, Joshua Banks. 2019. “Babbitt’s Beguiling Surfaces, Improvised Inside.” SMT-​V
5/​1–​5/​3.
Mailman, Joshua Banks. 2020. “Portmantonality and Babbitt’s Poetics of Double Entendre.”
MTO 26/​2.
Mandler, George. 2002. “Origins of the Cognitive (R)evolution.” Journal of the History of
the Behavioral Sciences 38/​4: 339–​53.
Margulis, Elizabeth. 2014. On Repeat: How Music Plays the Mind. New York: Oxford
University Press.
Martino, Donald. 1961. “The Source Set and its Aggregate Formations.” JMT 5/​2: 224–​73.
McClary, Susan. 1989. “Terminal Prestige: The Case of Avant-​Garde Composition.”
Cultural Critique 12: 57–​81.
McConville, Brendan. 2011. “A ‘Simple Composition’ of Charles Wuorinen: Isomorphism,
Self-​Similarity, and Nesting in Cello Variations.” TP 36: 151–​77.
McCreless, Patrick. 2006. “Anatomy of a Gesture: From Davidovsky to Chopin and
Back.” In Approaches to Meaning in Music, ed. Byron Almén and Edward Pearsall.
Bloomington: Indiana University Press.
Mead, Andrew. 1983. “Detail and the Array in Milton Babbitt’s My Complements to Roger.”
MTS 5: 89–​109.
Mead, Andrew. 1984. “Recent Developments in the Music of Milton Babbitt.” The Musical
Quarterly 70/​3: 310–​31.
Mead, Andrew. 1987. “About About Time’s Time: A Survey of Milton Babbitt’s Recent
Rhythmic Practice.” PNM 25/​1–​2: 182–​235.
Mead, Andrew. 1994. An Introduction to the Music of Milton Babbitt. Princeton: Princeton
University Press.
 References 283

Mead, Andrew. 1997. “Still Being an American Composer: Milton Babbitt at Eighty.”
PNM 35/​2: 101–​26.
Mead, Andrew. 1999. “Bodily Hearing: Physiological Metaphors and Musical
Understanding.” JMT 43/​1: 1–​19.
Mead, Andrew. 2004. “ ‘One Man’s Signal is Another Man’s Noise’: Personal Encounters
with Post-​Tonal Music.” In The Pleasure of Modernist Music: Listening, Meaning,
Intention, Ideology, ed. Arved Ashby, 259–​74. Rochester: University of Rochester Press.
Mead, Andrew. 2009. “The String Quartets of Milton Babbitt.” In Intimate Voices: The
Twentieth Century String Quartet, vol. 2, ed. Evan Jones, 210–​37. Rochester: University
of Rochester Press.
Mead, Andrew. 2011. “Parallel Processes: Milton Babbitt and ‘Total Serialism.’ ” In
Rewriting Recent Music History: The Development of Early Serialism 1947–​1957, ed.
Mark Delaere, 9–​27. Leuven: Peeters.
Mead, Andrew. 2012. “Making Connections: Some Thoughts about Milton Babbitt and
his Music.” PNM 49/​2S: 378–​89.
Miller, George A. 1956. “The Magical Number Seven, Plus or Minus Two: Some Limits on
Our Capacity for Processing Information.” Psychological Review 63/​2: 81–​97.
Miller, George A. 2003. “The Cognitive Revolution: A Historical Perspective.” Trends in
Cognitive Sciences 7/​3: 141–​44.
Milton Babbitt Collection. Library of Congress, Washington, DC.
Mirka, Danuta. 2009. Metric Manipulations in Haydn and Mozart: Chamber Music for
Strings, 1787–​1791. New York: Oxford University Press.
Monahan, Seth. 2013. “Action and Agency Revisited.” JMT 57/​2: 321–​71.
Morgan, Robert P. 2014. Becoming Heinrich Schenker: Music Theory and Ideology.
Cambridge: Cambridge University Press.
Morris, Robert, and Brian Alegant. 1988. “The Even Partitions in Twelve-​Tone Music.”
MTS 10: 74–​101.
Morris, Robert. 1987. Composition with Pitch-​Classes: A Theory of Compositional Design.
New Haven: Yale University Press.
Morris, Robert. 1993. “Why Weight? Pitch-​Class Duplication in Twelve-​Tone Polyphony.”
Paper presented at the Society for Music Theory National Conference in Montreal,
Quebec.
Morris, Robert. 1997. “Listening to Milton Babbitt’s Electronic Music: The Medium and
the Message.” PNM 35/​2: 85–​99.
Morris, Robert. 2010. The Whistling Blackbird: Essays and Talks on New Music, 3–​26.
Rochester: University of Rochester Press.
Morris, Robert. 2012. “Structure and Infrastructure in Milton Babbitt’s Music: Listening
to Babbitt’s Concerti for Orchestra.” PNM 49/​2S: 390–​408.
Moseley, Brian. 2017. “Sense, Structure, and Webern’s Mysterious Modernism.” TP
42: 81–​100.
Mukherji, Somangshu. 2014. “Generative Musical Grammar—​A Minimalist Approach.”
PhD diss., Princeton University.
Neff, Severine. 1993. “Schoenberg and Goethe: Organicism and Analysis.” In Music
Theory and the Exploration of the Past, ed. Christopher Hatch and David W. Bernstein,
409–​33. Chicago: The University of Chicago Press.
Oppel, Reinhard. 1921. “Zur Fugentechnik Bachs.” Bach-​Jahrbuch 18: 9–​48.
Parkhurst, Bryan. 2013. “Fraught with Ought: An Outline of an Expressivist Meta-​
Theory.” MTO 19/​3.
284 References

Parkhurst, Bryan. 2017. “Making a Virtue of Necessity: Schenker and Kantian Teleology.”
JMT 61/​1: 59–​109.
Pastille, William. 1984. “Heinrich Schenker, Anti-​Organicist.” 19th-​Century Music 8/​
1: 29–​36.
Pastille, William. 1990. “Music and Morphology: Goethe’s Influence on Schenker’s
Thought.” In Schenker Studies, ed. Hedi Siegel, 29–​ 44. Cambridge: Cambridge
University Press.
Pastille, William. 1995a. “The God of Abraham, Aquinas, and Schenker: Art as Faith in an
Age of Unbelief.” Indiana Theory Review 16: 105–​44.
Pastille, William. 1995b. “Schenker’s Value Judgments.” MTO 1/​6.
Payette, Jessica. 2008. “Seismographic Screams: Erwartung’s Reverberations through
Twentieth-​Century Culture.” PhD diss., Stanford University.
Peel, John, and Cheryl Cramer. 1988. “Correspondences and Associations in Milton
Babbitt’s Reflections.” PNM 26/​1: 144–​207.
Peles, Stephen. 1998. “Serialism and Complexity.” In The Cambridge History of American
Music, ed. David Nicholls, 496–​516. Cambridge: Cambridge University Press.
Peles, Stephen. 2001. “Heinrich and Michel’s Excellent Adventure: Review of Leslie
Blasius, Schenker’s Argument and the Claims of Music Theory.” JMT 45/​1: 176–​90.
Peles, Stephen. 2004. “ ‘Ist Alles Eins’: Schoenberg and Symmetry.” MTS 26/​1: 57–​86.
Peles, Stephen. 2012. “Any Theory Is a Choice: Babbitt and the Post-​War Reinvention of
Music Theory.” MTS 34/​1: 22–​24.
Perle, George. 1991. Serial Composition and Atonality: An Introduction to the Music of
Schoenberg, Berg, and Webern. 6th ed. Berkeley: University of California Press.
Peyser, Joan. 1995. The Music of My Time. White Plains, NY: Pro/​Am Music Resources, Inc.
Quantrill, Peter. 2007. Review of Schoenberg Moses und Aron, DVD performance by
Vienna State Opera. Gramophone. Originally published August 2007. https://​www.
gramophone.co.uk/​review/​schoenberg-​moses-​und-​aron.
Quine, Willard Van Orman. 1951. “Two Dogmas of Empiricism.” Philosophical Review
60/​1: 20–​43.
Quinn, Ian. 2007. “General Equal-​Tempered Harmony: Parts 2 and 3.” PNM 45/​1: 4–​63.
Quintillian. 1907. Institutes of Oratory. Translated by John Selby Watson. London: George
Bell and Sons.
Rahn, John. 1976. “How Do You Du (by Milton Babbitt)?” PNM 14/​2–​15/​1: 61–​80.
Reti, Rudolph. 1951. The Thematic Process in Music. New York: Macmillan.
Rothstein, William. 1986. “The Americanization of Heinrich Schenker.” In Theory Only
9/​1: 5–​17.
Rothstein, William. 1991. “On Implied Tones.” Music Analysis 10/​3: 289–​328.
Rufer, Josef. 1954. Composition with Twelve Notes. Translated by Humphrey Searle.
London: Barrie & Jenkins.
Salzer, Felix. 1952. Structural Hearing: Tonal Coherence in Music. New York: Charles Boni.
Sandow, Greg. (1982) 2004. “A Fine Madness.” Reprinted in The Pleasure of Modernist
Music: Listening, Meaning, Intention, Ideology, ed. Arved Ashby, 253–​ 58.
Rochester: University of Rochester Press.
Schachter, Carl. 1981. “Review: A Commentary on Schenker’s Free Composition.” JMT 25/​
1: 115–​42.
Schachter, Carl. 1996. “Schoenberg’s Hat and Lewis Carroll’s Trousers.” In Aflame with
Music: 100 Years of Music at the University of Melbourne, ed. Brenton Broadstock et al.,
327–​41. Melbourne: Centre for Studies in Australian Music.
 References 285

Schachter, Carl. 2001. “Elephants, Crocodiles, and Beethoven: Schenker’s Politics and the
Pedagogy of Schenkerian Analysis.” TP 26: 1–​20.
Schenker, Heinrich. 1954. Harmony. Edited and annotated by Oswald Jonas. Translated
by Elisabeth Mann Borgese. Chicago: The University of Chicago Press.
Schenker, Heinrich. 1979. Free Composition. Translated by Ernst Oster.
New York: Longman.
Schenker, Heinrich. 1996. “Further Considerations of the Urlinie II.” In The
Masterwork in Music 2, ed. William Drabkin and translated by John Rothgeb, 1–​22.
Cambridge: Cambridge University Press.
Schenker, Heinrich. 2004. “Beethoven’s Sonata in F Minor, Op. 2, No. 1.” In Der Tonwille
Volume 1: Issues 1–​5, 72–​95, ed. William Drabkin and translated by Joseph Dubiel.
New York: Oxford University Press.
Schoenberg, Arnold. (1941) 1975. “Composition with Twelve Tones (1).” Reprinted in
Style and Idea, ed. Leonard Stein, 214–​45. Berkeley: University of California Press.
Schoenberg, Arnold. (1949) 1975. “My Evolution.” Reprinted in Style and Idea, ed.
Leonard Stein, 79–​92. Berkeley: University of California Press.
Schoenberg, Arnold. 1995. The Musical Idea and the Logic, Technique, and Art of its
Presentation. Edited by Patricia Carpenter and Severine Neff. Bloomington: Indiana
University Press.
Schubert, Peter N. 1994. “A Multidetermined Moment in Milton Babbitt’s ‘Three
Cultivated Choruses.’ ” TP 19: 57–​82.
Schuijer, Michiel. 2008. Analyzing Atonal Music: Pitch-​Class Set Theory and its Contexts.
Rochester: University of Rochester Press.
Scotto, Ciro. 1988. “Preparing a Performance of Babbitt’s Arie da Capo.” PNM 26/​2: 6–​24.
Scotto, Ciro. 2002. “Review Article: The Conflict Between Particularism and Generalism
in Andrew Mead’s Introduction to the Music of Milton Babbitt.” JMT 46/​1–​2: 285–​345.
Sessions, Roger. (1935) 1979. “Heinrich Schenker’s Contribution.” In Roger Sessions
on Music: Collected Essays, ed. Edward T. Cone, 231–​ 40. Princeton: Princeton
University Press.
Sessions, Roger. (1938) 1979. “Escape by Theory.” In Roger Sessions on Music: Collected
Essays, ed. Edward T. Cone, 256–​62. Princeton: Princeton University Press.
Shannon, Claude E., and Warren Weaver. (1949) 1963. The Mathematical Theory of
Communication. Urbana: University of Illinois Press.
Sheppard, Richard. 1985. “The Poetry of August Stramm: A Suitable Case for
Deconstruction.” Journal of European Studies 15/​4: 261–​94.
Shields, Nathan. 2012. “Remembering Milton.” PNM 49/​2S: 350–​56.
Snarrenberg, Robert. 1994. “Competing Myths: the American Abandonment of
Schenker’s Organicisim.” In Theory, Analysis, and Meaning in Music, ed. Anthony
Pople, 29–​56. Cambridge: Cambridge University Press.
Solie, Ruth. 1980. “The Living Work: Organicism and Music Analysis.” 19th-​Century
Music 4/​2: 147–​56.
Stramm, August. 2013. Gedichte. Berlin: Berliner Ausgabe.
Straus, Joseph N. 1986. “Listening to Babbitt.” PNM 24/​2: 10–​24.
Straus, Joseph N. 1999. “Stravinsky’s Serial ‘Mistakes.’ ” The Journal of Musicology 17/​
2: 231–​71.
Straus, Joseph N. 2003. “Uniformity, Balance, and Smoothness in Atonal Voice Leading.”
MTS 25/​2: 305–​52.
Straus, Joseph N. 2012. “Babbitt the Analyst.” MTS 34/​1: 26–​33.
286 References

Straus, Joseph N. 2018. Broken Beauty: Musical Modernism and the Representation of
Disability. New York: Oxford University Press.
Stravinsky, Igor, and Robert Craft. 1982. Dialogues. Berkeley: University of
California Press.
Sullivan, Erin L. 2005. “Listening to Babbitt: Perception and Analysis of Milton Babbitt’s
Du for Soprano and Piano.” MA thesis, University of Georgia.
Swift, Richard. 1976. “Some Aspects of Aggregate Composition.” PNM 14/​2–​15/​1: 236–​48.
Taruskin, Richard. 2009. “Afterword: Nicht blutbefleckt?” The Journal of Musicology 26/​
2: 274–​84.
Taruskin, Richard. 2010. Music in the Late Twentieth Century. Oxford: Oxford
University Press.
Tindall, William York. 1962. A Reader’s Guide to Dylan Thomas. Syracuse: Syracuse
University Press.
Tommasini, Anthony. 1996. “Finding Still More Life in a ‘Dead’ Idiom.” New York Times,
October 6, 1996.
Uebel, Thomas. 2019. “Vienna Circle.” In The Stanford Encyclopedia of Philosophy (Spring
2019 Edition), ed. Edward N. Zalta. https://​ plato.stanford.edu/​archives/​spr2019/​
entries/​vienna-​circle/​.
Waller, Christopher. 1986. Expressionist Poetry and Its Critics. Leeds: The Institute of
Germanic Studies.
Walliser, Stephan. 1977. ‘That Nature is a Heraclitean Fire and of the Comfort of the
Resurrection’: A Case-​Study in G. M. Hopkins’ Poetry. Bern: A. Francke Verlag.
Walton, Kendall L. (1993) 2015. “Metaphor and Prop Oriented Make-​Believe.” Reprinted
in In Other Shoes: Music, Metaphor, Empathy, and Existence, 175–​95. New York: Oxford
University Press.
Webern, Anton von. 1963. The Path to the New Music, ed. Willi Reich and translated by
Leo Black. Bryn Mawr, PA: Theodore Presser.
Weisse, Hans. (1935) 1985. “The Music Teacher’s Dilemma.” Reprinted in TP 10/​
1–​2: 29–​48.
Westergaard, Peter. 1965. “Some Problems Raised by the Rhythmic Procedures in Milton
Babbitt’s Composition for Twelve Instruments.” PNM 4/​1: 109–​18.
Westergaard, Peter, ed. 1968. “Discussion: What Do You, as a Composer, Try to Get the
Student to Hear in a Piece of Music?” Proceedings of the American Society of University
Composers 1: 59–​81.
Wimsatt Jr., William K. (1954) 1958. The Verbal Icon: Studies in the Meaning of Poetry.
New York: The Noonday Press.
Wimsatt Jr., William K., and Monroe C. Beardsley. 1946. “The Intentional Fallacy.”
Sewanee Review 54/​3: 468–​88.
Winters, Yvors. 1937. Primitivism and Decadence: A Study of American Experimental
Poetry. New York: Arrow Editions.
Wintle, Christopher. 1976. “Milton Babbitt’s Semi-​Simple Variations.” PNM 14/​2–​15/​
1: 111–​54.
Wuorinen, Charles. 1979. Simple Composition. New York: C.F. Peters.
 Index

Figures are indicated by f following the page number

About Time (Babbitt), 212n17, 246–​48, all-​combinatorial hexachords, 15, 25, 53,
247f, 258, 259, 260–​62, 261f, 263, 264 68, 218, 218n22
Abrams, M. H., 5 A-​type, 18n32, 72f, 73, 112, 116, 133,
Adamowicz, Emily, 34 134, 144, 145f, 148, 149–​50, 151, 152,
aggregates, 15, 16, 19 153, 154, 155, 155n23, 156, 157, 160,
Allegro Penseroso, 199, 200 187, 189, 218n22
composite, 209, 209n15 B-​type, 18n32, 73, 134, 145f, 147, 148,
Composition for Four Instruments, 82, 152, 154, 156, 169, 218n22
109, 121 C-​type, 18, 18n32, 72f, 19, 134, 145f,
Composition for Tenor and Six 147, 148, 149, 150, 151, 152, 153,
Instruments, 81, 89, 90f, 91–​93, 154, 155, 155n23, 156, 157, 159, 160,
93n30, 96, 97, 98, 99, 101 188, 218n22
Composition for Twelve Instruments, 86 classification of, 18n32, 218n22
Danci, 209 Composition for Four Instruments, 110,
Du, 148, 149, 150, 151, 155 111, 111f, 112–​13, 115
Glosses, 19 Composition for Tenor and Six
hierarchical development and, 15–​16, Instruments, 88, 90f, 91, 93, 97, 98,
19, 21, 25 102, 104
It Takes Twelve to Tango, 254f, 255 D-​type, 18n32, 88, 91, 93, 98,
Lagniappe, 216 102, 218n22
My Ends Are My Beginnings, 19–​21, 20f Du, 143–​52, 145f, 152, 154, 155, 156–​57,
None but the Lonely Flute, 25, 211 159, 160
“Now Evening after Evening,” 223 E-​type, 18n32, 73, 145f, 152,
Partitions, 73 154, 218n22
Philomel, 193 F-​type, 18n32
Phonemena, 211 Glosses, 18, 19
Piano Concerto No. 2, 214 My Ends Are My Beginnings, 19, 207
Post-​Partitions, 231, 232f, 241–​42 Partitions, 73
Septet but Equal, 239 Philomel, 187–​88, 194
Soli e Duettini for flute and guitar, 211 Two Sonnets, 168, 171
Soli e Duettini for two guitars, 211–​12 “The Widow’s Lament in Springtime,”
“Sounds and Words,” 76 133–​35, 134f
String Quartet No. 6, 248, 252 Allegro Penseroso (Babbitt), 198n4, 199–​201,
Two Sonnets, 169, 171 200f, 201f, 203, 206, 223, 227, 228, 258
weighted, 93n30, 101 “Allmacht” (Babbitt), 142f, 147, 148, 153,
Alegant, Brian, 138, 155 156–​58, 158f, 159, 160
288 Index

all-​partition arrays, 25, 33, 54n21, 68n1, axiomatic logic, 4, 11, 58, 120, 270
91n28, 108, 195, 197–​98, 218,
220–​21, 268n34 Babbitt, Milton
Around the Horn, 211 About Time, 212n17, 246–​48, 247f, 258,
Homily, 224, 225 259, 260–​62, 261f, 263, 264
invention of, 71–​75, 84, 101, 104–​6, Allegro Penseroso, 198n4, 199–​201, 200f,
193, 241–​42 201f, 203, 206, 223, 227, 228, 258
It Takes Twelve to Tango, 255, 255f All Set, 84, 84n17, 86–​87, 87f
in Robert Morris’s music, 75n8 Arie da Capo, 208
Mehr “Du”, 211 Around the Horn, 211, 243
My Ends Are My Beginnings, 20–​21, 207, Ars Combinatoria, 27
208, 211 “As Long as It Isn’t Love,” 126,
None but the Lonely Flute, 21–​23, 22f, 126n2, 127f
25, 31f, 211 “The Autobiography of the Eye,” 195n44
“Now Evening after Evening,” 223, Canonical Form, 212n17, 258, 259, 260f,
256, 257f 262, 263, 264
Phonemena, 210 Clarinet Quintet, 32n51
Piano Concerto No. 2, 213–​16 Composition for Four Instruments (see
Post-​Partitions, 231, 232f, 240–​42 Composition for Four Instruments)
Preludes, Interludes, and Postlude, 211 Composition for Synthesizer, 68–​69n1,
Relata I, 68, 73, 74f, 228 227
Soli e Duettini for flute and guitar, 211 Composition for Tenor and Six
Soli e Duettini for two guitars, 211–​12 Instruments (see Composition for
A Solo Requiem, 203–​4, 206, 211 Tenor and Six Instruments)
String Quartet No. 6, 63–​64, 64f, Composition for Twelve Instruments,
211, 251f 57n24, 84, 86, 86n22, 136n8, 207n11
All Set (Babbitt), 84, 84n17, 86–​87, 87f Concerti for Orchestra, 212n17, 216,
analepsis, 26–​33, 54, 218, 266 224, 227
Anson-​Cartwright, Mark, 229n36 Concerto Piccolino, 32n51
Arie da Capo (Babbitt), 208 Consortini, 207
Around the Horn (Babbitt), 211, 243 Danci, 32n51, 53, 209–​11, 209f,
arrays, 16, 26, 26n41, 34, 219, 233, 243, 210f, 228
267–​68, 267n33. See also all-​partition Du (see “Allmacht”; “Begegnung”; Du;
arrays; superarrays; trichordal arrays “Schwermut”; “Traum”; “Verzweifelt”;
Allegro Penseroso, 199 “Wankelmut”; “Wiedersehen”)
Composition for Tenor and Six Dual, 196–​97, 217, 217n20, 243
Instruments, 76, 88–​99, 92f, 95f, 98f, “The Effect of the Text,” 172n31
99f, 100f, 101f, 103f, 211 Emblems (Ars Emblematica), 36–​39, 37f,
late piano pieces, 258, 259, 264 38f, 45, 46, 50, 51, 228, 258
Philomel, 193, 194f Envoi, 258
Schenker and, 9 Fabulous Voyage, 126, 126n2
“Sounds and Words,” 76, 226 Four Play, 125
Ars Combinatoria (Babbitt), 27, 228 Glosses, 16–​19, 17f, 53, 62, 107,
“As Long as It Isn’t Love” (Babbitt), 126–​28, 207n11, 227
126n2, 127f Groupwise, 54n21, 207, 212n17
Aufbau (Carnap), 41–​44, 58, 60, 66 The Head of the Bed, 54n21, 125, 128n3,
autism, 217n21 129n5, 207, 212n17
“Autobiography of the Eye, The” Homily, 221, 221n27, 224–​26, 225f, 226f
(Babbitt), 195n44 Images, 227
 Index 289

It Takes Twelve to Tango, 227, 254–​55, String Quartet No. 4, 93n30, 208
254f, 255f, 257 String Quartet No. 5, 55n22, 212n17
The Joy of More Sextets, 22n37 String Quartet No. 6 (see String
Lagniappe, 207, 216 Quartet No. 6)
Manifold Music, 32n51, 207n11 Swan Song No. 1, 227
Mehr “Du,” 139n12, 211 Tableaux, 208, 243
More Phonemena, 208, 227 “That Nature is a Heraclitean Fire and of
Music for Four Instruments, 52 the Comfort of the Resurrection,” 85,
My Complements to Roger, 213n17 86f, 162, 163f, 164–​65, 170, 171f
My Ends Are My Beginnings, 19–​21, 20f, Three Compositions for Piano, 57n24,
81n14, 207, 208, 208n12, 211, 221, 203n7, 207n11, 228
221n27, 243 Three Cultivated Choruses, 210–​11
None but the Lonely Flute (see None but Three Theatrical Songs, 126
the Lonely Flute) Triad, 32n51
“Now Evening after Evening,” 195n44, Tutte le Corde, 32n51, 243, 258, 263–​64,
216, 223, 224f, 255–​57, 257f, 268 263f
The Old Order Changeth, 32n51, 258 Two Sonnets (see Two Sonnets)
“Pantun,” 195n44, 216, 227 “The Virginal Book,” 125, 195n44
Paraphrases, 54n21, 207, 207n10, Vision and Prayer (see Vision and
213n17, 227 Prayer)
Partitions, 53, 68, 71–​73, 72f, 227, 264 “A Waltzer in the House,” 128, 128f
Philomel (see Philomel) When Shall We Three Meet Again?,
Phonemena, 211 32n51, 207n11
Piano Concerto No. 2, 32n51, 213–​16, Whirled Series, 222n30
214f, 215f, 217, 227 “The Widow’s Lament in Springtime”
Piano Quartet, 32n51 (see “Widow’s Lament in
Playing for Time, 213n17 Springtime, The”)
Post-​Partitions, 208, 227, 231–​34, 232f, Woodwind Quartet, 207n11
233f, 236, 240–​42, 243, 258 Bach, J. S., 223
Preludes, Interludes, and Postlude BaileyShea, Matthew L., 240, 242
211, 258 Ballade No. 1 (Chopin), 109
Reflections, 93n30, 198, 243, 268 Bartók, Béla, 7, 12–​13, 223
Relata I, 68, 73, 74f, 193, 223, 227, 228 Bauer, Marion, 1
Semi-​Simple Variations, 84, 85f, 227 Beardslee, Bethany, 130, 138, 177
Septet but Equal, 227, 236–​39, 237–​38f, 248 Beethoven, Ludwig van
Sextets, 243 Sonata in F Minor, Op. 2, No. 1,
Soli e Duettini for flute and guitar, 211 10, 10n18
Soli e Duettini for two guitars, 211–​12 Symphony No. 1, 109
A Solo Requiem, 139n12, 203–​6, Symphony No. 2, 117
204f, 205f, 208, 208n12, 211, 221, “Begegnung” (Babbitt), 141f, 144, 153–​54, 155f
221n27, 228 Bernhard, Christoph, 259n23
“Sounds and Words” (see “Sounds and blocks (of an all-​partition array), 30,
Words”) 30n49, 31f, 32, 63, 64f, 224–​25, 240,
“Spelt from Sybil’s Leaves,” 162f, 164–​69, 241, 260
165f, 167f, 168f, 170, 170f Boretz, Benjamin, 233, 267
String Quartet No. 2, 53, 69, 82, 93n30, Boulez, Pierre, 56
102, 108, 207n11, 220 Brahms, Johannes, 4
String Quartet No. 3, 208 Brower, Candace, 262
290 Index

Burnham, James, 41 Cold War, 59n27, 61

Burnham, Scott, 230 combinatoriality, 13, 19. See also all-​
combinatorial hexachords
Canonical Form (Babbitt), 212n17, 258, Composition for Twelve Instruments, 86
259, 260f, 262, 263, 264 Philomel, 187, 188, 189
Carlin, Laurence, 27, 28 “Sounds and Words,” 187n40
Carmichael, L., 60 Two Sonnets, 168
Carnap, Rudolf, 2–​3, 2nn3–​4, 39–​47, 51, 59 completeness, 196–​229, 252
Aufbau, 41–​44, 58, 60, 66 exhaustive (see exhaustive completion
Babbitt’s meeting with, 41 [of a list])
intellectual background of, 40 lack of general systemic model
phenomenalistic construction and, for, 216–​17
41–​47, 52, 58 multiple, 198–​99, 207, 216
caudate sonnets, 162 over-​, 198–​99, 201, 202, 203, 216,
Cavell, Stanley, 2 223, 224
Chomsky, Noam, 49, 49n16 under-​, 198–​99, 199n5, 201, 211–​12,
Chopin, Frédéric, 109 213, 216, 219, 223
chunking, 48–​51, 49n17, 58, 229. See also composite aggregates, 209, 209n15
cognition Composition for Four Instruments
Citkowitz, Israel, 10n18 (Babbitt), 25, 53, 82, 82f, 83f, 89f,
Clarinet Quintet (Babbitt), 32n51 107–​24, 111f, 114f, 115f, 117–​19f,
closing signals, 222–​26, 230. See also 122–​24f, 202–​3f, 223–​24, 227, 246
rhetorical closing gestures completeness and, 198, 198n4, 201–​3
codas, 85, 85n20, 101, 124, 228 generative polyvocality in, 82
Allegro Penseroso, 199–​201, 200f series of, 107–​21
Composition for Four Instruments, 201–​ surface of, 107–​24
3, 223–​24 Composition for Synthesizer (Babbitt),
Homily, 224–​26, 226f 68–​69n1, 227
cognition, 47–​52, 69n1, 130, 235, 243, 264, Composition for Tenor and Six Instruments
268, 269, 270. See also embodiment; (Babbitt), 53, 68–​106, 78f, 80f, 90f,
gestures; kinesthetic empathy; 92f, 94f, 95f, 96f, 97f, 98f, 99f, 100f,
memory; temporality 101f, 103f, 104–​5f, 126n1, 193,
Babbitt’s compositional techniques 201n6, 211
informed by, 52–​57 arrays of (see under arrays; trichordal
completeness and, 196–​97, 212 arrays)
Du and, 161 difficulty of, 68–​71
Emblems (Ars Emblematica) and, generative polyvocality in, 88–​89, 99
37, 38–​39 large-​scale time-​point structure
embodied, 62, 246, 246n12, 268 of, 75–​81
language and, 59–​61 Composition for Twelve Instruments
Leibniz and, 28 (Babbitt), 57n24, 84, 86, 86n22,
metaphorical, 62 136n8, 207n11
Miller and, 47–​49 concatenationism, 219–​20
organicism as cognitive heuristic, 57–​59 concealed repetition, 29
series and, 107, 110 Concerti for Orchestra (Babbitt), 212n17,
time-​point system and, 21 216, 224, 227
cognitive revolution, 49 Concerto for Nine Instruments
Cohn, Richard, 120 (Webern), 18, 53
 Index 291

Concerto Piccolino (Babbitt), 32n51 divine proportion, 75, 75n10

Consortini (Babbitt), 207 Du (Babbitt), 129n5, 130, 138–​61, 144f,
construction, 268, 269 145f, 146f, 149f, 150f, 155f, 158f, 159f,
Carnap and, 39–​47 176, 195, 195n44, 227, 244–​45, 244f.
phenomenalistic, 41–​47, 52, 58 See also “Allmacht”; “Begegnung”;
rational reconstruction, 42, 43, 44, 45, 50 “Schwermut”; “Traum”;
contextual compositions, 13, 13n21, 43–​44 “Verzweifelt”; “Wankelmut”;
continuity, convincing, 69, 70, 75, 75n9, “Wiedersehen”
101, 106 basis of text, 125, 138
counterpoint, 16, 44, 50–​51, 58, 68, 70–​75, bifurcated structure of, 147
87, 88n24, 93n30, 97, 101–​6, 108, identity confusion theme in, 142–​43n16
113, 120, 136, 162, 190, 191, 191n41, interludes of, 151–​52, 151n21,
192, 193, 193f, 199, 212, 231, 234, 156, 244–​45
239, 243, 259, 264, 267 narrative structure of, 140
Cox, Arnie, 235–​36, 239, 242, 268 poems selected for, 140, 141–​42f
cross-​modal mimetic motor imagery, 236 Dual (Babbitt), 196–​97, 217, 217n20, 243
cross-​references, 26–​27, 29–​33, 44, 50, 52, Dubiel, Joseph, 34, 39, 106, 106n36, 110,
54–​55, 62–​66, 208, 218–​19, 266 116, 198, 198n4, 200, 218–​19, 229,
None but the Lonely Flute, 30–​32, 31f 233, 239, 259, 265
Piano Concerto No. 2, 213 dyads
String Quartet No. 6, 63–​66, 248 Composition for Four Instruments, 114
Composition for Tenor and Six
Dallapiccola, Luigi, 223 Instruments, 82, 96, 99
Danci (Babbitt), 32n51, 53, 209–​11, 209f, Philomel, 192
210f, 228 Semi-​Simple Variations, 84
Davidovsky, Mario, 231 Two Sonnets, 85, 171
derived series, 16, 16n30, 44, 52, 53, 68, dynamics, 56–​57, 222
85–​86, 125. See also trichordal arrays; About Time, 262
trichord orderings Allegro Penseroso, 200–​201, 201f, 227
Composition for Four Instruments, 82, 108 Canonical Form, 259
Composition for Tenor and Six Composition for Four Instruments, 83f,
Instruments, 70n2, 80f, 81, 82, 121, 121n9, 122–​24f
89, 101–​2 Composition for Tenor and Six
Composition for Twelve Instruments, 86 Instruments, 77, 104
Du, 148, 149, 151, 161 Du, 149, 157, 158, 161, 244–​45
Glosses, 18, 19 Emblems (Ars Emblematica), 36
Partitions, 73 Homily, 225, 225f
Philomel, 126n1, 184, 188f, 193 None but the Lonely Flute, 23–​24
Semi-​Simple Variations, 85f “Now Evening after Evening,” 223, 224f
Two Sonnets, 85 Post-​Partitions, 231, 240
Vision and Prayer, 175 Septet but Equal, 236–​239
“The Widow’s Lament in Springtime,” Soli e Duettini for two guitars, 211
133, 134, 136, 137 A Solo Requiem, 206, 206n8, 228
Der Tonwille (Schenker), 10 String Quartet No. 6, 248–​49
determinism, 14–​15, 15n25 Two Sonnets, 256
disambiguation, 45n13, 53, 108, 115,
119–​21, 160, 198 echo song, 178, 182, 190
discontinuity, 245–​52, 264 echo verse, 178
292 Index

“Effect of the Text, The” (Babbitt), 172n31 Post-​Partitions, 231–​233, 240–​41

Electronic Study No. 1 (Davidovsky), 231 prototypical four-​part, 231, 234f, 259
electronic synthesis, 46–​47, 68, 70, 75, 177, rhetorical closing, 257–​65
182, 191, 192, 243. See also R.C.A. Septet but Equal, 236–​39
Mark II Electronic Music Synthesizer serial anomalies and, 253–​57, 264–​65
Emblems (Ars Emblematica) (Babbitt), String Quartet No. 6, 248–​52
36–​39, 37f, 38f, 45, 46, 50, 51, 228, 258 Tutte le Corde, 262–​64
embodied cognition, 62, 246, 246n12, 268. Two Sonnets, 256
See also energetics; gesture virtuosity and, 242–​45
empathy, kinesthetic, 235–​40, 265 “The Widow’s Lament in Springtime,” 254
energetics, 234–​42. See also embodied Girard, Aaron Robert, 2
cognition; gesture Glosses (Babbitt), 16–​19, 17f, 53, 62, 107,
enjambment, 132, 134, 176, 176f 207n11, 227
Envoi (Babbitt), 258 Goethe, Johann Wolfgang von, 5–​6, 5n10
epistemology, 44–​45, 45n13, 51–​52 golden ratio. See divine proportion
“Erfüllung” (Stramm), 140 Goodman, Nelson, 41
“Erhört” (Stramm), 140 Grimes, Ev, 49
exhaustive completion (of a list), 25–​26, Group for Contemporary Music, 69
26n38, 33, 44, 52, 106, 197–​99, 200f, Groupwise (Babbitt), 54n21, 207, 212n17
201, 207, 216, 218, 219, 223, 228–​29, Grundgestalt, 30n47
266. See also completeness
Handke, Robert, 6–​7n13
Fabulous Voyage (Babbitt), 126, 126n2 Harmony (Schenker), 7, 48n15
Faulkner, William, 4 Hatten, Robert, 234, 245, 258
formalist discourse, 265–​68. See also Head of the Bed, The (Babbitt), 54n21, 125,
cognition, language and 128n3, 129n5, 207, 212n17
Four Play (Babbitt), 125 Hempel, Carl, 41, 41n5
Free Composition (Schenker), 7, 29 Herring, Christoph, 139
Frege, Gottlob, 40 hexachordal areas, 44, 218
Fulmer, Mimmi, 173 Philomel, 187, 189
Fünf Orchesterstücke (Schoenberg), 50 Post-​Partitions, 241
Schoenberg and, 218, 219n24
Galanter, Eugene, 49, 49n16 String Quartet No. 4 (Schoenberg), 29
generalized gravity, 240 hexachords
generative polyvocality, 82, 99 all-​combinatorial (see all-​combinatorial
gestures, 213, 220, 229, 231–​68 hexachords)
About Time, 246–​48, 260, 262, 264 Philomel, 185, 188
Canonical Form, 259, 264 trichordal generators of, 19, 19n34
defined, 234 hierarchy, 12, 13, 61, 71, 105–​6
discontinuity and, 245–​52, 264 arrays and, 15–26
Du, 244–​45 Carnap and, 44
Emblems (Ars Emblematica), 36 cognition and, 12, 28, 48–​53, 58–​59,
embodied energetics and, 234–​42 120–​21, 196–​97, 216, 217n21, 219,
formalist discourse and, 265–​68 223, 228–​29
It Takes Twelve to Tango, 254–​55 Composition for Four Instruments,
“Now Evening after Evening,” 255–​57, 268 120–​21, 124
periodicity and, 245–​52, 253, 264 Schenker and, 3–​10, 58, 120
Philomel, 182, 255–​56, 266, 267 Hogan, H. P., 60
 Index 293

Holbrook, David, 174 Lanksy, Paul, 10n17

holism, 8, 21, 56, 212 Larson, Steve, 240
Hollander, John, 125, 177, 178, 190, 193 late piano pieces, 257–​65, 258f
Homily (Babbitt), 221, 221n27, 224–​26, Leibniz, Gottfried Wilhelm, 27–​29, 66
225f, 226f Leibowitz, René, 270
Hopkins, Gerard Manley, 85, 85n20, 125, Leong, Daphne, 221, 233
126, 127, 162, 164, 166, 168. See also Lerdahl, Fred, 252
Two Sonnets Levenson, Michael, 4
horizontal weighting, 91n28 Levinson, Jerrold, 219
Lewin, David, 143
Images (Babbitt), 227 liminal periodicity, 252
information processing, 47–​49, 61, 61n32 live-​voice series (of Philomel), 187–​88, 188f, 195
“In Love in Vain” (Kern), 129 logical empiricism, 2, 3, 40, 66, 75
integral serialism, 56 Logical Structure of the World, The
intentional fallacy, 13n23 (Carnap). See Aufbau
interdimensional parallelism, 21, 52, 55–​56, London, Justin, 245
212–​13, 252
interval classes, 99, 113, 114f, 124, 187, 250 Machaut, Guillaume de, 207n9
interval-​class vector, 186 madrigalism, 125, 128, 130, 162, 166, 177
invariance, 43, 44, 47, 52, 69, 71, 84 “Magical Number Seven, Plus or Minus
Composition for Tenor and Six Two, The” (Miller), 47–​49, 52, 53, 55
Instruments, 96, 101–​6 Manifold Music (Babbitt), 32n51, 207n11
Philomel, 182, 191, 192 Martino, Donald, 10n17
Semi-​Simple Variations, 84 maximal diversity principle. See
Two Sonnets, 168 exhaustive completion (of a list)
inversion, 16, 18, 44, 62, 91n27, 192 McClary, Susan, 266–​66
It Takes Twelve to Tango (Babbitt), 227, McCreless, Patrick, 231, 259
254–​55, 254f, 255f, 257 McNutt, Elizabeth, 221, 233
Mead, Andrew, 3, 25, 34, 71, 107, 113, 183,
Jackendoff, Ray, 252 207, 220, 221, 233, 241
jazz, 243 Mehr “Du” (Babbitt), 139n12, 211
Johnson, Mark, 8 memory, 21, 28, 36, 38–​39, 47–​55, 58, 60,
Jonas, Oswald, 10, 10n19 61, 196, 222
Jordan, Stanley, 10n17 metaphor, 3, 7, 8–​10, 14–​15, 62, 70,
Joy of More Sextets, The (Babbitt), 22n37 125, 136n8
Joyce, James, 4 conceptual metaphor, 8, 262–​63
metaphysics, 5n9, 40
Karpman, Laura, 10n17 Miller, George A., 47–​49, 52, 53, 55
Kern, Jerome, 129 mimetic disinvitation, 268
kinesthetic empathy, 235–​40, 265 mimetic hypothesis, 235–​36, 239, 240, 242,
Klavierstück, Op. 33a (Schoenberg), 60, 61 256, 268
Kramer, Jonathan D., 217, 219 modernism, 4, 217n21, 266n32
Krieger, Murray, 13n23 moduli, 21, 23, 57, 57n25, 227
Krosnick, Joel, 217 Composition for Tenor and Six
Kunitz, Stanley, 128 Instruments, 77–​79, 78f, 81
“Sounds and Words,” 77, 77f
Lagniappe (Babbitt), 207, 216 String Quartet No. 6, 65, 250, 251,
Lakoff, George, 8 252, 252n15
294 Index

Monahan, Seth, 6n11, 240, 242 motivic or thematic conceptions of, 6–​8,
Mon fin est mon commencement 7n14, 12–​13, 29, 43–​44, 120
(Machaut), 207n9 roots of concept, 5
Monod, Jacques, 138 Schenker’s, 3–​10, 11, 13, 14–​15, 33,
More Phonemena (Babbitt), 208, 227 57–​59, 197, 269
Morris, Robert, 75n8, 84–​85, 138, 155 time-​point system and, 21, 213
mosaics, 84, 96, 96f, 96n31, 160 Oster, Ernst, 10, 10n19
Mozart, Wolfgang Amadeus, 48, 55, 121 over-​completeness, 198–​99, 201, 202, 203,
M transformation, 155–​56, 155n23, 211 216, 223, 224
multiple completeness, 198–​99, 207, 216
musical forces, 240–​42 palindromes, 201n6
musical idea, 45n13, 108n1 “Pantun” (Babbitt), 195n44, 216, 227
Music Educators Journal, 49 parallelism
Music for Four Instruments (Babbitt), 52 interdimensional, 21, 52, 55–​56, 266
mutual partitioning, 190–​92, 192f, 193 motivic, 29, 29n45, 58
My Complements to Roger musical, 127–​28, 130, 166
(Babbitt), 213n17 phonemic, 126–​30, 139, 166, 167f,
My Ends Are My Beginnings (Babbitt), 19–​ 168, 176
21, 20f, 81n14, 207, 208, 208n12, 211, of process, 26n39
221, 221n27, 243 prolongational, 15
sonic, 127–​28, 166
Naturklang, 7–​8 transformational, 15
Nazi Germany, 1 vocalic, 174
New Criticism, 13n23 Paraphrases (Babbitt), 54n21, 207, 207n10,
New York City, 1–​2, 40 213n17, 227
None but the Lonely Flute (Babbitt), 21–​25, Parkhurst, Bryan, 9
22f, 23f, 24f, 30–​32, 31f, 54, 55, 65n35, partitions, 19–​22, 26, 55, 83–​87, 219.
200, 211, 212n17, 213, 248 See also all-​partition arrays; arrays;
“Now Evening after Evening” (Babbitt), swapping; trichordal arrays
195n44, 216, 223, 224f, 255–​57, Composition for Four Instruments, 201
257f, 268 Composition for Tenor and Six
Instruments, 91–​93, 96, 96n31
octaves, 23, 56, 57n25, 91n28, 93n30, 102, Glosses, 19
192, 193f, 226, 226f, 226n34, 259–​63 My Ends Are My Beginnings, 19–​21
Old Order Changeth, The (Babbitt), Partitions, 73
32n51, 258 Post-​Partitions, 231, 232f, 241–​42
opening techniques, 226–​28 Relata I, 73
Oppel, Reinhard, 6–​7n13 Two Sonnets, 171
order position, 182, 182n37, 184, 185, Partitions (Babbitt), 53, 68, 71–​73, 72f,
186f, 187,, 189 227, 264
organicism, 32, 34–​35, 56, 107, 110, 119–​ Peles, Stephen, 59, 267
20, 121, 235, 268, 270 perception, 12, 15, 21, 38–​39, 45–​46, 47,
aesthetics and, 10, 10n17 60–​61, 62, 71, 75, 75n9, 106n36, 120,
as cognitive heuristic, 57–​59 143n19, 176, 219–​20, 245
completeness and, 197 periodicity, 245–​52, 253, 264
Leibniz’s, 28 Petrarchan sonnet form, 85, 162–​70
metaphysics of, 5n9 phenomenalistic construction, 41–​47,
methodological vs. rhetorical, 9 52, 58
 Index 295

Philomel (Babbitt), 88n24, 126n1, 130, Reflections (Babbitt), 93n30, 198, 243, 268
177–​95, 183–​84f, 185f, 186f, 188f, register, 57, 219, 220, 221, 227, 242, 245,
192f, 193f, 194f, 207n11, 227, 255–​56, 246, 253
266, 267 About Time, 260–​62, 264
basis of text, 125 Allegro Penseroso, 198n4, 199–​201, 200f,
differences between original text and 206, 223, 227
finished score, 178, 190–​91 Canonical Form, 259, 262, 264
principal parts of, 177-​178 Composition for Four Instruments,
program note, 177, 177f 83f, 114f
psychological progression in, 188, 189 Composition for Tenor and Six
rhyme in, 178, 193 Instruments, 91, 93, 97, 99, 102n34
text of, 179–​81f It Takes Twelve to Tango, 254–​55
uniqueness of, 193 late piano pieces, 258–​60, 258f
Phonemena (Babbitt), 211 “Now Evening after Evening,” 256–​57
phonemic parallelism, 126, 128, 129, 139, Post-​Partitions, 231, 233, 236, 240,
166–​68, 167f, 175, 175f, 178 241, 242
Piano Concerto No. 2 (Babbitt), 32n51, Septer but Equal, 236–39
213–​16, 214f, 215f, 217, 227 Tutte le Corde, 263–​64
Piano Quartet (Babbitt), 32n51 Relata I (Babbitt), 68, 73, 74f, 193, 223,
Picker, Tobias, 10n17 227, 228
pitch-​vowel correspondence scheme (of retrograde, 16, 18, 76, 187, 201,
Vision and Prayer), 174–​76, 174f, 175f 201n6, 211–​12
Playing for Time (Babbitt’s), 213n17 retrograde inversion, 16, 18, 36, 76, 91n26
polytonality, 12, 161 revelatory narrative, 107–​8, 108n1
Post-​Partitions (Babbitt), 208, 227, 231–​34, rhetoric (musical), 229, 268, 270. See also
232f, 233f, 236, 240–​42, 243, 258 closing signals; opening techniques;
Preludes, Interludes, and Postlude rhetorical closing gestures
(Babbitt), 211, 258 Du, 151
Princeton University, 41, 41n5, 132 Relata I, 227
prolepsis, 26–​33, 54, 218, 266 Two Sonnets, 169n29
prolongational parallelism, 15 rhetorical closing gestures, 199, 222,
prop oriented make-​believe, 9, 14–​15, 223n31, 229, 257–​59, 264–​65
58, 235 About Time, 260–​62, 264
psychology (Babbitt’s), 61–​67 Allegro Penseroso, 199
pulse pieces, 212–​13, 212n17, 246, 248 Canonical Form, 259, 262, 264
Composition for Four Instruments, 110,
Quine, Willard Van Orman, 41, 41n7 113, 120, 121, 121n8
Quinn, Ian, 124 Tutte le Corde, 262–​64
rhythm, 55–​57, 195n44, 234, 245–​46
Rahn, John, 140 Allegro Penseroso, 200, 201
rational reconstruction, 11, 42, 43, Composition for Four Instruments,
44, 45, 50 201–​3, 202–​3f
R.C.A. Mark II Electronic Music Composition for Tenor and Six
Synthesizer, 68, 87, 194 Instruments, 70, 71, 76
recoding, 48, 49, 49n17, 50, 53, 55, 58, discrepancy between pitches and, 207–​8,
60n29, 107, 110 212–​13
redundancy, 21, 48, 48n14, 50, 51, 52, 53, It Takes Twelve to Tango, 255
54, 55, 56 Leibniz and, 27
296 Index

rhythm (cont.) Composition for Tenor and Six

My Ends Are My Beginnings, 207 Instruments, 73, 82, 89, 91, 98, 99
Philomel, 182 Du, 143, 144, 151, 152, 155, 160
Post-​Partitions, 232 Emblems (Ars Emblematica), 38
sprung, 164 Glosses, 18
Two Sonnets, 164–​67 My Ends Are My Beginnings, 20
Vision and Prayer, 176 None but the Lonely Flute, 30, 32, 55
RICH (RI-​chaining operation), 91n26 Philomel, 187, 188, 189, 190
richness of relation, 27, 34, 46, 52, 67, “The Widow’s Lament in
130, 270 Springtime,” 133
Robin, Leo, 129 Semi-​Simple Variations (Babbitt), 84, 85f, 227
Romanticism, 5, 231, 232 Septet but Equal (Babbitt), 227, 236–​39,
row. See series 237–​38f, 248
Russell, Bertrand, 40 serial anomalies, 253–​65
series, 8, 14, 15–​16, 15nn26–​27, 25, 26n38,
Salzer, Felix, 10, 12–​13 30n49, 34, 44–​45, 53, 54, 54n21, 68,
Sandow, Greg, 233, 236n4 106, 110, 195, 197
scales of measurement, 56, 56n23, 212n16 Allegro Penseroso, 200
Schachter, Carl, 8 arrays and, 15–​26
Schenker, Heinrich, 2–​15, 2nn3–​4, 16, 26, Carnap and, 44, 52
27, 33, 39, 40, 66, 120, 197, 240, 269 Composition for Four Instruments, 107–​21
Babbitt’s first encounter with ideas Composition for Tenor and Six
of, 10–​11 Instruments, 73, 73n7, 77, 82, 85, 88,
Babbitt’s philosophical differences from, 89, 89f, 93, 101–​6
4n7, 5n9, 8–​9, 11 Composition for Twelve Instruments, 86
Der Tonwille, 10 derived (see derived series)
Free Composition, 7, 29 Du, 138, 138n11, 143, 147, 154, 156,
Harmony, 7, 48n15 159, 160, 161
and memory, 57–​59 Emblems (Ars Emblematica), 37f, 38, 51
organicism and (see under organicism) Glosses, 16–​19, 107
Schlick, Moritz, 40 Goethe and, 6n12
Schoenberg, Arnold, 1–​3, 2nn3–​4, 3n5, Homily, 225
4, 6, 13, 39, 40, 54, 59, 62, 199, 218, identifying, 107–​21
219n24, 223, 253n16 My Ends Are My Beginnings, 19, 21
Fünf Orchesterstücke, 50 None but the Lonely Flute, 25, 32
Grundgestalt, conception of 30n47 path to composition from, 14
Klavierstück, Op. 33a 60, 61 Partitions, 73
the musical idea, conception of Philomel, 182, 182n37, 183–​95, 185f
45n13, 108n1 Post-​Partitions, 231, 232f, 233, 240–​41
“Tot,” Op. 48, No. 2, 267 Schoenberg and, 30n47
Schubert, Peter, 210 A Solo Requiem, 204–​6, 228
“Schwermut” (Babbitt), 140, 142f, 147, “Sounds and Words,” 76
153, 157, 160–​61 Tutte le Corde, 264
science, 3, 46, 59–​60. See also cognition Two Sonnets, 168, 168f, 169, 169n28
secondary sets, 17f, 19, 19n33, 20f, 25, 73, 101 “The Widow’s Lament in Springtime,”
segments, 16, 26, 51, 54, 68, 197 133, 136, 137
Composition for Four Instruments, 109, Sessions, Roger, 8, 10, 10n18
111, 113, 114, 115, 116, 120 Sextets (Babbitt), 243
 Index 297

signals. See closing signals subsets, 45n13, 53, 107, 108

simultaneities, 26, 26n41, 37, 54, 55, 88n24 Composition for Four Instruments, 25, 114
Composition for Four Instruments, Composition for Tenor and Six
113–​15, 114f, 117, 119, 121–​24 Instruments, 89
Composition for Tenor and Six Du, 143, 147–​48, 150, 152, 153, 154, 159
Instruments, 88n24, 93, 102 Glosses, 19
Du, 157, 159 Philomel, 191, 192, 192f
Philomel, 182n37 subvocal participation, 236n4
Soli e Duettini for flute and guitar superarrays, 33n54, 101, 211, 221, 228
(Babbitt), 211 About Time, 260
Soli e Duettini for two guitars (Babbitt), 211–​12 Danci, 209, 209f, 209n14, 210f
solipsism, 42, 42n8 Soli e Duettini for two guitars, 211–​12
Solo Requiem, A (Babbitt), 139n12, 203–​6, surface, 7n15, 21, 26–​27, 26n41, 29–​33, 38,
204f, 205f, 208, 208n12, 211, 221, 44, 54–​56, 106, 120–​21, 137, 213, 267
221n27, 228 Swan Song No. 1 (Babbitt), 227
Sonata in F Minor, Op. 2, No. 1 swapping, 83–​85, 85f, 86, 91, 113, 171
(Beethoven), 10, 10n18 Swift, Richard, 212
Sondheim, Stephen, 10n17 Symphony of Psalms (Stravinsky), 13n22
sonic parallelism, 127–​28, 166 Symphony No. 1 (Beethoven), 109
“Sounds and Words” (Babbitt), 70n2, Symphony No. 2 (Beethoven), 117
76–​81, 76nn11–​12, 77f, 79f, 187n40, Symphony No. 40 (Mozart), 48, 55
201n6, 208, 226, 226f, 226n34
“Spelt from Sybil’s Leaves” (Babbitt), 162f, Tableaux (Babbitt), 208, 243
164–​69, 165f, 167f, 168f, 170, 170f tactus, 252, 252n15
Sprachkritik, 4n7 tape-​voice series (of Philomel), 184,
Sprechstimme, 173, 174, 175 184n38, 185f, 186f, 187–​88, 188f,
sprung rhythm, 164 188–​89, 191, 192, 195, 256
Stockhausen, Karlheinz, 56 tempi, 242–​43
Stramm, August, 125, 126, 138, 139–​40, Composition for Tenor and Six
139n12, 148, 157, 160 Instruments, 69, 70, 77–​79, 78f
“Erfüllung,” 140 “Sounds and Words,” 77, 77f, 78–​79
“Erhört,” 140 Two Sonnets, 166
“Werben,” 139 “The Widow’s Lament in Springtime,”
Straus, Joseph N., 124 135–​36
Stravinsky, Igor, 13, 13n22, 137, 223, 253n16 temporality, 62, 197–​98, 235, 265. See also
String Quartet No. 2 (Babbitt), 53, 69, 82, rhetorical closing gestures, verticality
93n30, 102, 108, 207n11, 220 tetrachords
String Quartet No. 3 (Babbitt), 208 Composition for Tenor and Six
String Quartet No. 4 (Babbitt), 93n30, 208 Instruments, 82, 89–​91, 89n25, 96
String Quartet No. 4 (Bartók), 12 Du, 155
String Quartet No. 4 (Schoenberg), 29, Philomel, 190–​92
32n52, 223n31 Semi-​Simple Variations, 84
String Quartet No. 5 (Babbitt), text setting, 125–​95. See also Du; “Now
55n22, 212n17 Evening after Evening”; Philomel; A
String Quartet No. 6 (Babbitt), 32, 32n51, Solo Requiem; Two Sonnets; Vision
33, 63–​66, 63f, 64f, 65f, 211, 212n17, and Prayer; “Widow’s Lament in
248–​52, 249–​51f Springtime, The”
Structural Hearing (Salzer), 10 serial anomalies and, 255–57
298 Index

“That Nature is a Heraclitean Fire and of trichordal arrays, 25, 54n21, 68, 69, 70n3,
the Comfort of the Resurrection” 195, 199, 220
(Babbitt), 85, 86f, 162, 163f, 164–​66, All Set, 86–​87, 87f
170, 171f Composition for Four Instruments,
Thomas, Dylan, 125, 172–​73, 176 82, 109, 111, 111f, 115f, 116, 120,
Three Compositions for Piano (Babbitt), 124, 201
57n24, 203n7, 207n11, 228 Composition for Tenor and Six
Three Cultivated Choruses Instruments, 70, 71–​73, 83–​87, 88–​91,
(Babbitt), 210–​11 90f, 96, 97–​98, 101, 106
Three Theatrical Songs (Babbitt), 126 Danci, 209–​11, 209–​10f
time-​point system, 30n48, 55–​57, 62–​63, Du, 143, 144, 146f, 147, 149, 149f, 151,
208, 221, 227, 253, 266 152–​54, 155–​56, 159
Allegro Penseroso, 200, 201f extending, 83–​87
cognition and, 21, 56–​57, 212 Glosses, 16–​19, 17f
Composition for Tenor and Six Paraphrases, 207
Instruments, 70, 76, 79–​81, 80f, 211 Partitions, 72f, 73
Danci, 209, 209n14, 210, 210f Two Sonnets, 168, 169, 170f, 171
difficulties caused by, 71n4 “The Widow’s Lament in Springtime,”
Homily, 225, 226f 133–​34, 134f, 136
invention of, 21, 21n36, 68 trichord orderings, 54, 106
meter and, 248–​52 Composition for Four Instruments,
My Ends Are My Beginnings, 207, 208 109n3, 111–​15
None but the Lonely Flute, 23–​25, 23f, Composition for Tenor and Six
24f, 30nn47–​48, 32, 213, 248 Instruments, 80f, 81, 88–​102, 103f
Philomel, 182, 183–​84f Du, 143, 143n18, 144f, 146f, 148, 149, 156
Septet but Equal, 237 Emblems (Ars Emblematica), 36–​38,
Soli e Duettini for flute and guitar, 211 37f, 51
Soli e Duettini for two guitars, 211–​12 Glosses, 17f, 18, 62
A Solo Requiem, 204, 205f, 206, 228 Partitions, 72f, 73
“Sounds and Words,” 70n2, 79–​81, 81n16 Two Sonnets, 86, 168, 169
String Quartet No. 6, 63f, 64–​65, 248–​52, trichords, 16n30, 19n34, 52–​54. See also
249–​51f trichord orderings; trichordal arrays
Three Cultivated Choruses, 211 Composition for Four Instruments, 82,
Tindall, William, 173, 174 82–​83f, 109–​24
Tin Pan Alley, 126 Composition for Twelve Instruments, 86
tonal system, 3, 11–​12, 15, 15n26 Du, 86n21, 147–​60
“Tot,” Op. 48, No. 2 (Schoenberg), 267 Music for Four Instruments, 52
total serialism, 56 Partitions, 264
Tractatus compositionis augmentatus Semi-​Simple Variations, 84, 85f
(Bernhard), 259n23 Two Sonnets, 85–​86, 86f
transformational parallelism, 15 Vision and Prayer, 174
transposition, 16, 18, 44, 53, 57 Tutte le Corde (Babbitt), 32n51, 243, 258,
Du, 144, 147 263–​64, 263f
Philomel, 182, 184, 185, 186f, 185n39, twelve-​tone row. See series
187, 188f, 188, 189 twelve-​tone series. See series
“Traum” (Babbitt), 142f, 143, 144, 147, twelve-​tone system, 15–​26, 52–​57, 229,
153, 157, 158–​60, 159f 253, 267, 270
Triad (Babbitt), 32n51 Carnap and, 43–​44
 Index 299

Schenker and, 3, 4, 11, 12, 13–​15 Vivas, Eliseo, 13n23

Schoenberg and, 2, 4, 218, 223n31 vocalic parallelism, 174
Twentieth Century Music (Bauer), 1
Two Sonnets (Babbitt), 53, 85, 86f, 111, Walcott, Derek, 256
161–​72, 165f, 167f, 168f, 170f, 171f, Waller, Christopher, 140
195n44, 207n11, 227, 256. See also Walter, A. A., 60
“Spelt from Sybil’s Leaves”; “That Walton, Kendall, 9
Nature is a Heraclitean Fire and of the “Waltzer in the House, A” (Babbitt),
Comfort of the Resurrection” 128, 128f
basis of text, 125, 162 “Wankelmut” (Babbitt), 139, 141f,
rhyme scheme of, 162, 166–​70 152–​54, 245
Vision and Prayer compared Washington Square College, 41
with, 161–​62, 176-​177 Webern, Anton von, 6
Concerto for Nine Instruments, 18, 53
under-​completeness, 198–​99, 199n5, 201, Variations for piano, Op. 27, 223
211–​12, 213, 216, 219, 223, 252 weighted aggregates, 93n30, 101, 268n34
University of North Carolina, 40 Weisse, Hans, 10n18
University of Pennsylvania, 40 Welch, Roy Dickinson, 132
Urlinie, 11–​12, 14, 240 “Werben” (Stramm), 139
Ursatz, 7–​8, 7n15, 14, 120 Wheelwright, Philip, 41
When Shall We Three Meet Again?
Varèse, Edgard, 7, 13, 223 (Babbitt), 32n51, 207n11
Variations for piano, Op. 27 (Webern), 223 Whirled Series (Babbitt), 222n30
verticality, 217–​22, 228, 229 Whitehead, Alfred North, 40
“Verzweifelt” (Babbitt), 141f, 151, 153, “Widow’s Lament in Springtime,
154–​56, 157 The” (Babbitt), 53, 129n4, 130–​38,
Vienna Circle, 2, 2nn3–​4, 40, 41, 60 133f, 135f, 176–​77, 195, 195n44,
Vienna Triangle, 2–​3, 2n3, 39 227, 256
“Virginal Book, The” (Babbitt), 125, 195n44 basis of text, 125, 130
virtual participation, 235n3 Du compared with, 138, 143
virtuosity, 242–​45 text of, 131f
Vision and Prayer (Babbitt), 68–​69n1, “Wiedersehen” (Babbitt), 141f, 147, 148–​51,
161–​62, 161n25, 172–​77, 176f, 150f, 153, 156, 160, 244–​45
207n11, 208, 227 Williams, William Carlos, 125, 127, 130,
basis of text, 125, 172 132, 136, 137, 138
pitch-​vowel correspondence scheme, Winters, Yvor, 132n6, 135f, 137
174, 174f, 175, 175f Wolpe, Stefan, 69–​70
rhyme in, 175 Woodwind Quartet (Babbitt), 207n11
shape of stanzas, 172, 173, 173f Wuorinen, Charles, 81n16