This article was downloaded by: [University of North Texas]

On: 29 November 2014, At: 12:31

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Contemporary Music Review

Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/gcmr20

Symmetry, the twelve-tone scale,

and tonality
a
George Perle
a
San Francisco Symphony
Published online: 24 Aug 2009.

To cite this article: George Perle (1992) Symmetry, the twelve-tone scale, and tonality,
Contemporary Music Review, 6:2, 81-96, DOI: 10.1080/07494469200640151

To link to this article: http://dx.doi.org/10.1080/07494469200640151

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever
as to the accuracy, completeness, or suitability for any purpose of the Content. Any
opinions and views expressed in this publication are the opinions and views of the
authors, and are not the views of or endorsed by Taylor & Francis. The accuracy
of the Content should not be relied upon and should be independently verified
with primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages, and other
liabilities whatsoever or howsoever caused arising directly or indirectly in connection
with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms
& Conditions of access and use can be found at http://www.tandfonline.com/page/
terms-and-conditions
 Contemporary Music Review, 9 1992 Harwood Academic Publishers GmbH
1992, Vol. 6, Part 2, pp. 81-96 Printed in the United Kingdom
Photocopying permitted by license only

Symmetry, the Twelve-Tone Scale,

and Tonality
George Perle
San Francisco Symphony
Downloaded by [University of North Texas] at 12:31 29 November 2014

My earliest composition in the harmonic language I call "twelve-tone tonality" was published fifty
years ago. The brief explication of this language which follows was written in response to some
queries by David Zinman when he conducted the premiere performance of my Sinfonietta with the
St. Paul Chamber Orchestra.
Several excerpts from this letter which previously appeared in my article, "The First Four Notes of
Lulu," in The Beng Companion, ed. Douglas Jarman (London: Macmillan, 1989) are reprinted here
with the kind permission of Mr. Jarman and the publisher.

F e b r u a r y 6, 1988
D e a r David,
W e ' v e listened to the t a p e a n d f o u n d S u n d a y ' s p e r f o r m a n c e e v e n better t h a n I
h a d realized in the hall.
Enclosed are a f e w of m y arrays. I'll try to explain w h a t this musical l a n g u a g e
that I call "12-tone tonality" is all about. I w a s invited s o m e time ago to write a n
article on the subject for a n e w international c o m p o s e r s ' journal, but h a v e n ' t
b e e n able to take the time a w a y f r o m m y c o m p o s i n g . M a y b e this letter will get
m e g o i n g on it. G o d only k n o w s w h a t it m a y do to you!
W h a t I ' m d o i n g h a s m o r e direct c o n n e c t i o n s with Bart6k a n d Berg t h a n with
a n y other c o m p o s e r s , b u t it also intersects w i t h S c h o e n b e r g a n d Webern, w i t h
Stravinsky all the w a y back to Firebird, w i t h Scriabin, etc. In fact, the v e r y last
e x a m p l e in m y b o o k , Twelve-tone Tonality, is f r o m Chopin.
W e start w i t h the a s s u m p t i o n that intervals can be identical in t w o different
w a y s . O n e of these w a y s is familiar to e v e r y b o d y . C-E, D-F~, G -G, are different
instances of the s a m e interval. We'll call it "interval-4" b e c a u s e that's the
difference in s e m i t o n e s b e t w e e n the t w o notes.
The other k i n d of identity, w h i c h is absolutely basic to Bart6k as well as to the
12-tone system, has to do w i t h axes of s y m m e t r y . C-E b e l o n g s to a family of
s y m m e t r i c a l l y related dyads, as follows:

D D~ E F F# G G#
D C~ C B A~ A G~

T h e s e d y a d s are identical as s u m s . Taking C as O, C# is 1, D is 2, etc. Just as

before w e h a d a c o n s t a n t difference of 4, n o w w e h a v e a constant s u m of 4. F r o m
this p o i n t of v i e w D-F~ a n d G -G are not the s a m e as C-E. D-F~ is a m e m b e r of
the sum-8 family of dyads, a n d E~ -G is a m e m b e r of the sum-10 family. The s u m s
fall into t w o categories, e v e n a n d odd. H e r e , for example, are the sum-3 a n d the
sum-9 series of dyads:
81
 82 GeorgePerle

SUM-3
(C~) D D# E F F~ G (G~)
(D) C~ C B Bb A G~ (G)

SUM-9
(E) F F~ G Ab A Bb (B)
(F) E Eb D Db C B (Bb)

Just to show at once that this stuff really means something and that it has a
significant history, here's a tetrachord that is important in both the Bart6k 4th
quartet and in Lulu: B E F Bb. We can look at this as consisting of two sum-3
Downloaded by [University of North Texas] at 12:31 29 November 2014

dyads, B-E and F-Bb. Both Bart6k and Berg interpret it in this way and relate it
symmetrically to other tetrachords that can be derived by combining two sum-3
dyads. But the same tetrachord can also be interpreted as consisting of two sum-
9 dyads, E-F and Bb-B, and both composers use this alternative interpretation as
well, as a way of connecting two series of symmetrical dyads that are a minor 3rd
apart. I've traced this sort of thing all the way back to Berg's Opus 3 string
quartet (1910). At the beginning of the 2nd mvt. symmetrical relations based on
sum 8 play a dominant role, and a chord that has a special importance here is the
augmented 6th chord consisting of two sum-8 dyads, C-Ab and D-F~. In this
context E (=B~) is the axis of symmetry. This same chord is subsequently
reinterpreted as consisting of two sum-2 dyads, C-D and A~-F#, and through
this reinterpretation we modulate from sum 8 to sum 2 and a new axis of
symmetry, G (=C~):

E F F~ G A~ A Bb
E Eb D Db C B Bb
G Ab A Bb B C C~
G F~ F E Eb D C#

Inversional symmetry, based on the unfolding of pitch classes through a

single sum, is one of the two kinds of symmetry that you can have w h e n you're
dealing with a 12-tone scale. The other kind is based on the unfolding of pitch
classes through a single interval, i.e., an interval cycle. If you do this through
any interval other than 1 (or 11) or 5 (or 7) the pitch classes are distributed into
partitions. Thus we have two different whole-tone (interval-2) cycles, three
different diminished 7th chords (interval-3) cycles, four different augmented
triads (interval-4 cycles), and six different tritones (interval-6 cycles). (It's also
ultimately necessary, for the coherence of the system, and in fact it has import-
ant practical consequences, to say that we have twelve interval-0 cycles.) Both
kinds of symmetry are a natural consequence of the replacement of a diatonic
scale of unequal intervals between scale degrees by a semitonal scale of a single
recurring interval. Diatonic scales can be differentiated as to pitch-class content
and mode. There is only one semitonal scale (as to pitch-class content) and only
one semitonal mode (as to interval structure).
I've arbitrarily laid out my above collections of symmetrically related dyads as
interval-1 cycles, which is the obvious thing to do if you're not defining cyclic
relations. If you will look at the opening bars of the Bart6k 4th quartet you will
immediately see that they are based on the sum-9 collection. If you will look at
 Twelve-ToneTonality 83

the opening bars (following the introduction) of Berg's Lyric Suite you will see
that they too are based on the sum-9 collection, but in this case these dyads are
laid out in the following 12-tone row:

FE CA GD Ab Db Eb Gb Bb Cb

If y o u invert this, beginning on E, you will get the same dyads:

EF AC DG Db Ab Gb Eb Cb Bb
Downloaded by [University of North Texas] at 12:31 29 November 2014

(Tritone transpositions will give you the same dyads, in retrograde order.) But
this is not the inversion that Berg uses. He uses the inversion beginning on B or
F. No w if you will compare this to the prime set beginning on F you will see that
this duplicates another dyadic segmentation:

(F) EC AG DAb Db Eb Gb Bb Cb F
(B) CE GA DAb Eb Db Bb Gb FB

This is the segmentation that you get at the beginning of the recapitulation (cf.
bars 2-11 and 45-47). This alternative segmentation unfolds sum-4 dyads except
at the midpoint, where you have a sum-10 tritone.
Successive statements of the row will give you what Babbitt calls a "secondary
12-tone set" (shown in bold typeface):

F E C A G D Ab Db Eb Gb Bb Cb F E C A G D Ab Db Eb Gb Bb Cb

Which brings us to what I call the "cyclic set." The alternative notes of the cyclic
set unfold the same interval cycle in opposite directions, which is most easily
seen if y o u lay it out this way:

Ab Eb Bb F C G D A E B F~ C~ (G~ . . .
C~ F~ B E A D G C F Bb Eb Ab (Db . . .

This is the same collection of sum-9 dyads that was laid out above in an interval-
1 (or 11) cycle. Here it is laid out in an interval-5 (or 7) cycle. What I call it
d ep en d s on whether I subtract the left element from the right or vice versa. The
former gives us a P-7 cycle for the top line and the latter gives us an I-7 cycle for
the bottom line. Or, if you prefer to define the cyclic interval as 5, right from left
will give you I-5 for the top line and left from right will give you P-5 for the
bottom line. (I consider my figuring out how to name the cycles one of my most
brilliant achievements as a theorist!) You can read the above paired P-7/I-7 cycles
in two ways. If the P-7 cycle gives you the left note of each sum-9 dyad you get a
series whose alternative segmentation gives you sum-4 dyads (the principal
series of the Lyric Suite). If the I-7 cycle gives you the left note of each sum-9 dyad
you get a "cognate" inverted series whose alternative segmentation gives you
sum-2 dyads. Here they are:

Sums 9/4: Ab Db Eb Gb Bb Cb F E C A G D
Sums 9/2: Db Ab Gb Eb Cb Bb E F A C D G
 84 GeorgePerle

If you want the cognate inverted set that Berg uses you have to shift the I-7 cycle
one degree to the right relative to the P-7 cycle:

Ab Eb Bb F C G D A E B F# C~ (G~...
G# C~ F~ B E A D G C F Bb Eb (Ab...

Here are Berg's two sets, cyclically permuted to start with their second
hexachord:

Sums 9/4: Ab Db Eb Gb Bb Cb F E C A G D (D
Downloaded by [University of North Texas] at 12:31 29 November 2014

Sums 11/4: Ab Eb Db Bb Gb F B C E G A D (D

(Just for the record, I have to tell you that it was more than 30 years after I began
to compose with these cyclic sets that I figured out this aspect of the first
movement of the Lyric Suite/)
Cyclic sets can be generated by any pair of inversionally related cycles. The
paired sum-9 semitonal cycles that we started with will give us the following
inversionally related cognate sets:

Sums 8/9: E) E F Eb F~ D G Db A~ C A B Bb (Bb

Sums 10/9: F) F E F~ Eb G D A b Db A C Bb B (B

Here is a cyclic set based on interval-2:

Sums 7/9: Eb E F D G C A Bb B Ab C~ F~ (D~

The sums 10/8 interval-2 set requires two partitions to generat e all twelve notes:

Sums10/8: F) F Eb G C~ A B (B A C~...//

E F~ D Ab C Bb (B~ C . . .

Everything in what I call "12-tone tonality" is based on cyclic sets, but I have
never used these in a way that bears any resemblance whatever to "serialism" or
the so-called "12-tone system", which I regard as an essentially primitive notion,
like playing the Eb scale over and over again in order to compose a piece in Eb.
My very first piece in the system (1940) was based on the following paired
interval-7 sets:

Sums 0/7: C) C G F D Bb A D~ E G~ B C~ F~ (F~

S u m s 1 0 / 3 F) F Bb C Eb G Ab D C~ A F~ E B (B

(I don't want to worry you with this now, but here's a parenthetical note on
nomenclature for future reference to avoid some elementary confusions. These
two sets are inversionally related. If I count up from the even sum of the first I
get 7 and if I count down from the even sum (or up from the odd sum) of the
other I also get 7. I use lower-case letters to name sets and upper case to name
cycles. The two sets above are respectively P0P7 and i3i:0. If I call the interval
 Twelve-Tone Tonality 85

cycle 5 instead of 7 then the two sets are respectively called i7i0 and P10p3.) This
piece [Example 1] is mainly based on the chords that you will get if you align
these two sets so that they share one of the two interval cycles (the shared I-7
cycle is shown in boldface here):

Sums 0/7:
C C G F D B~, A D~ E G~ B C~ F~ F~ C~ B G~ E D~ A Bb D F G (C
Sums 3/10:
Eb C Bb F F Bb C EL, G Ab D C~ A F~ E B B E F~ A C~ D Ab G (Eb
Downloaded by [University of North Texas] at 12:31 29 November 2014

If I subtract each note of the second set from the note above it in the first set I
have alternate intervals of 9 and 0. You will see that each of the unison "axis-
note" dyads has a minor-3rd neighbor on each side of it. The two minor-3rd
dyads together I call a "cyclic chord" and the cyclic chord plus the axis dyad an
"axis-dyad chord". The cyclic chords progress by perfect 5ths in one direction
and the axis dyads by perfect 5ths in the opposite direction. However, I feel no
more constrained to progress along the series, which is the conventional rule of
12-tone music, than a tonal composer would feel constrained to progress in-
variably along root progressions by a fifth. On the difference array based on this
pair of sets [Example 2] the collection of chords derived from the above align-
ment is shown at the top of page 88. If you displace one set relative to the other
by 2 or 4 or any other even number of degrees, you will get another series of axis
dyads and another series of cyclic chords. All such alignments and the chord-
series they generate are shown in the complete difference array [Example 2].* I
can, for example, realign my two sets so that I will get exactly the same cyclic
chords with axis dyads of interval 6 instead of interval 0 (see the 3 6 3 segment of
the difference array):

Sums 0/7:
D~ A Bb D F G CC GF DBb A D~ E G~ B C~ F~ F~ C~ BG~ E (D~
Sums 3/10:
C Eb G Ab DC~ AF~ EB BE F~ A C~ D Ab G Eb C Bb F F Bb (C

[Example 3] shows the sum array that you can get for the same pair of sets, if you
align them so that opposite cycles meet. For example:

Sums 0/7:
G CC G F D BbA D~ E G~ B C~ F~ F~ C~ B G~ E D~ A Bb D F G
Sums 10/3:
Eb CBb F F Bb C Eb G AbD C~ A F~ E B B E F~ A C~ D Ab GEb

Where each scale of dyads gave you all twelve statements of a single interval
before, this time each scale of dyads gives you all twelve statements of a single
sum. If I add vertically aligned notes I get alternating "secondary sums" of 10
and 0. (I didn't intend to get into this, but these particular "secondary sums"
bring up something so beautiful that I can't refrain from mentioning it. 10 and 0
happen to be, in this array, "primary sums" as well - that is, they are two of the
constant sums that generate the array. Thus every secondary dyad of sum 0 also
appears as an adjacency in the first set, and every secondary dyad of sum 10 also
 86 George Perle

Perle, Modal Suite, I (1940)

Moderato (J.) F C l ~'D~ --'--I
Downloaded by [University of North Texas] at 12:31 29 November 2014

P0 P7: eab b
t3il0 ~ gaDd dbbad~e (@)dr e= ab b
__ c el~g) o" d aD ( m) a
iG. I
~.~" L ~1 ~z~ , I~ ~~

t am,,
g aba
e~b
gebd
,~b
clPg gf, d
fff"c
eabb
cebg
hesitatingly I N - ]
9[K I IL " I: ~ fast

I gc dzob b :lb. t
ab9 eD f=" o c,~ o~' a 0b dzra Db
s o m e w ~ t sZowe~" t i m . at f i T s t Cx d ~b f= a cz
I0 [ ~ ?'e~z'rding a Lf.ttl~

ceb 0 gebdcZe d x a I~ 9~"

f= 9 ( ~ ) bbd(f) e bb
c#d ab
CoD'r lght 1941 bv Boletin iahno-omericano de mOmca, Y Usedby permission

Figure 1 Perle, Modal Suite, I (1940)

 Twelve-Tone Tonality 87

appears as an adjacency in the second set. That's obvious. More astonishing is

the fact that e v e r y cyclic chord that has an axis d y a d of s u m 0 occurs as a four-
note s e g m e n t of the second set, and every cyclic chord that has an axis d y a d of
s u m 10 occurs as a four-note segment of the first set. I call such axis-dyad chords
"tonic chords" because they have a special m e a n i n g and special functions in the
system.)
The axis-dyad chords are exactly the same as before, but in a n e w arrange-
ment. (Just to m a k e it easier for y o u to see this, I've laid the s u m and difference
tables out so that they b o t h begin with the same chord.)
The axis-dyad chords are the same in the s u m and difference tables, but w h a t I
call the " s u m tetrachords" are not. Sum tetrachords are w h a t y o u get if y o u just
Downloaded by [University of North Texas] at 12:31 29 November 2014

look at two staves of the s u m or difference table at a time, instead of three . . . .

In the difference array y o u have every tetrachord that consists of a sum-0 d y a d
and a sum-3 dyad, and every tetrachord that consists of a s u m 7 d y a d and a sum-
10 dyad. In the s u m array the tetrachordal combinations are 7/3 and 0/10. I
should explain that I d o n ' t n e e d to look at the tables w h e n I compose. But it's
nice to k n o w that t h e y ' r e there. Like O z a w a always having the score resting o n
the stand in front of him, but conducting from m e m o r y .
I'll skip the m a n y and vast intervening steps that take us from these elemen-
tary beginnings to the Sinfonietta and give y o u a couple of arrays from the second
m o v e m e n t . M y first array here pairs sets of two different interval cycles, 3 (or 9)
a n d 6. I'll call one set i3i0 (which m e a n s that I'm calling it an interval-9 set) a n d
the other ilp7. Here they are [with no particular alignment represented]:
i3i0: C~ D B~ F G A~'E B (C~ /! C Eb A F~ (F~
ilp7: C C~ F~ G (C // B D F G# (B // Bb Eb E A (Bb
[Example 4 shows] the difference table a n d [Example 5] the s u m table. The
n u m b e r s u n d e r the chords [Example 6] are p r i m a r y sums and the letters are axis
notes. [I've indicated, on the array,] the chords I use in the first three bars. The
last chord of the 3rd bar,
3/0: Bb F G
1/7: C C~ F~
is reinterpreted so that I can have an 8/5 tetrachord that w o u l d otherwise be
unavailable:
11/8: Bb C~ G
5/11: C F F~
T h e n it is reinterpreted again to p u t me into [another] array:
3/4: F Bb F~
1/7: C# C G
I could go o n and on and on, with h o w different arrays intersect with one
another, h o w y o u have tonic arrays as well as tonic chords, about the different
ways of m o d u l a t i n g from one array to another, about master arrays (each of which
contains 144 arrays, instead of merely 144 axis-dyad chords as the individual array
does), about the voice-leading implications of different interval systems, etc., etc.
Instead I will conclude all this with an excerpt from the quartet I'm writing for the
 88 GeorgePerle

01 fferen~..Orr~y:
P o P7 9 - ~" -" ~- . ~_ . k~. :

9 9 ~ ~" . L. : . ,
0 - : ~- . '.. L_ ,._ ,-

,-~ ' ~-' -~" ~. =~-._ . -~-

: . I. : L_ L_
Downloaded by [University of North Texas] at 12:31 29 November 2014

m a ~a T-- m mm A rL-
--

d. a' ~dm 8" J~"

M-- IL mm
: - ~. -- :_- - ~- "- .

2 --
L_
~- . L. : L_
~- .
L-

!,- .,,. I., -- "- "

m-- II

- k~ . k. _
. L_ L- I
2 L L-
T-- m ---- T-- W I"m' "0' ~

L. L. .o- C.
4

Figure 2
 Twelve-Tone Tonality 89

6 L- ~; " w: L=
r--
=
9
r-

_ ~.
3 - = - a-
Downloaded by [University of North Texas] at 12:31 29 November 2014

6 "~; v~
: z~
~ ~ = L-

1 _ ,i= ~.. - ,.-

. ~. - - ~ p
L_ ~ ;* . k. o L.
8 -; ,_: = ,k L .
V]~
" --
9 -- m a L-

11

lO

11

10
-O- llr

Figure 2 continued
 90 GeorgePerle

S u m arr~/:
P7 DO

13 ~10
10 I 0

9 ~m 4b ~ : J: 9 9 t:

0 " g" - ~ L. _ [- , t|. 9

4,. i, k=," " - ~- " : ;= "- k-
9 - a- a ~ 4-
Downloaded by [University of North Texas] at 12:31 29 November 2014

o ~ E ~ , ,. 9 - ~- : ~ ~:

dA + ~aD -O-

,n Jl.

-4P- lP" 6

6 6.. ~ i, ~., - "- -

4
! ww
-O-

6 9

4 . P . . . . ~t. : :.= " ~'-

Figure 3
 Twelve-Tone Tonality 91

ii = L= . ~E 9 9 . . I, f f ~ -- k -- '
w 4. 1 r iTM , 9 II

+ v ]mr ~. . d. -*-

I ~l ~ T i; ,

v h, * - ~: " : ~" " Q~

1 #- ~ a . ,&--
Downloaded by [University of North Texas] at 12:31 29 November 2014

~_ am 16l .dl- ,

9 L- " . - ;;" - " "" : :--

9 a- . d.

9 ' -- = ;-= ;" =~- L - tl.,

_ a_ . Pal "*"

- k- ~ . ~. ~" ~" - ~"

3 " "- : ~= ~= ~ L. ;;" h- "-

5 . [,. tl-- ;.; - .= :']='~' ~-

iL . ~. 4-
.- I- ' . L. tt-- L! - ;=
.t.~ I " " g- r ~ ! .
r-- ~ Fw .diP- 11" ~
,-@. ~_ . , - a= 9 ~a
=j t ,= ,- - ~- :9 "it- '-

" . " -" ='-

Figure 3 continued
 92 GeorgePerle

i! Pl ! c__ = = ~,,= L 9 . L- " -

-'- -- t,,," . . . . - "- - l..

1 C.-- - - ' me
n ~- m ~m --" ~_ a - -

. , -,,- l-,- ~-1 L,. : , L_

4 ~..L- = ._ ,. :.
Downloaded by [University of North Texas] at 12:31 29 November 2014

10 " -.. ,' ; - DIP ~' -- ~~'-

- " -" ", - -
.

m 6. _ L_
7 . L. - L_
- r- . . [~.
4- v ~ ~ - ~-

. L. _ L- .e. ~. . I,. _
10 '~- - . . . . -- ~: . [.. _

" ~ ?" ', -- . L- - L- ~ IP~'I

8 ' I I . L. = c;, ~- ." .= ~= I 1-

6 -'- ~.'---; It. 9 = , -. ,--.

L_ 4., e~ . Tum 9 . L-

~, It,, - .- - I,- .,,.

Figure 4
 Twelve-Tone Tonality 93

S u m ~FFW:
2
i3 iO
PZ iL
11

11
Downloaded by [University of North Texas] at 12:31 29 November 2014

10

4 ~_ ,- ,, r

I
m ;, u ~ -- : m-
6 . " 9 . ffi ~-- "- . :L-

,, F,, 4. ~ W " a- ., ~. 4.

7 . - r- = _IL_ ;
- i ~- ~ w "0" .'IF
d m ~ 4.
w H., w m " ---- #--
3 __"~- -- 9 - ~ ,.- - ;- - L.
"4" w

8. : - ": ;; ,"- L-

Figure 5
 94 George Perle
Downloaded by [University of North Texas] at 12:31 29 November 2014

~ 3 "~o : o o ~ A~ Ab & o 3 A o 3 o F

i.p.: F

~l~ J~-: ~" Ct~

; , t~ . c /4 ~" c "/

Figure 6

Second Movement, Sinfonietta 1, by George Perle. Copyright 1990 by Galaxy Music Corporation9
Reprinted by permission. Sole seling agent: E.C. Schirmer, Boston.

Juilliard. [Example 7] shows a progression of three tetrachords of sum-couple 0,9.

[In Example 8] I have reduced the progression to its basic form by eliminating 8ve
displacements and figuration (suspensions and passing notes). In the first bar [of
Example 8] you see the primary symmetrical relations, with all the voices moving
equally in parallel or contrary motion. In the second bar I show the actual voice-
leading of the two top parts and how these become the inner voices through a
three-octave displacement of the F# in the bass, which moves to the highest part
 Twelve-Tone Tonality 95

and thus opens space between the first two chords. That space is filled in by cycles
that have nothing to do with any specific array. This kind of use of cycles is the
only thing I'm aware of ever having directly and consciously borrowed from Berg,
who first used it in this way as long ago as the Opus 3 quartet. In [Example 9] I
show how the three chords also represent a secondary sum-couple, 1 and 8. My
reason for showing this is that through this secondary sum-couple another
tetrachord of the array, of primary sum-couple 11, 10, momentarily becomes part
of the progression. In the last bar of the analysis I show the final level, with
dissonant suspensions that seem to me as clear in character and function as such
things are in traditional tonal music. This sort of thing, which I also do in the
Sinfonietta, represents a new stage in my understanding of 12-tone tonality.
Downloaded by [University of North Texas] at 12:31 29 November 2014

Perle, Windowsof Order

Figure 7 Windowsof Order (String Quartet No. 8)

Copyright 1990 by Galaxy Music Corporation. Reprinted by permission. Sole selling agent: E.C.
Schirmer, Boston.
Hope to see you at your concert and give this stuff to you then. I'm sure you
didn't suspect what you were in for when you asked if you could keep a couple
of my arrays!
Sincerely,

* [I have appended, to Example 1, an analysis which shows how each of the bracketed segments of
the piece may be interpreted as deriving from combined segments of the two sets that generate the
array. The analysis given here corresponds to that which is given in my book, Twelve-ToneTonality
(Berkeley, University of California Press, 1977). I give a somewhat different interpretation of the
relation between the composition and the array in The Listening Composer (Berkeley, University of
California Press, 1990)].
 96 GeorgePerle

a b c G b c
J
L.~. L _ :. L_ t._
I~ v v ii

Sum 9
A i

bR
" - et
Downloaded by [University of North Texas] at 12:31 29 November 2014

Sum 0
i vmo e
0 I,-

Figure 8

secondary sums
a b ~. b t c

mr

~- ! I t
t
primary sums~ 9 I0 9
0 II 0
* consonant d o u b l e suspension
t" dissonant suspensions

Figure 9