Webern's Sketches (II)

Author(s): Roger Smalley

Source: Tempo, New Series, No. 113 (Jun., 1975), pp. 29-40
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/942774
Accessed: 25/10/2008 17:18

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=cup.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Tempo.

http://www.jstor.org
WEBERN'S SKETCHES (I )

RogerSmalley

QUARTET OP. 22 FOR CLARINET, TENOR SAXOPHONE, VIOLIN AND PIANO

The sixtieth birthday of Adolph Loos. I dedicated my Quartet op. 22 to him. For it I had planned
a third movement on which I worked in August and September. Finally decided to leave this work
also 16 in two movements.17
The most extensive selection of plates in the volume ( 3 in all) is devoted
to this work. It was originally conceived as a Concerto for Violin, Clarinet,
Horn, Piano and String Orchestra and, like the Concerto op.24, was inspired
by (and presumably during) Webern's walking tours in the Carinthian Alps.
This is made clear by his verbal outline of the work (top of plate 12) which reads
as follows:
I4.ix. 928 Concerto (3 movements) for Violin, Clarinet, Horn, Piano and string orchestra.
17.i.1929 st movement: Quiet (Annabichl, mountains) perhaps variations.
2nd movement: Slow: Introduction to 3. (Schwabegg) (soloists only)
23.i.I929 3rd movement: Rondo 18
Main theme Secondary themes
I Coolness of the first spring (Anninger, first flora, primroses,
anemones [hepatica, pulsatillal)
I Cosy, warm, sphere of the highest pastures
II Dachstein, snow and ice, crystal clear air
II Soldanella, flowers of the highest region
III The children on ice and snow
I Repetition of the ist secondary theme (sphere of the
Alpine roses)
II 2nd secondary theme, light, sky
IV
Coda: Outlook into the highest region

The first group of sketches (Plates 11 - 14) are a continuous series of pages
(53 - 56) from Sketchbook I. They show the evolution of the basic set, combined
with attempts to form the evolving set into a violin melody. It is some time before
the set attains its final form (Ex.8, P0) which is labelled 'gilt (alle Intervalle)'
['is valid (all intervals)'].

? 9 7 5 by Roger Smalley
30 TEMPO

Ex.8
Po. _R . o(ll)

..+Rlo

The chronology of this work is difficult to determine, because not all the
relevant pages are reproduced, and because of the rather chaotic nature of the
sketches themselves. My conclusion, after studying the available evidence, is
that the three movements were actually composed in reverse order-III, II and
finally I. But the third movement to be composed (intended to be the first) was
never completed, and the two movements of the published score therefore cor-
respond to the second (II) and third (III) of Webern's original scheme. The evi-
dence for this is as follows:
(i) According to Webern's list of contents on the title page of Sketchbook I
(reproduced as Plate i) the sketches on Plates I I - 14 were for the third move-
ment of the original scheme. Although they bear no resemblance to the eventual
third movement (second in the published score) they must gradually have evolved
into it because at the top of Plate 19 (page 4 of the next sketchbook, with which
the reproductions of the Quartet sketches begin again) the last five bars of this
movement appear, in a form which corresponds almost exactly (down to the
correct bar numbers, I88-I92) with the published version. The only major
difference suggested by this fragment is that the entire movement was originally
notated in 2/8 (in the published version it is in 2/4, with all the durations doubled
in length). Beneath the double bar-line at the end of this sketch is written
'I2.iv.3o Mbdling'.
(2) Now follows another gap in the reproduced sketches, but in his commentary
Krenek tells us that 'On August 2oth [ 930o, only six days after having completed
-on August 14th-what he called the "second movement" of his Quartet op. 2 2
(in the published version it became the first), Webern began to contemplate a
third movement for this work: "Sehr langsam" ' (plates 2 g to 33, which are a
continuous sequence of pages-29 to 38-from Sketchbook II). Krenek con-
tinues 'It is not possible to conjecture whether it should have followed the
published movements or taken a different place in the context of the whole work';
on the contrary, I believe that the textual evidence to which I now turn proves
conclusively that this movement was intended to be the first.
(3) In Webern's 'programme' the first movement is designated as 'ruhig'
(quiet). The most highly evolved sketch on Plates 2 - 33 (Ex. 16) is marked
'sehr ruhig' (very quiet).
(4) The plan also suggests 'perhaps variations' as the form of the first movement,
and it is highly probable that Ex. 9 - i6 do, in fact, trace the evolution of a
theme for variations. It was first conceived as a melody for saxophone, accom-
panied by the piano. In the later sketches the melody is shared between violin,
clarinet and saxophone, but still remains a single melodic line (only in the
sixth bar of the final sketch-Ex. 16-do melodic overlaps occur). The piano
retains its accompanying role and emphasizes the two cadential points at bars
4 - 5 and 7. The second of these is a perfect fifth lower than the first and they thus
provide a distant reminiscence of the half- and full-close of a tonal variation theme.
That Ex. 16 is definitely a theme (as in the Symphony op. 2 , II) and is not itself
the first variation (as in the Piano Variations op. 27, III) is proved by the indica-
 WEBERN'S SKETCHES (II) 31

tion 'i Var.' immediately following Ex. 2 (in the book this will be found above
bar 8 on Plate 27).
(5) The published first movement is quite different in both texture and form.
But its tempo marking-'Sehr massig' (very moderate), and preludial character,
do correspond closely to Webern's description of the second movement ('Slow:
Introduction to 3'). Its introductory relationship to the second movement is
further emphasized when we compare their respective durations-the second
movement takes almost twice as long to perform as the first.
(6) The most significant evidence of all is provided by Webern's choice of set-
forms. The set labelled P0 in Ex. 8 is taken from the sketches-it also corres-
ponds to that given by Friedrich Wildgans in his book on Webern, in which he
states that 'the notes-row tables appended .... were those written in Webern's
own hand and found among his effects'. 9 On this basis the first two set-state-
ments of the published first movement are of P7 and 13 (there are no statements
of P0 in this movement), whereas the second of the two published movements
begins with P0 and ends with R0, and the theme of the projected third movement
consists of I0 and R0. This strongly suggests that the published first movement
was originally intended to occupy a central position, with the other two move-
ments on either side of it. Thus George Perle,20 who identifies the set-form
beginning on the saxophone in bar i as P0, and Walter Kolneder 21 who, less
understandably, selects the inversion of this set which starts on the saxophone in
bar 6 as P0, are both, strictly speaking, wrong, although it is quite reasonable
to assume, in the absence of any evidence to the contrary, that the first thematic
statement of a twelve-tone work will be of P0.
It is true that even the movement at present under discussion, which was to
have been the first, does not begin with a statement of P0. In fact the only move-
ment which does open with P0 is the published second movement. This was
originally intended to be the third, but was actually the first movement to be
composed. This suggests that for Webern-unlike Schoenberg-the thematic
significance of particular forms and transpositions of the set was not crucial. Far
more important were the relationships created between the sets used within each
separate movement. I think it is for this reason that Webern was able to compose
the movements of the work in the 'wrong' order and subsequently to change his
initial plan so that it opens with a movement containing no statements of P0 at all.
When I first drafted this article (as long ago as I97 )22 I was unaware of
the existence of an extensive article by George Perle on the same subject.23 In
the event it turns out that our two articles complement one another very well.
Apart from the general description of the book and its contents there is very
little duplication of material. As his title suggests, Perle looks at the sketches
primarily from the point of view of Webern's use of the I 2-tone technique. He
examines its origins in the pre-I2-tone works of both Berg and Webern, and
its use in the Piano Suite op. 2 of Schoenberg, before looking at the different
ways in which Webern employed the technique from op. 17 onwards. He devotes
as much space to the published movements of opp. 19-22 as to the unfinished
ones. My principal aim is to trace the evolution of musical material within the
sketches themselves. Perle, on the other hand, takes the most complete form of
each sketch as the starting point for a more general discussion. His article, there-
fore, sets the preoccupations of my own in the wider context of Webern's
development.
In one respect in particular I am glad that I had not read Perle's article when
32 TEMPO

I wrote my own. This is because in relation to opp. I9-22 we have both specu-
lated, on the basis of evidence drawn from the sketches, about the order of
composition and intended sequence of movements, and the possible reasons for
their abandonment. And in almost every instance our conclusions are identical.
This does not invariably prove them to be true, but the fact that they were arrived
at independently by two people who have studied the sketches very closely must
surely be counted significant.
With regard to op.22 Perle agrees that 'presumably . . . the projected
additional movement was intended to open the work'; citing, in one form or
another, most of the six points I have made above. In addition he has noticed
that the opening bars of the unfinished movement are based on the same set forms
as the concluding bars of the rondo (the published second movement) and suggests
that 'the F-sharp centricity of the opening bars would in all probability have been
confirmed in the projected movement as a whole, as it is in the last movement,
and the first of the two published movements, beginning and ending on C-sharp
and avoiding the principal group of row forms entirely, would thus have served,
in terms of its "tonality", as a contrasting middle movement'.24 Further
comments from Perle's article will be quoted at the appropriate points.
Examples 9 to 16 show all the most significant stages in the evolution of the
variation theme, laid out (as far as is practicable) beneath one another so that the
changes can be observed as clearly as possible. The first two sketches for the
theme are Ex.9a. The version on the third system down was probably sketched
first because (i) the theme was at this time intended for saxophone (it is the only
one of the three melody instruments which can play the low Bb])and the composer
would naturally begin to write on the saxophone system; and (2) it is structurally
in a more primitive state than the version which was subsequently written on the
upper two systems.
Webern's aim in constructing this theme is to combine set-forms Il (which
he labels III) and R0 (which he labels II) into a continuous melodic line. The
crucial point will obviously be at the junction of the two sets, which occurs in
bars 6-7 of Ex. ga. This is the place where any structural weakness will show up
most obviously, and Webern therefore concentrates on this moment during the
first few sketches (up to Ex.9c). In his first sketch (Ex.9a, third system down)
he uses the pitch Cf (12 of Il and i of R0) as a pivot, but omits the second pitch
of R0 (the note under bar 6 reads 'g is missing'). The second version (upper two
systems of Ex.ga) exploits the fact that pitches 9 and Io of I0 and 3 and 4 of Ro
are identical (C# - BL) and uses this invariant dyad to link the two sets. The
notes above the top stave read 'leave out c' and 'left out f c and c g' (the first
dyad of I0 and the first dyad of R0 respectively). The reasons for these omissions
will become clear when we go on to examine the structure of the piano part.
We can now follow the painstaking series of stages through which Webern
refined this melody. Ex. 9b is a renotation of Ex. ga (second system to bar 5,
first system from bar 6) in 4/4, with a few small changes of duration. In Ex. 9c
the melody attains its first definitive shape, which remains unchanged until Ex. 13.
The most significant alteration here is that the repetition of the connecting dyad
C$ - Bb (see Ex. 9b, bars 4 - 5) is omitted so that the two sets are drawn even
closer together.
The contour of this version of the I 8-note melody is symmetrical around the
opening F# and closing Gb. Eight of the remaining 16 pitches are above this
central pitch and eight bl2ow it; furthermore the seven pitch-classes which are
 WEBERN'S SKETCHES (II) 33

20 viii-l

fchlt g

Ex.96
(ts.]

Ex.9d
. ir Ir:'
t~:~
!J l,,a t i I Jt
Yjj , , " -

~^ J iti?ii $ ,lI Itu1 !~t'

itr ~t" fIl, .r^, 2l i ~ ~t1"I^,t
~it 1'I
Ex.9e
ID m EDD ri0 ED M1111

_,
Ex.9
ED21
3M
ID ID LI

5a*.

~~~~~
p~~~~cr ~~~~i PP~~~P
fehlen fc u cg

repeated all recur in the same octave position (see Fig. 2). This limitation of
pitch mobility, highly characteristic of Webern's later style, has important con-
sequences for our perception of his music. Firstly it provides a kind of harmonic
unity which binds together long stretches of music (for example the first 32 bars
of the String Quartet op.28). Secondly it directs the listener's attention away
from the pitch micro-structure and towards the perception of larger-scale re-
lationships. This makes the music easier to grasp because it involves the recogni-
tion of identities and similarities which are not abstract but are set up within the
context of the work itself. For example, in Ex. 9c it is possible to perceive that
the climactic pitches of each half of the melody are identical (the En's in bars 3
and 6) without having to know what pitch it is or why, according to the choice of
set-forms, it should occur at these two particular places. It is sufficient to recog-
ise that they are the same pitch, and one of the purposes of the principle of

Fig. 2.
r E E
1JS -./
D A ,A _ _ _ A
/ \
st_ -- X

FV
Fr
cerl \ i V"
-- K/ \
- -i---...
--^- ...... /
A i, .

C> *. i<
/0
/^/
\
I

B
34 TEMPO

fixed octave positions is both to create and render audible such relationships.
Webern's sets often possess a very complex micro-structure (like that of op. 24)
but in listening to a work we are not meant-nor, indeed, is it possible-only to
follow this repeating set-structure. Ex. 9c is a good example of how Webern
superimposes higher levels of order onto the continual rotation of individual sets.
During our analysis of the complete theme we shall see how the theme's overall
form is articulated by a very subtle use of such relationships. Ex. gc has a time-
signature of 5/16, and now that the registral distribution has been fixed Webern
continues to experiment (as we have already seen in the opening of Op. 24) with
different meters, altogether trying out the same melody in 3/4, 4/4, g/i6, 1/2,
2/4 and (in Ex.Ioa) 4/8. These sketches are mainly concerned with minute
changes of grouping and accentuation. The most significant decisions appear to be:
(i) whether the opening F$ is to be separated from the following Ab - Bb
-
dyad by a rest (as in Exx. ga d), or not (as in Ex. ge). Webern finally settles for
no rest, but differentiates the two events by means of articulation (Ex. 9f).
(2) Whether the next five pitches are to be grouped 2 - 3 or 3 - 2-in other
words should the EL (6 of Io) be attached to the preceding Gh-G# to form the
climax of a rising three-note phrase, or is it to be preceded by a rest so that it
forms the high point of a descending three-note phrase (Eb - Eb - D)) ? Again
Webern chooses a solution involving the greatest degree of ambiguity by eliminat-
ing the rest (in Ex. 9e) and placing the E] on the downbeat of bar 3 (Ex.9f); in
this delicately balanced position it is heard as belonging to both the antecedent
and consequent limbs of the phrase.
(3) The accentuation of the central 4-note phrase beginning on C#. In
Exx.9c and d this C# is double the length of the three succeeding pitches; the
phrase thus tends to 'sit down' on its first note. In Ex. 9e and f this effect is
avoided by making all the durations equal in length. The new distribution of
accents (the C$ is now off the beat and the final Ab on a downbeat), coupled with
the ascent of the phrase from the lowest pitch of the entire melody (B1), generates
a much greater sense of forward movement. The three alternative phrasings in
bars 4 - Y of Ex. gf show that Webern was preoccupied with this problem of
metrical stress. The rhythmic structure, rising contour and large intervals of
this phrase together create an impetus which projects the melody towards its
main climax (the EL on the downbeat of bar 6 in Ex. 9f) from which it falls
away, in two elegantly descending phrases, to the concluding Gb.
I will call Ex.9f the first definitive form of the melody, because it is retained
unaltered through the next three sketches (Exx.lo - I2), in which the piano
accompaniment begins to evolve. This uses the same two set-forms as the theme,
but in reverse order-i.e. Ro followed by Io. Now we can see the purpose of the
four omitted pitches in the middle of the theme, and how they are used to create
the closest possible interrelationship between melody and accompaniment. In
bar 4 of Ex. ioa the Cb and Gh of the first piano chord are pitchesI and 2 of
R0 and initiate the second half of the melody (pitches 3 to 12 of Ro) which is
continued and completed by the saxophone. The Ff and Ch of the second piano
chord, on the other hand, are pitches Ii and i 2 of I0 and thus complete the first
half of the melody (pitches i to l o of I0) begun by the saxophone. Still with bar 4
the F# in the first piano chord is pitch i of I,, which is then completed by the
remainder of the piano part (bars 5 - 7); and the G, in the second chord is pitch 2
of R0, retained from chord i. This second chord, which occurs at the precise
mloment the four sets cross over, is thus the perfect pivot chord, containing one
 WEBERN'S SKETCHES (II)

[KlIv.

" -
, I

' a
[ki,.] 1 =I
.
... ... !;

Ex.11
rn Sehr langsam *= ca. m

pitch (CL) common to both I and R0, one pitch (FL) From Il and one (Gb)
from R0.
In the next four sketches Webern explores different ways of grouping the
pitches in the piano part. Fig. 3 facilitates a detailed comparison of the various
changes and renders further description unnecessary. The first definitive version
of the entire theme (melody and accompaniment) is reached in Ex. 2. I-Iere all
the grace-notes have been eliminated from the accompaniment and the three-
part chords in bar 4 are retained. The Bb in the final bar of the piano part is
completely new (it comes from RI,, set number 28 in Webern's nomenclature)
and means that the cadential motive of bar 4 is reproduced in a transposed form.
36 TEMPO

Fig. 3.

BaS --> 2 2 3 4 - --, -s- 7

En^ @ .A.2 2Z / ? @ I@? ^3 @

E 1oM) -1/1-/ 1+@9/@+1 L3 / 33? Z+1 1+2 o

?i /1+
-2
ExFlo() +Z
I+z +1 /1 t 1 t0 +(2+I 1 /+2 / 1+2
f( 1C or
f( + Zt 2
kx1~Ci
11?1
t~, /k
En1
2y 14-l+/+1 1t/lZ+1 I' 2
z2zli / 1 +

Ext 1 +@/?+1 +QI+1 @t @Q@+ 1 1 t<

+ Fp mit;ih

I. Numbersin circles ( 3 etc.) indicate that the pitches are soundedsimultaneouslyas chords.
2. A grace-noteflag (A) above a numbershows how manyof the pitches are articulatedas grace-
notes.
Even the apparent perfection of Ex. 12 did not satisfy Webern's ideals. The
sketches continue, showing that he still wanted (i) to employ all the melody
instruments-not only the saxophone-in presenting the theme and (2) to inten-
sify the thematic role of the piano part. The next two sketches are markedly
different from those preceding them. In Ex. I3 a first attempt is made to divide
the melody between all three melody instruments; that this was Webern's main
preoccupation here is indicated by the fact that only the first three bars of the
piano are written in (although even these are different from any preceding
version). The violin part in bars 7 and 1 3 is particularly interesting in that it
shows Webern exploring a further possibility (which was not followed up) of
establishing a thematic relationship between the half- and full-close. Ex. 14
tackles the other unresolved problem, that of creating a closer motivic relation-
ship between melody and accompaniment. Webern returns (two days after
Ex. I3) to a version of the theme on saxophone. The piano part remains sub-
stantially the same as in Exx. i 2 and I 3, but Weber tries to integrate the melody
by making it take on many of the characteristics of the accompaniment. It is
noticeably less legato than before and is subdivided exclusively into I- and 2-note
segments; the 3- and 4-note segments of Ex. 12 have all disappeared. Many of
the 2-note groups are separated by rests (e.g. bars 2-4, 8-io) into which the piano
part slots. This, coupled with the rhythmic similarities between the motives
and the many pitch identities has the effect of making the music sound less
like a melody with accompaniment and much more like a (quasi-canonic) dia-
logue between two equal partners.
The final version ('gilt') of the theme is Ex. 16. A remarkable point here is
that whilst in the earlier sketches there are often only minute alterations between
one sketch and the next, this radically new version sprang into being almost fully
formed. It is preceded by only three other sketches, the first of which is Ex. I 5.
I have not included the other two because they differ from Ex. i6 only in a few
details of octave placement. The main differences between Exx. I and 16 are
already resolved by the next (omitted) sketch. These are (i) the conversion of
the time signature from 4/8 to 5/8 (the figures above the top stave of Ex. i 5 show
 WEBERN'S SKETCHES (1I) 37

Ex.13
c-] ?J^"
I:~~~
I;~ ~ ,
^ij^^ I ..-
:tj
27-viii1930 'I^ 'jC"ii"
[]i

Webern already working this new distribution out) and (2) the further compres-
sion of bars 7-8,accomplished by moving the FL - D] dyad one quaver backwards.
By Ex. I6 all Webern's aims have been achieved. The melody is shared not
only between all three melody instruments but also the piano, which takes over
the third pitch of I (Bb) in bar 2; furthermore the piano part is fully integrated
into the overall motivic structure.

Fig. 4. --
I A1 I B, 1 71 Cq 7r A, 11 3 17

'
d
jg ,J 1.
5 f 7' L
- 7 r, 7
t i' I
'1 7
C-
7*I
.nk ,!!. 7~~~ ~~ J
ei
'J

I C, 1 yI A3
Me!.1 1 7 7 I
7 Ut
'1
LJ 7 4
 38 TEMPO

Precisely how closely integrated is shown by an analysis of the rhythmic and

pitch (as opposed to serial ) structure of the entire theme. Fig.4 is the rhythmic
structure of Ex. 16. It is made up of three cells-A, B, and C-each of which
occurs two or three times in more or less varied forms. These cells are all
formed from a combination of melody and accompaniment, clearly demonstrating
their inseparable unity. Cell A appears three times: in bar i, bar 4 (the 'half-
close') and bar 7 (the 'full close'). It consists of two motives, a minim followed by
a staccato quaver (d) and a group of two quavers (e). In A1, d is played on the
saxophone and e on the piano; in A2 and A, these roles are exchanged. In A1, d
is in single pitches, in A2 and 3 three-part piano chords. Rhythmically d remains
unaltered, whilst motive e undergoes a process of progressive rhythmic simpli-
fication; in A2 the grace-note is omitted and in A3 the two quavers become a
single crotchet.
Cell B occurs twice, in bars 2 and 5. It consists of two motives, a dotted
crotchet (f) and a group of two staccato quavers (g). In B1 these are both played
by the piano; in B2 they are shared between clarinet and piano. The dotted
crotchet (f) has become a crotchet followed by a quaver.A very subtle point here
is that since each part contains the same two pitches (Bb - Ah and AL - Bb) we
do actually hear a dotted crotchet's worth of the B[b.
The most complex cell is C, which consists of four groups of two quavers
(h, i, j, k) plus a crotchet (1). In C1 the two groups of two quavers (h, i; j, k)
are presented in canon, at a distance of one quaver. In B2 (bar 6) this process is
intensified; in addition to the canon between melody and accompaniment the
quaver groups within each part are also in canon at a distance of one quaver.
This naturally-reduces the total length of the group by two quavers. Thus although

Ex,15 calando tempo

?[19]30 m
(4x
[ LJ30
m
U] O-[41]
E
5 1
E 5
-3 2 3 4 5
E1
2 3 4
m
51 2 3
3
4 5
[.
1 4 51
El
2 4

fpf

b~
r?..3?
1- S
"f
- -jt- - _- -j_p C

12 3 1 Xl =.-
'
' . a eP
i
PP 4
r P

J,
<hK1??3j mI J. ?^ . 3- J.- 4=-1_-
,t -J PP fp PP
 WEBERN'S SKETCHES (II) 39

the other cells both have the same durationand appearance(A1/2/afive quavers,
B1/, four quavers), C1 is seven quaverslong whilst C2 is contracted to only five
quaversand forms the climax of pitch eventuation(i o attacksin the space of four
quavers),density (the only 3- and 4-note simultaneitiesoccur in this bar), pitch
(the EL in the violin and piano parts is the highest note in the entire theme) and
dynamic(forte in all four instruments).
The overallform hingesaroundthe three appearancesof A, separatedby two
occurrences of cells B and C. As previously suggested this formal structure is
clarified by the fact that each of the three cells uses pitches selected from a
different area of the total chromatic. Fig. 5 shows the distributionof pitches
among all 7 cells. Visuallyit suggestssome strikingparallels. The two B-cells
use exclusively the pitches Af -Bb -Bf, the A-cells are concentratedaroundB]
-Cf -C# and the C-cells aroundDf -Eb -E -Ab. In Fig. S the principalpitch areas
of the three groups of cells are indicatedby continuousbrackets; less frequently
used pitches by extensions of these as dotted brackets. It can be seen that the
total chromatic is sub-dividedinto five well-definedareas, the B-cells using one
areaand the B- and C-cells two interlockingareaseach.
Fig. 5.

r i A 1 r A
I B I___1 r---I C 1--1 r Ic I

A CB Cc D Eb E' -- F F Ah
A1 x X , X x

A xx x x x xx

XX sx 0

Ix ____I
'
CZXX X<X % X \X X

This seven-bar theme seems to me to be as masterly a piece of music as

Webern ever wrote. It can surely only have been for reasons connected with the
overall balance of the work that he abandoned an idea so rich in possibilities and
so perfect in itself. 25 The same reason must also, I think, be invoked to account
for his giving up a third movement for the String Trio op. 20, of which a very
substantial torso of 31 bars remains.
To be continued in TEMPO114.
Materialrelating to Weber's op. 22 quoted by kind permissionof
UniversalEdition (Alfred A. Kalmus)Ltd.

NOTES
6. i.e. as well as the String Trio op.20 and the Symphonyop.2 i, for both of which Webern had begun a
third movement.
I7. Webern's diary, Io December 1930. Quoted in Wildgans, Friedrich: Anton Webern.Calder and
Boyars, London I966, p. I39.
i8. The following materialhas been slightly rearrangedin order to give a clearer impressionof the probable
sequence of events. Anningerand Dachsteinare mountainsin the AustrianAlps. See also PartI of this
article, (TEMPOI 2) page I i, paragraph3.
19. Wildgans, op. cit., p. 18.
40 TEMPO

20. Perle, George: Serial Compositionand Atonality. 2nd (revised) edition, Faber and Faber, London 1968,
p.8o (Example I I 3).
2I. Kolneder, Walter: Anton Webern. Faber and Faber, London I968, p.I20.
22. The following 3 paragraphs were added in I975.
23. Perle, George: 'Webern's Twelve-Tone Sketches' The Musical Quarterly, Vol. LVII, No. i, January 1971,
pp.I - 25.
24. Perle, op. cit., p.20.
25. cf. Perle p.2o: 'Why Webern should have discontinued the projected additional movement after the
promising beginnings that are found in the sketchbook remains a mystery. There is no question of a tech-
nical cul-de-sac, as in the projected third movement of the Symphony, and he does not appear to have
been distracted by the conception of another work, for he makes no use of the sketchbook again until
four months later'.

ROGER SMALLEY

MONODY
for piano with live electronic modulation
?3.00

MISSA BREVIS
for 16 solo voices ? io5

PIANO PIECES I-V

60p

VARIATIONS FOR STRINGS

for string orchestra?I.o5
Other works by Roger Smalley are available on hire

FABER MUSIC LTD

38 Russell Square London WCiB gDA.