You are on page 1of 12

B Major Hamony

as

Dominant or

Mediant, in Ch,opin's,Opus ilDlf.z


Schenker's Graphs from Free
Comp o sition Reconsidered

Howard Cinn'amon
At some point in

every undergradua music theory'curriculum it

becomes necessary to select pieces from the repertoire to illustrate the


of chorale-style voice leading, upon

commonly based, and those of free

composition. A freent first sp in this direction is the selection of a


piece based upot strigborward arpeggiations in which the figuration
pattern can be easily represented as a series of four- or five-voice
chords that manifest basic chorale-style voie leading and harmonic
progression. This practice is particularly true in courses focusing on,
or incorporating some, elemenB of Schenkerian analytical methods, for

which the unfolding of harmoniesl is a basic concept. But it can also be


a useful first step for approaches using other methodologies as well.
One piece commonly chosen for this purpose is Chopin's tude
in C Major, Opus 10/1. Two factors often contribute to this selection:
the structural parallelisms between this piece and J: S; Bach's Prelude
in C Major from Book I of The Well-Tempered Clavier (an equally

rUnfolding of harmonies is defined as the transformation


of harmonies tbat
finction as simultaneities on one lerrel futo ielodic paerns (e.g., arpeggiations) on a
more foreground level.

Indiana TIny Revew Vol. l5l1

Cinnamon, E Major Hannony as Dominant or Mediant

popular choice to lll this rnction)2 and, for those wishing to


incorporate schenkerian methodologies, the fact that two of schenker's
os,n voice-lelding graphs of this piece :re readily available in Free
Cnmpositon}- Th" parallelisms with the Bach pielude allow for a
disc'ssion of the nature of style in general anr $tistic differences
between these pieces iir particular, while the availability of schenkerian
graphs provi{es models th. can be used as a gqide (both for the
student and the teacher) an allow for relevant references to Schenker's
own writings.
If one is not careful, however, these factors can become negative
elements by unduly influencing both instructor and students in their
interpretation of these pieces, sometimes leading to an overenphasis on
the similarities between them and./or promoting the acceptance of
Schenker's analyses without the healthy degree of skepticism so
necessa to the development of the critical analfical skills that are the
ultimate goal. One factor that mitigas against the selection of the
Chopin hde is the questions prompted by seeming inconsistencies that
appear when Schenker's voice-leading graphs are closely compared
with the piece and the in-rference with the learning process that may
result. This surdy re+*arnnes the piece from a p"i"cti"" pro-p
by some of these questions and offers an alternative view of its tonal
strucre. TVhile this discussion employs Schenkerian methodology and
thus will be of direct relevance to those who employ similar .ihos
in their approach, the reconsideration of tonal orientation within the
middle section of the tude will be of interest to all analysts and may
cause even non-Schenkerians to re-examine their perspective on the

2Serreral
Scherkerian analyses ofthis Bach prelude are availablead rcle\ant her:
Heinrich Schenker, Five Grqhic Mnsicol Anolyss, ed. Felix Srl'er (Norr York Dover,
l9@), 3G37; Aller Forte and Swer E. Gilbert, Intrtdon n Schen,rion Anasis
(New York \\. Vf. Noror, 1982), 18&190 and,2V2; David Nanmeyer and Susan
Tepping, A Guid m Sankrian Anafrsis (Englewood Ctffs: Preotice Hall, 192), 6870; and rffilliam Drabkin, "A Lsson in Analysis ftom Heimich Schrker," Mnsr:c
,lnosis 4 (19&!: 241-258.

3Heinich Schenker,
Free @mposition, trans. ad ed. by Emst Ost (New York:
I-ongman, 1979), Supplement, Figures 130.,4 and 153,2.

controlling harmony
piece.

[.e.,

key oriention) of the middle section of this

In considering the voice.leading graphs inFree Composton, one

essential point must be kept in mind. Unlike the graphs in Der


Tonwlle, Das''Metenryelk in der Musk and some of Schenker's
specifically analical writings, the voice-leading graphs in Free
Composition are intended as illustrations of specific theoretical points
made in the t9&, not complee enalyses of the pieces discussed. As a
result they focus on the point being made and include only as much
information about the entire piece as is necessary to provide context.
In cases where a more detailed analysis,exists elsewhere, it often
provides essential information neither includ nor suggested in the
illustration in Free Compositon In cases where no other source exists
(like that of this Chopin nrde), w have nothing to guide our
inrpretion of the analysis or our understanding of Schenke's view
of the piece other than the information specifically repiesented in the

illustration. While these graphs ae almost certainly based on


comprehensive analyses, the only way to arrive at those analyses is
through extrapolation from the information provided, based upon an
understanding of the principles employed. This necessarily involves a
considerable amount of speculation and analyticl input by the
interpreter and can often be a matr of considerable deba among
advocates of Schenker's theories. The discussion can be simplified
considerably when there is more than one voice-leading graph to
consider and compare, as is the case here, but its difculties can be
compounded by apparent inconsistencies between these graphs as well.
The two graphs of Opus 10/1 included in Free Composition,
found in Example l, are innded to illustrate distinct theoretical
concepts, but they appear, for the most part" to reflect a consistent view
of overall form and harmonic organization in the piece. Schenker
considers this piece to be an example of three-part form in which the
middle section is articulad by a prolongation of VI (A minor); in fact,
the second graph of Example I (his Figure t53,2) is used specifically
as an illustration of a category of three-part form in which a neighbor

Indinno Thcory Revew

!r

ia
lr

'!

l
ie

9
.7
g

ir

t^
u)

go

-t

l
o-

to

<Cl

t,)

c
t

9.

ND

,t
oa

Vol' 15/l

Example l. Schenker's voice-leading graphs of Opus l0/l;


Figures 130,4 and 153,2 from Free Compostion

4,

M$or Harnony Dominant or Medottt

2.

Forleand Gilbert's representation of tonal stfucEre in

133.

5lbid., l13.

'1bid.,

The opening sixteen measures conclude o I rather than V,


and with Jreassertion of scale degree 3 in ttre melody'

from an intemrPtion theY state:

Allen For and Steven E. Gilbert, in the chapter on form in their text

Opus 10/l

Example

from

a d 47 uras it

A (W)

Here is another remarkable use of the divider. Only

motion occurs in an inner voice rather than tte soprano.4 In his


discussion of the first graph of Example I (his Figure 130,4), schenker
elaborates somewhat on the nature of form, and the harmoniccontrapuntal press that generates it in this piece:

Cinnamon, E

Indiana Theory Review

Vol.

l5ll

Measures 1748 continue to prolong (tto* over VI), and


the inverted dominan^ t at thp end of m. 48 (harmonizing the
upper neighbor o 3, not 2) is clearly felt as a transition
rather than a point of rest.6

Example 3.

While Schenker considers. the harmonic organization among the


of this piece to be based o. n a familiar I-VI-I harmonic
organition, he apparently views the prolongation of VI as somewhat
unusual in that some of its main elements consist of dominant
harmonies that prolong their tonic, but do not resolve to iu what he
calls "fifth dividers" (sometimes also called "back-relating
dominants"). This is, in fact, the specic point that the first graph in
Example 1 is intended o illustrate. Such "fifth dividers" function as
offshoots of a tonic harmony that precedes them, but serve to close off
a musical idea rather than resolve to I.7 In Schenker's view, the E
major harmonies that occur within the prolongation of YI, atm-24 and
. fi, ae extensions of tonic (i.e., A minor [or major]) harmonies that
precede them in m. l7 and m. 25 respectively. For this to be the case,
the tonic harmonies must be clearly established as such by other
contextual elements. While these two E major harmonies seem clearly
to be the goal of the harmonic+ontrapuntal motion, the tonic nature of
sections

the harmonies they are supposed to be prolonging is not nearly as cleaThe most serious question involves the harrrony in m- 17, which

Schenker identifies as I of A minor (VI of C major)- While this


harmony does employ A as it bass and two of the upper voices do
contain C and E, its highest voice adds an F to this harmony, resulting
in a 6/5 chord that begins a sequence which ultimaly arrives on the

E
re
F

23-U

enker's

shing to
in A minor, such a harmony would require the establishment of an
Forte and Gilbert,

Innofuaion to Schenkeri Attatysis,2V)-2V3.

?For a clear, concise discussion of ''back-relating dominants" see Edwad

and

Aldwell

cal schachtf, Harmory atdvoiceLcading,?rlded. (New York Iarcoutt, Brace,

Jovanovich, 1989), 14547.

Cinnamon, E Major Hannony as Domnan or Mediant

Mm.

15-24; seguence leading from

I to V^/I

IndaruTheoryRevew Vol- l5l1

minor context prior to its appearance. Furthermore, while the sequence


17-24 an be inrpred in terms
in terms of C major.
well
understood
equally
can be

that.follows in mm.

of A minor, it

Example 4.

a. Sequence that would lead to Y in m.24


b. Modified version as it appears in mm- 1724
mr.

17

Cinnarron, E MQor Hanrry as Domhant or

Mediant

redirected to an E major harmony (made


by the augmented-sixth chord of m. ?2,
inner voice from F to D# (a chromaticized
appears in mmr 3-7; see the,first graph in Example 1)- Th first real
indication of A minor as a tonic, therefore, occurs only with the arrival

on its domiqant in mm. 23-24, six measures after schenker assefts a

circle of dominant-seventh chords, arriving ultimaly back on an A

Insntnerus) for further clarification. He states:


When

This view is reinforced when one examines the sequence more


closely. Example 4a shows how its voice leading seems to be-direcd,
not to V of A minor, but to V of C major through a consisnt G5
soprano-bass voice-leading pattern that would normally result in a G
mjor harmony in mm.23-24. Example 4b shows how the sequence is

A is sounded for

Free Contpsition,

ll3.

the second time

in

mea'sure

l0

Indiaru Tluory Raievv Vol. 15/l

While it is certainly possible for a harmony to function as a


dominant in the foreground while serving as a structrral I in the
middleground (or even background), this har,mony, like ttrat of m. 17,
seeN to initiate a transitional process rather tban serve as a point of
structural repose. This process, a circle of dominant-seventh chords,
leads first to a brief resting point on the dominant of C major (mm. 2il28) and then on to an A mjor harmony in mm. 35-36 (the only
harmony in the passage that does not contain a seventh, making it the
most stable harmony within it). The transitional process then continues
further, again through a circle of frffs that is (this. time) mostly
diatonic in A minor, to the V/VI of mm. 4:748. The asertion that the
dominant-seventh harmony of m. 25 be considered strucrral, while the
only consonant potential tonic in the passage, that in mm. 35-36, is
considered inerely transitional, seems troubling.
An equally troublesome segment of Schenker's graph is the
representation of the brief retransition to I in ntm. 4749. The
hamonies represented in Figure 130,4 of Free Composition (see
Example l), along with the analysis of scale-sps (Srufe presented
below the graph, indicates that Schenker considers this transition to be
based upon a succession of fifth-related harmonies in which the
prolonged VI of mm. 1747tds first to tr and then to V, which in
hrrn continues to I and the beginning of the reprise at m. 49. The
meaning of the parentheses arorlnd the tr chord is unclear. It probably
sugge.sts that tr is implied though not explicitly stated (or perhaps
omitd entirely), but no indication is given of any modification,
zubstitution, or alteration to the V chord. On the basis of this graph one
would expect to find a root-position V chord in m. 48 that leads
directly to I, but no such chord exists in the music. The actual
transition is achieved rather subtly and suddenly through a secondinversion \Z chord that occurs at the last possible moment, on the last
quarter of m. 48.
One might say that it is precisely this appearance of Va/3, with D
in its bass, which prompted Schenker's suggesion of a tr chord: tlat,
in some way, the Val, represents a merging of II and V in such a way
as to produce smoother voice leading. While such an elision might be
conceivable in some circumstances, this explanation represents a view

Cinnamon, E Major Hatmory as Domunt or

Medittt

11

of the-piece that seems needlessly conVoluted, especially considering


the familiarity of the acral progression and voice leading that occurs
in mm. 4749. This view does offer some rationale to justiff
Schenker's analysis, but his representation simply cannot be aken
lirally in light of a more viable alternative.
Example 5.

Mm. 47-50, the retransition to I

Example 5 presents mm. 47-50 of tle piece whi Example 6


illustrates the foreground Voice leading operating within them. The Va/t
seems to be a straightfonvard passing chord. In fact, this entire
progression seems quite familiar. In this context the E major harmony
(viewed up to now as V/VI, a dividing dominant) is considered Itr# of
C major (exhibiting secondary mixhrre as III often does)- It functions
as a substitrte for f and returns to f/r through a passing chord with no
root position dominant present or needed-e (No that Forte ard Gilbert
(and ltrilt) as a substitu for f and its opention in
Aldwell and Schachr, zLzll , 218-219, and 361364.

Tor a discussion of Itr


passages

like this,

see

12

IndanaTheory Review

Example 6. Voice leadingof return to

asaPassingY

., 48

(v=)

Vol. l5ll

I in mm. 4749, employing Va/,

4s

u$( r

support this view of the renrrn to I through a passing V%, both in their
graph [Example 2] and in their discussion of the piece cid e.arlier.)
this view is accepted, it calls into question Schenker's interpretation not
only of this Eansition, but of tonal strucrre and tonal hierarchy within
the entire middle section of the piece. Va/r, after all, represents a
perfectly reasonable transition from ltr|, to I, but an inrpreation that
views it as a link between VI and I requires one to consider both the E
major harmony and ttre Va/, to be voie-leading chords, with V of VI
reinterpreted as IIIfi only in the foreground as part of tlis transition.
Employing this E major harmony in such a direct voice-leading
relationship to I, however, suggests that it might have a mor strucral
role as IIIif throughout the section. As the earlier discussion indicated,

If

a similar E major harmony was the goal of harmonic+ontrapuntal


motion following the cadence on I in m. 16. Might there be a

connection benveen this firststable harmony of the middle section (mm.


'some
23-24') and the last one (mm. 47-48)? Might there be
way to

consider the middle section of the piece to be based upon a


prolongation of this harmony rather than VI?
Examples 7a-c present a view of voice leading and harmonic

Cinnamon, E Major Hannony as Dontitttt or Medant

l3

stncture in the entire piece that incorporates a prolongation of ltr#


rather than VI within the middle section; Example Za illustrates rrm.
123, the ening section of the piece and the transition to ltr#;
Example 7b illustates mrt. ?-049, the end of the trasitional sequence
leading to III#, the middle section of the piece,and the retransition to
I; and Example 7c illustras mm. 49:77, the closing section of the
piece. Tonal structure between mm. 23 a47 @xample 7b) can more
clearly (and more consisntly) be viewed as a prolongation of E major
(Itrf) in which.the potential ambiguity of function inherent in this
harmony that allows it to be interpreted as V/YI is exploid for
expressive purposes. When viewed in thiq way, IILff is seen to be
prolonged, in a very conventional way, witfl a middleground harmonic
progression, I-IV-V-I. Though the E major harmonies are approached
each time as if they were V of A minor (or major), this is only a
foreground element.
. After its anival in m. 23, III# proceeds towads its own fV
through a normative descending-thirds progression that arrives at IVI
(of E) in m. 29 and then proceeds on to fV (m. 3. This IVI is
marked off as a midpoint in this progression through the durational
accenJ given its own dominant (mm. 27-28), by the emphasis given the
resolution of this dominant as a result, and by a change in the
sequential pauernthat follows it. The IYI chord is approached through
a succession of dominant-seventh harmonies. The circle of dominantseventh chords continues after m. 29, but vith inrvening passing
chords thatproduce half-sp melodic motions in the bass, anticipating
the arrival on IV (of through a dominant-seventh chord on B b that
is slightly alred and reinrpreted as a French augmented-sixth chord.
fV arrives in m. 35 in the form of an A major harmony: diatonic
if considered IV within a prolongation of E major, bu repreSeiting
mixre if considered I in A minor. As nod earlier, this is the oily

harmony between nn. VI and 46 that does not contain a seventh. It is


thereby distinguished from the surrounding harmonies, all of which are
relatively less stable. Earlier it was observed that the apparent stability
of this harmony was inconsistentwith Schenker's complete omission of

it

from his analysis.

It is, however, totally

consistent

with

the

interpretation of its function offered here. This fV is then prolonged by

14

Indona Tlory Reviqv

Vol.

15/1

another sequential passage based on a circle of fiftbs sequence, this


lime employing a 1G7-10 sopranebass voice-leading pattern to descend

by step to its third (now minor), which arrives in m. 44. The voiceleading pattern that follows then continues.briefly, as a link to what
would be tr of A minor, which arrives in m. 45. Tliis harmony is
immediately altered, however, becoming V of E major and leading to
a perfect authentic cadence on ltr# in m. 47. The seguence beveen
mm. 36 and 4 offers the strongest support yet of an interpretation of
this section in rms of A minor. It is completely diatonic within that
key, suggesting the middleground unfolding of a tonicized A minor
harmony. This interpretation, however, is eally compatible with the
view taken here, which also considers the A minor harmony to be
unfolded, but on a more foreground level and within a conxt that
interprets it as fV of E (I[#) rather than I of A.
While the IV chord of m. 35 is major Qn keeping with its
function within E major), the use of a sequence that is predominantly
diatonic in A minor achieves two effects. First" it perpetuates the
suggestion of A minor, begun with the approach to III# as VAy'I, to
substantial expressive effect. Second, it results in a disproportionate
emphasis on the anival of V (of when the B diminished chord,
approached as tr of A minor, is chromatically alred to become a
major triad (V of , causing it to stand out from the sequence as a
structural harmony that frnctions on a deeper level. This differentiation
of the B harmony is further reinforced by a shift of regisr in the bass,
which had until m. 44 continued consisntly downward by sp from
A to C. This octave shift places the bass of the V chord (m. 4 in the
same regisr as that of the fV chord (m. 35), zupporting their voiceleading and harmonic associations. Particular notice should be paid to
the role of the harmonies that Schenker identifies as tonic within the
middle section of the piece. Here, both are considered relatively
foreground elements that function within trasitional passages, an
interpretation that seems more in keeping with their roles as parts of
ongoing processes. The apparent 6/5 harmony of m. 17 is seen as the
initial chord in a sequence leading from I Qn C major) to III1, while
the A chord of m. 25 is considered part of a circle of fifths, connecting
I (of with its lYI.

Cinnamon, E Major Harrony as Domilnnt or

Medant

15

Example 7a. Voice leading of opening section and transition to

III#

(mm. l-23)

.ivr.
rY

Itr

(mm. 2G4t9)

Indiana Theory Reviqv

Vol. l5ll

El=
H

l'

r\

return to

t.

lr
lt

tl

H-

I
t

tl

tl

tl

.t

t)

',

tt

li

rl

,l
tl

r'

tl
l

Example 7c, Yoice leading of closing section (mm. 49-77)

Cinnamon, E Major Hannony as Domnont or Medaw

't

Example 7b. Voice leading of end of transition, middle section, and

16

o
o

17

-e

vrI
#

III

( vI)

FIGURE

5! Yotce Leading of Eatlre Pl Interprtfng the

Prolonglg III'.

1dd1e

Stlon

69

7t

vlr

'

Indiau Theory Revew Vol.

l5ll

Itisi

t9

Example 8. Summary of the tonal structure of the entire oiece

contradicts

within

Cinnamon, E Major Hamony as Domirunt or Mediant

the

perspective,
assertion of a key and the prolongation of its tonic. Bear in mind that

15

2J S5a6a?

a6 .9

"H:
chords
taken here acknowledges the implications of A, but seqs them as an
expressive element, employed as a foreground device within a
prolongation of III that treats it as if it were v/Tr. It draws a crucial
and often-overlooked distinction be tut n prolongaton and ncznrion:
specifically, that to be prolonged a harmony need not be treated as
tonic nor even oocur in a context where it could be considered the tonic

eof
any

that
Whi
'

(v)

YI

two pieces share.r' But this reading also promotes discussion of other
concepts_not t'pically addressed in analyses guided by Schenker's
graphs ofthis piece.

eading within this piece

Gilbeft's fxanpt Z-

IVI-I, it is no less viable, and this interpretation appears to overcome


the most serious problems pointed out in schenker's analysis. It also
ess conventional than

offers several pedagogical advantages, which make this tude even


more usefirl as a "teaching piece." of course, this inrpretation still
includes the strucnual similarities with the Bach prelude: the E-F-E
neighbor motion. that
the employment of
parallel tenths (in mm.
a descending octave
transfer of E in the sop
f tonic and dominant
pedals as a means of emphasizing the prolonged harmonies that form
the background close of the piece are all strucurral features that these

rsee

the analyses of the Bach prclude cited above for illustntions of the
roles of
these featrres within thatpiece.

Indiana Theory

Rqew Vol.

l5ll

thc rrnrn !o its tonic are associated with changes in design that call

antlon to those strucnral events. In a piece such as thiJ, where a


rcpCltd.fl8Uration pattern seerrs to go on indefinily, even slight
chfncr ln dceign can signal important strucral events.
Iltio picce also offers a useful example of how foreground

tructutoE, particular towad the end of a piece, can summarize tonal

The Place of Schenkerian Analysis


fui Undergraduate and
Graduate Curricula

David Gag4
the piece is an excellent example of this common structurar practice, a
discussion of which might compare the rehrn to I in m. 49 with ttrat

of the role of

each augmented-sixth chord as an indication of a


structural harmony to follou and as a mens of generating harmonic
ambiguity by suggesting the next chord as a dominant might be
pursued. One could begin by pointing out how both V and III in the
second phrase

of

A'

ixth chords,

calling them to one's


dominant in
a yet to be esblished key. One could then continue to consider how
approaches through augmented-sixth chords affect the perception of Itrl
in m- 23 and fV (of III#) in m. 35.
Most important of all, however, is the lesson this nalysis can
teach us about the analytical pr(rcess itself. A comparison with
Schenker's origi
r's canpoint
outquiclearly
orinstnrctor

prefers) the need of every analyst to consider all possibilitie and weigh
all ambiguities before arriving at a conclusion. schenker's analysis is
so appealing because it coincides with our perception of the implied A
minor tonality on the surface and confirns our suspicions about its role
in the piece- If this alrnative analysis teaches us nothing else, it shows
us the value of considering alrnatives for what they can contribute to
our overall view of a piece, even if we ultimaly reject them.

Introduction
Since Schenker's death some sixty years ago, many principles and ways

of thinking that he first introduced have become an integral part of


musical discourse. Conceps such as prolongation and the notion of
strucrral levels are 4ow freguently taught o music students, and
analytical graphs are commonplace in theory journals.
At the sarne time, the field of Schenkerian analysis has expanded
and diversified in the last three decades- Sndies of pr-Baroque and
twentieth*ennrry works have explored strucnre from a Schenkerian
perspective in repertoires that Schenker himself did not address.
Current reseach is expanding the consideration of rhythm, texIe,
orchestration, form, and other compositional featrres in relation to

structure.l
tSee

d.,

Trends in Sdmkeian eserclt (New


1990); William Rothstein, Phrase Rthm inTorcl Music (New York:
Schirmer, 1989); Janet Schmalfeldt" "Tomrds a Reconciliation of Schenkerian Conce'pts

for orample Aller Cadwallads,

York Schirmer,