Musical Proportion and Formal Function in Classical Sonata Form: Three Case Studies
from Late Haydn and Early Beethoven
Author(s): James S. MacKay
Source: Theory and Practice , 2004, Vol. 29 (2004), pp. 39-67
Published by: Music Theory Society of New York State
Stable URL: https://www.jstor.org/stable/41054354
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Music Theory Society of New York State is collaborating with JSTOR to digitize, preserve and
extend access to Theory and Practice
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Musical Proportion and Formal Function in
Classical Sonata Form: Three Case Studies from
Late Haydn and Early Beethoven
James S. MacKay
"Sonata form" - the analytical brainchild of Antonin Reicha, A.B. Marx, and Carl
Czerny in the first quarter of the nineteenth century - remains today a problematic
paradigm. The inadequacy of the "textbook" model to explain the musical
choices of especially Haydn and Beethoven is evident whenever one examines
their sonata-form compositions. Consequently, writers from Donald Francis
Tovey to William S. Newman and Charles Rosen have sounded various cautionary
notes concerning this model, and have inspired others in recent times to search
for new, more flexible means of describing how Classical sonata form functions
musically.
The past few years have witnessed the appearance of two analytical
approaches that can assist us in the quest. First, there is William Caplin's taxonomy
of Classical instrumental music at the level of the four-measure phrase and
the two-measure phrase member.1 It provides us with the technical means to identify
thematic and transitional units and to distinguish between them on the basis
of their syntactical components. Second, there is James Hepokoski and Warren
Darcy's identification of breaks in musical action (i.e., medial caesuras and essential
expositional closures) as analogues to punctuation in language, which they
use to demarcate formal events in sonata expositions.2 Both approaches draw
heavily on historical views of musical form: Caplin's ideas about formal function
derive from Schoenberg and Ratz;3 Hepokoski and Darcy's interest in musical
breaks as large-scale formal determinants is an extension of Heinrich Christoph
Koch's concept of "melodic punctuation."4
Hepokoski and Darcy suggest a rather wide - but nonetheless distinct range
of possible temporal locations, within a sonata-form exposition, for each
type of melodic punctuation. Their data suggests that the temporality of musical
events is more integral to a listener's perception of musical succession than has
been hitherto assumed. Indeed, it could be argued that the musical proportion of
an exposition's formal components (i.e., main theme, transition, subordinate
theme, closing section)5 influences subconsciously how we formally partition a
sonata exposition, regardless of the literal form-functional meaning of each expositional
segment, as determined by Caplin. Accordingly, this study seeks to
39
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
40 Theory and Practice
reconcile and combine the two approac
how a listener experiences sonata form
expositions, but the principles sugg
understanding of sonata-form movem
conclude with an examination of a c
Symphony No. 90.
In introducing the discussion at hand,
Hepokoski and Darcy's analytic approa
Caplin proposes that each musical c
ciñc formal function within a composi
as an initiating, intermediate, or conclu
units then combine in different ways
eight-measure sentence is formed from
ately repeated and then followed by fr
common structure for a main theme)
non-thematic region such as a transiti
mal functions can be employed differ
thematic locations within the form. T
main themes, subordinate themes, and
resemble the last-mentioned formal re
they contain, and the order in which
between tight-knit and loose thematic
tactical components. Tight-knit th
temporally balanced phrase members
sisting of a pair of two-measure phra
theme). Techniques that create mus
(such as fragmentation, extension, a
organization.6 Caplin shows how form
main themes, while formal loosening
group. (Transitions and developments
themes as to their formal structure,
between formal regions of relative ton
Although they are conversant wi
Darcy take a different tack. Rather t
from small motivic kernels (à la Scho
musical punctuation as a way of demar
sonata form. As of this writing, their
in full; an article entitled "The Media
Century Sonata Exposition" (to whi
article, "Beyond the Sonata Principle")
However, to the extent that ideas pres
those of Caplin, the article provides
sonata form.
As a preliminary step in demarcating the various events of a sonata-form
exposition, Hepokoski and Darcy identify points at which the musical flow
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 4 1
breaks, taking their lead from Koch's notion of "melodic
emphasis on musical punctuation displays the authors' interes
based analogies to explain the exposition's succession of even
forms of musical punctuation are the medial caesura and th
ordinate theme's conclusion, which effects essential expositio
presence and location of the medial caesura is crucial for dete
structure of a sonata-form exposition, I will describe it at len
As Hepokoski and Darcy note, the medial caesura typ
exposition more or less at its midpoint. This cadence can be o
half cadence (HC) in the tonic, a HC in the subordinate key,
tic cadence (PAC) or (more rarely) imperfect authentic c
subordinate key; the second of these types is by far the most
ures leading to this caesura are often rather dynamic in char
Darcy note that the harmony preceding the medial caesura
chord (e.g., a secondary dominant or augmented-sixth chord)
orous, highly active," and the dynamic level is quite loud, usu
threefold "hammer blows" reiterating the goal harmony of
this juncture in the musical flow is often highlighted with a gran
ish.
One of Hepokoski and Darcy's major contributions to a deeper understanding
of sonata form is their acknowledgement that temporal location is a vital
determinant of formal boundaries. (Their three subcategories of medial caesura,
listed above, occur progressively later in the exposition.) This attention to temporal
location is tied to a specifically Classical sense of musical balance. For a
sensitive listener attuned to the norms of the style, it is often possible to get a
sense of where one is situated formally in a work by Haydn, Mozart, or
Beethoven, on the basis of musical proportion. The ensuing discussion will
explore how this balanced musical proportion operates in Classical music at various
levels of structure.
Even at the level of the two-measure phrase member (the smallest formal
level in Caplin's system), a certain temporal balance is expected in Classical style.
Indeed, a sense of balance and regularity is often considered to differentiate
Classical music from that of the bordering Baroque and Romantic styles. Caplin's
theory considers the four-measure phrase, made up of a pair of equally weighted
two-measure phrase members, to be the normative model. Any alteration of this
standard pattern is an example of musical extension or contraction. Thus, one
would hear (for example) a three-measure phrase member in the context of the
typical two-measure unit (or four-measure phrase) that it deforms.12
This notion of balance occurs at the thematic level of structure as well.
Classical themes are typically built out of phrases of equal length: the proportions
of phrase 1 tend to demand an equivalent response, temporally speaking, in phrase
2. Thus, the standard four-bar antecedent typically demands a consequent of
roughly equal length. Similarly, an irregular antecedent often demands an equally
irregular consequent as a counterbalance. This point is illustrated by the minuet
from Haydn's Quartet in G Major, op. 54, no. 1 (see Example 1): its opening tenmeasure
theme balances a five-measure antecedent phrase with a consequent
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
42 Theory and Practice
Example 1. Haydn, String Quartet i
third movement, mm. 1-10.
ANTECEDENT CONSEQUENT
Basic idea Contrasting idea (extension) Basic idea
(HC)
Contrasting idea (expanded)
7
^(P
ph
an
m
en
w
Th
op
in
th
ic
ph
ex
te
w
co
fe
ba
on
ba
de
Ha
m
Ca
a
lar
se
m
ke
sim
te
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 43
Example 2. Haydn, Piano Trio in G Major, Hob. X
first movement, mm. 1-23.
PARTI
ANTECEDENT (4 measures) CONTINUATION (6 measures)
(HC)
(PAC, D major)
PART 2
CONTRASTING MIDDLE (4 measures) CONTINUATION (7 measures)
(HC)
_^ (codetta)
(PAC)
the overall procedure in the larger for
smaller one that it expands.
The concern for proportion implici
view of sonata-form expositions. Their
arating the Principal theme-Transitio
departs from the home tonality, from
which introduces and then confirms t
sion feels like a "halfway" mark in its
temporal location. They furthermore d
basis of the presence or absence of suc
in structure if a medial caesura is pre
exposition") if it is absent.15
As Hepokoski and Darcy explain, t
determined by its harmonic goal) helps
ment in a sonata exposition - at lea
exposition establishes a home tonality
purpose of establishing a complemen
departure from the original tonality, s
along in the form the various types o
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
44 Theory and Practice
occur. Not suq^risingly, a HC in the
default" for the medial caesura) typical
does a HC in the subordinate key, or a
Darcy's second-level and third-level
cadence types are each a progressively
the home tonality and establishing the su
tive temporal placement in the flow of
If a medial caesura arrives "too so
likely be unconvinced musically of its
this moment-by-moment perception
might break down a complex type of d
even at first hearing. To begin with,
length. This length, moreover, is alre
opening phrase (since, as suggested a
paired with a second phrase of equal le
cal proportion for the succession of m
work being experienced for the first t
Regardless of initial impressions, o
poorly) executed proportions in subse
upon revisiting the complex forms of
the formal subtleties and ambiguities
giving way to a complementary tonali
will now be considered through two
ambiguous medial caesura: the first
Major, op. 10, no. 2; and Haydn's Piano
Beethoven's Sonata in F Major has b
Haydn's joking style.18 Certainly its o
norms and thematic devices of Classical sonata form in a willful and comic manner.
As illustrated by Example 3, the movement departs its home tonality very
quickly: after only twelve measures of main-theme material in a fast duple meter,
Beethoven leaves the home key in the subsequent phrase (mm. 13-18) and leads
(rather oddly and abruptly) to a HC in A minor! However, in m. 19, he immediately
establishes the expected subordinate key of C major; and so one might
consider the anomalous cadential goal (V of the minor mediant key) to be the
exposition's medial caesura, despite its early placement and "incorrect" tonal ori-
entation.
If there were no other candidates for this point of punctuation, we would be
forced to accept m. 18 in retrospect as the true medial caesura. To cite a related
case, the first movement of Beethoven's "Waldstein" Sonata in C Major, op. 53,
modulates to V of III (a HC in E major) after departing the home tonality. Because
no conventional half cadence in the dominant appears later in the exposition, this
odd cadential goal, leading toward the mediant major key, is apparently to be
taken as the exposition's true medial caesura. (This reading is confirmed by a
twelve-measure passage prolonging the dominant harmony - Hepokoski and
Darcy's "caesura fill"19 - and subsequently the onset of the subordinate theme in
E major.) However, such a cadence in the "wrong" tonal location could certainly
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 45
Example 3. Beethoven, Sonata in F Major, op. 10, no
first movement, mm. 1-19.
MAIN THEME ^
Piano < p
^Tf ^x p-r-^ TRANSITION
8 ^^^ JÖ
(PAC, F maj
/^j" "% i -*
< 14 WWf^ * /^r - ^ ^t~^n ,^r~"x ^
(HC, A minor)
produce, in the stylistically aware listener, an expectation that a more normativ
medial caesura (i.e., HC in the home tonality, or a HC or PAC in the compleme
tary tonality) will occur at a later point in the exposition. Such is Beethoven's
procedure in the F-major sonata. Both the odd tonal goal of the putative media
caesura and its seeming premature appearance in the exposition hint that a subs
quent musical correction will need to occur as the sonata movement unfolds.
Given the main theme's twelve-measure span, combined with the Classica
ideal of balanced musical proportion, one might surmise that the transition will
of similar length. When this segment cuts off abruptly in the "wrong" key, aft
only six measures, the effect is deliberately perfunctory. Nonetheless, there is l
tle that is unusual about the putative thematic unit that begins in m. 19 - at lea
not from a form-functional standpoint. The theme can be interpreted as a (slight
expanded) sixteen-measure sentence, one of Caplin's compound theme mode
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
46 Theory and Practice
Example 4. Beethoven, Sonata in F
first movement, mm. 18-37.
SENTENCE
Piano < 18 i* ^ ~^IZ
, /-- - 1 "ß- W ^" ^ (continuation) u ^ - ^/*~~~ >s
< 24 ^^
-^
^
HC (C major)
34 6 6 6 J^^a
Mm. 19-26 can be unde
a basic idea stated twic
harmonic shape of the
the subordinate tonalit
length of this sentence
as opposed to the usual
switched in midstream
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 47
ically analogous to the two-measure norm. (Such a shift is
Caplin never uses in his analyses.) However, that reading seem
factory. Perhaps the irregular length of the material's formal uni
theme is not what it appears to be, at least with regard to its f
Moreover, in spite of the apparent initial aptness of this p
the subordinate theme, the arrival of this new thematic reg
tonal region seems distinctly premature - after all, we are bu
ures removed from the main theme. As a result, the subordina
be called such) must do some of the rhetorical work usually al
tion, as Rosen has noted.21 Probably due to its early appearanc
this theme, following its normative presentation phrase, quick
acter of transitional material, leading (as traditions are wont t
proper subordinate key (as shown in Example 4). It is as if
introducing the "correct" medial caesura, which he had with
"correct" cadence on the dominant of C major is more convin
porally than the early HC in A minor: the requisite G-major ha
30, and is then prolonged through m. 37. Because the expositio
ures in length, this harmony and its subsequent prolongation
at the midpoint. In contrast, the "wrong" cadence in m. 18 can
as having occurred far too early.
Beethoven's playing with formal boundaries and listener e
exposition has still not reached its conclusion, however. The
the medial caesura (m. 38ff.) doesn't sound like the onset of a
shaped like the concluding half of a longer thematic unit. Acc
and Darcy, we would normally expect a beginning gesture to o
caesura, but Beethoven follows the grand rhetorical pause on
a thematic region that seems to begin somewhere in the midd
As shown in Example 5, mm. 38-41 can be explained as
cadential gesture (following Caplin), immediately repeated
parallel minor. One might argue for an IAC in m. 41, had not
ment in the musical flow been so ambiguous. M. 45, whic
clarified tonal and form-functional matters with a PAC
resolves deceptively to a remote Ab-major chord (bVI of C ma
ing" of mm. 38^1 is revealed instead to be a pseudo-cadential
that subsequently opens up into the subordinate theme's "true
(mm. 47-55). Although it should be noted that Classical th
begin as if in the middle of a longer idea (e.g., the main theme
ment of Beethoven's op. 31, no. 3, can be understood in this
to complicate further the clear establishment of the expositi
aries.
What ensues is the true conclusion of the subordinate theme: Beethoven
follows the pseudo-cadential initiating gesture of mm. 38-45 with an expanded
cadential progression (ECP),22 leading to the expected PAC that confirms the
dominant tonality in m. 55. Following this cadence, the final twelve measures of
the exposition are a non-problematic closing section: the harmonic content almost
exclusively alternates tonic and dominant chords, thus serving to ground the formal
tension that Beethoven had created earlier in the movement.
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
48 Theory and Practice
Example 5. Beethoven, Sonata in F
first movement, mm. 38-55.
Basic idea? (pseudo-cadential)
Piano < 38
(basic idea repeated)
ñ
I jf
*^ S/^
(ECP) _ . ^ . JL ^ A ^
«
< 46 PP cresc.
(PAC)
One could say that both Caplin's
analytic systems are helpful (to some
placed the formal boundaries in this
tional reading of the exposition's fo
theme arrives early (if a bit strange
later onset for the subordinate them
mine the location of the subordina
30-37 is surely far more emphatic t
discrepancy in analytical results is m
rather, it aptly illustrates this expo
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 49
Example 6. Haydn, Piano Trio in Fft Major, Hob. XV:2
first movement, mm. 1-16 (piano part only).
Antecedent
mf J fz PPiano < J
' (HC, F# minor)
Failed consequent
' F ^
u<;H|
theory of formal funct
tively demonstrates th
accomplished from the b
ushers in the exposition
Beethoven's deformati
sonata greatly resembles
his Piano Trio in Ft Min
be segmented into a mai
six measures that begins
in Example 6, Haydn acc
common time. The open
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
50 Theory and Practice
the second phrase begins with the same
consequent phrase) but subsequently and s
in the relative major (i.e., A major).
Caplin views the opening phrase as a fo
sition ensuing in m. 5.26 To make this re
movement as R=1/2N, so that the open
equivalent to an eight-measure senten
include the opening movements of Beeth
and the "Waldstein" Sonata, op. 53. The ap
the opening material cuts across formal b
lowed by a transition that begins with t
initiating procedure for the transitional
of mm. 1-4 as the complete main theme,
in two different ways. Mm. 5-8 could serv
m. 9 would be the onset of an extended su
in m. 28. Alternatively, mm. 5-21 could
sition.29 In this reading, Part 1 of the t
theme incipit, and subsequently leads (so
tual subordinate key; and Part 2 (mm.
rhetoric (increased harmonic and rhythm
sage terminates with a second HC in the
goal is prepared at length and confirmed
To illustrate a different view of how t
we turn to H.C. Robbins Landon. He anal
bridge passage (i.e., transition).30 Thus, h
parallel but oddly modulating mm. 5-8 to
gests that m. 22 comprises the unequivoc
A major. This tonal arrival, along wit
Hepokoski and Darcy's signals for the on
and the onset of distinctive thematic mat
tion, leads Landon to identify the p
movement's subordinate theme (see Exam
I have suggested a range of possible for
exposition, informed by the observation
Landon. It is not my intention to privileg
them are plausible. Which reading one p
which HC in the subordinate key is inter
and second, how one reads formally the o
As noted above, if the movement's fir
theme and transition regions, then th
Caplin's criteria - must be a single sub
beginning in m. 29). 32 Despite its prese
clear articulation of the subordinate key (
be considered a continuation phrase in C
onset of the subordinate theme region. Th
ceding segment fully, and Caplin requires
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 5 1
Example 7. Haydn, Piano Trio in F# Major, Hob. XV:26
first movement, mm. 16-28 (piano part only).
"" ' | ¡ »t if " " | Sr " Fr I t
(medial caesura plus fill)
<P f r r r r
l^V^ffî r r rHC(A+)
(Subordinate theme?) £ ^ _
24 6 3 6 6 fi3 6
PAC(A+)
authentic cadence.33 Landon's reading acknowledges the initiation function of m.
22, but fails to explain how mm. 9-21 can be a bridge passage (transition) when
it is in the key of its cadential goal from the outset. However, the rhetorical signals
of transitional material in these measures, combined with the convincing
medial caesura in mm. 19-21, strongly support Landon's formal reading over that
derived from applying Caplin's analytical principles.
Caplin's theory does provide a graceful way out of this quandary, although
he does not invoke it for this movement in particular. Much like the piano-trio
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
52 Theory and Practice
Example 8. Haydn, Sonata in E Ma
Antecedent
Piano <
íg&^íp=í - F - t - k=^=4i-J^
1 IC in E major
Failed consequent
5 _
HC in C# minor
movement under consideration, the finale of Haydn's Sonata in E Major, Hob.
XVI:31, begins with a main theme whose first two phrases both end with HCs (as
shown in Example 8, the first phrase ends on V of the home key, while the next
phrase ends on V of the submediant minor). In effect, the theme's opening phrase
creates the expectation of a period, which is not carried out in conventional fashion
by its second phrase. As with the opening measures of the piano-trio
movement under discussion, it would be problematic to designate the initial eightmeasure
unit of Haydn's E-major sonata as "antecedent plus antecedent." The
remote tonal goal of the second phrase has a different musical effect than the first
phrase's normative tonal goal (i.e., the tonally further-ranging second phrase has
continuation function instead of initiation function). Thus, Caplin designates the
second phrase as a "failed consequent" that - instead of returning to the tonic effects
a greater tonal departure.34
This strategy, though anomalous by high Classical standards, is not uncommon
in the mid-eighteenth century. Alan Campbell has pointed out that a periodic
structure in which the consequent phrase has a tonally more-remote cadence than
the antecedent phrase is a common theme-type in the music of Domenico Scarlatti
and his contemporaries.35 Such a procedure is not atypical in Haydn's works up
to about 1765 (Caplin's example of this theme-type dates from a decade later
still), at which stage in his career Haydn's formal structures were still greatly
influenced by the phrase-structural strategies of the late Baroque. It is a bit surprising,
however, to see the composer returning to this procedure in 1795, the date
of the Piano Trio in F# Minor. However, this theme-type was obviously part of his
vocabulary even in his mature period. It is plausible to read the opening eight
measures of the trio as an anomalous period (antecedent plus modulating "failed
consequent"). After such a formal designation, the exposition proceeds convenThis
content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 53
tionally as to its formal function: mm. 9-21 would be the tran
a conventional SC region (mm. 22-37).
This later placement of the S region seems rhetorically co
above, m. 22 is marked by a new and more forceful establishme
major, and by new triplet figuration. Although the subordinate
m. 9 (Haydn even precedes this moment with the requisite HC
ity), one senses, even on first hearing, that it is likely prema
begun the subordinate theme: if m. 8 is the medial caesura, it
early, only 22% of the way through the thirty-seven-measur
reads m. 22 as the onset of S, the proportions seem more satisf
caesura of m. 19 occurs 5 1% of the way through the expositio
normative placement from the standpoint of musical balance.
As a final example of how musical proportion affects our
form, we will examine the finale of Haydn's Symphony No. 9
annotated score reduction of the entire movement appears in t
article. This quicksilver movement (marked allegro assai) begin
square eight-measure period, immediately repeated with new
sixteen-measure span comprises a main theme that is almost u
forward, permitting Haydn plenty of room for formal loosenin
unfolds.
The remainder of the exposition is problematic as to it
whether one invokes Caplin or Hepokoski and Darcy. The
returns in the dominant key in both mm. 30 and 54. Either of
tures could be heard as a principal-theme-based (P-based) subo
P-based S is a typical strategy for Haydn, as Hepokoski has not
thematic statement is preceded by a proper medial caesura. The
ible cadence near the midpoint of the exposition is a HC in G
considerably after the first putative P-based S, and too far in
ond P-based S (m. 54) to be heard as a medial caesura follow
Since the exposition spans seventy-eight measures, the weak H
tually at the exposition's midpoint; but this juncture lacks the r
that might have made it a convincing candidate for a medial ca
Of the three possible choices (mm. 29, 38 and 47), the last
sage is probably the best candidate for a medial caesura. In spit
harmonic goal (vii0? of G major replaces the expected V chord)
rhetorically. Haydn prolongs this dominant-functioning chord thro
which a monophonie caesura- fill continues through m. 53. Th
falls short, however: it is surely insufficient for the rhetorical eff
caesura to be present if the proper cadential articulation is not
earliest candidate for the exposition's medial caesura (m. 29) h
lem; and m. 38, the only HC following the main theme, is
plausible S material, as we will explore further below.
The absence of clear cadential articulation complicates the
subordinate theme according to Caplin's principles as well.
beginning of the transition with no difficulty: this formal reg
with a false closing section (containing post-cadential mat
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
54 Theory and Practice
pedal)37 followed by a move to V of V
cadence on this scale-step (V of G ma
thwarting the expected HC in the subor
material on the dominant in mm. 30beginning
of the subordinate theme grou
sition continues past m. 30, then the HC
However, the material that follows this
mm. 38-46 are the most "transitional" t
repetition and heightened chromaticism
harmonic goal of this phrase (m. 47) is n
Haydn ought to precede the subordinate
tonal substitute. Haydn then states main
nant (i.e., a P-based S) in m. 54. Due
preparation of this juncture with a domi
key, the ensuing passage (mm. 54-71) se
ordinate theme, but neither Caplin's nor
us to view this reading as incontestable.
The best solution for this formally p
abandon the search for a normative S re
enty-eight measures as a three-part exp
terminology, a continuous exposition).3
1-16) elides with an expansion section
with closing material (mm. 71-78). The r
measures is a bit lopsided, but it seems
that the material from m. 16 onward ne
a convincing S region. As Hepokoski a
search for a second theme [in a continu
derstanding of [its] internal processes."
regarding this exposition, as its musical ef
inant tonality. The complementary tonal r
of P material on the dominant (mm. 30
exposition's concluding measures.
Haydn follows this problematic exp
development based largely on main-the
rial (orchestrated for solo flute with lig
tonic, and thus is a possible candidate f
this passage begins in m. Ill, only twen
one would not expect it to represent al
Even though development sections that
sitions are not unknown in Classical styl
too early, proportionally speaking.40 (M
material with the rhythm of S from mm.
a formal restatement.) In any event, thi
a sequential progression as the developm
major is in place from m. 138, creating
precedes the recapitulation's "double ret
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 55
The recapitulation begins in m. 142. Because Haydn had st
material four times in the exposition (beginning in mm. 1, 9
would expect him to do a bit of musical pruning in the recapi
ments of main-theme material, all in the tonic key, would be
Haydn doesn't merely cut away at the recapitulation with pr
ever; instead, he uses a chainsaw! He recapitulates the first eig
main theme, follows it with a transposed restatement of mm.
nate theme by default, minus its opening phrase), stated this time
concludes with a four-measure post-cadential unit followed b
silence.
Although the passage from mm. 142-67 recapitulates both P and S versions
of main-theme material and ends with an emphatic cadence, one can scarcely
believe (given this material's brevity compared to the exposition and development)
that the movement is truly over. The recapitulation has all of the syntactical
components that are necessary to effect formal completion, but proportionally
speaking it is far too brief to be rhetorically adequate. However, the four measures
of silence that ensue shake one's confidence. Perhaps the movement has
indeed concluded, albeit strangely.41
Then, in m. 172, the coda begins in the remote key of Db major! (One imagines
that Haydn would have taken great pleasure had an unwary soul dared to
applaud during the four-measure rest, on the reasonable expectation that the
emphatic cadential gesture of mm. 162-63, and the four measures of stock postcadential
material that had ensued in mm. 164-67, had marked the end of the
movement.) The coda extends for seventy measures, virtually the same length as
the exposition, a vastness that is unprecedented in Haydn's sonata-form movements.
As the coda unfolds, Haydn explores various combinations and reworkings
of the exposition's main thematic material. Once the section has extended sufficiently
to counterbalance the exposition, Haydn ends in a great cadential flourish.
The finale's recapitulation and coda together span exactly one hundred measures,
an adequate counterpart to its seventy-eight-measure exposition.
In this essay, I sought to explore how a sense of musical proportion assists the
analyst in determining formal boundaries in the Classical sonata. As the preceding
examples demonstrate, both Haydn and Beethoven occasionally use the
expectation of musical balance to deceive the listener about formal location. It is
apparent that a sonata exposition sets particular expectations of section length
upon completion of the main theme (or even upon completion of the theme's first
phrase). If these expected boundaries are at odds with the musical material's rhetoric,
then there can be a leakage of characteristics from one formal region to
another. Main themes that depart from the home tonality too soon (given the proportions
of the exposition) blend with the ensuing transition. Subordinate themes
that arrive too soon may take on transitional features. Subordinate themes that
arrive too late (e.g., those that follow Hepokoski and Darcy's "third-level default"
medial caesura after a PAC in the subordinate key)42 tend to sound like closing
material.
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
56 Theory and Practice
This sense of Classical proportion c
sonata-form movement, as the length o
length of both the development and t
return" (of main theme and home key) in
be a false recapitulation; the tonal and th
viously established temporal proportions of the movement. Although
recapitulations tend to be shorter than expositions (because transitional and thematically
redundant material tend to be omitted or truncated in the recapitulation),
an excessive amount of musical pruning can still give the listener a sense of
rhetorical imbalance, as we observed in the concluding movement of Haydn's
Symphony No. 90, which required an extended coda to reestablish this balance.
As demonstrated by both Hepokoski and Darcy's article, and Caplin's book,
every thematic event in a sonata-form exposition sets up expectations as to formal
function and musical proportion, and these are either fulfilled or denied in subsequent
measures. The methods by which Haydn and Beethoven play with these
expectations, making the stylistically aware listener a partner in the creation of a
sonata-form movement's inner logic, constitutes one of the many musical riches
of Classical style.
NOTES
1. William Caplin, Classical Form: A Theory of Formal Functions for the
Instrumental Music of Haydn, Mozart and Beethoven (New York: Oxford Univ.
Press, 1998).
2. The latter is a work in progress, part of which appeared a few years ago: James
Hepokoski and Warren Darcy, "The Medial Caesura and its Role in the EighteenthCentury
Sonata Exposition," Music Theory Spectrum 19/1 (1997), 115-54. An
expanded version of the theory outlined therein will appear in Elements of Sonata
Theory: Norms, Types and Deformations in the Late Eighteenth-Century Sonata
(New York: Oxford Univ. Press, forthcoming). Hepokoski also touches upon many
of these points in "Beyond the Sonata Principle," Journal of the American
Musicological Society 55/1 (2002), 91-154.
3. Caplin, Classical Form, 3.
4. Hepokoski and Darcy, "The Medial Caesura," 115.
5. This terminology follows Caplin (cf. Classical Form, chapters 3-7, for the main
theme; chapter 8 for the subordinate theme and closing section; chapter 9 for the
transition). Where more concise, I also use the P-T-S-C abbreviations (Principal
theme-Transition-Subordinate theme-Closing) adopted by Hepokoski and Darcy
(cf. "The Medial Caesura," 121-22).
6. Caplin, Classical Form, 84-86.
7. Caplin, Classical Form, especially 125-77.
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 57
8. As Hepokoski and Darcy acknowledge in "The Medial Caesu
9. Hepokoski and Darcy, "The Medial Caesura," 115.
10. Hepokoski and Darcy, "The Medial Caesura," 1 19-23.
1 1 . Hepokoski and Darcy, "The Medial Caesura," 1 24-25.
12. Caplin, Classical Form, discusses such proportional irregula
from the norm"; see especially 47-48, 55-57.
13. The most extreme example of formal imbalance in the Classical
Haydn's Sonata in D Major, Hob. XVI:42. This binary-form c
eight-measure first part and a ninety-three-measure second
repeated, making the binary division incontrovertible). In the gr
ond part, the expected return of the home key is repeatedly su
avoided.
14. Cf. Charles Rosen's discussion of Koch in The Classical Style, rev. edn. (New
York: Norton, 1997), 87.
15. Larsen's two-part and three-part exposition models appear in "Sonata Form
Problems," in Handel, Haydn and the Viennese Classical Style (Ann Arbor, MI:
UMI, 1988), 269-79. Hepokoski and Darcy 's "continuous exposition" model
includes Larsen's three-part exposition model (in which the boundary between T
and S blurs, thus forming an expansion section) with other similar exposition models
that Larsen does not consider; see Hepokoski and Darcy, "The Medial
Caesura," 118-21.
16. Caplin has noted that all three of these (potential medial caesura) cadence types
may be present in a sonata exposition {Classical Form, 196, table 13.1). In some
instances, the presence of both a HC in the tonic and a HC in the subordinate key
may signal a "two-part transition" {Classical Form, 135-38).
17. See Hepokoski and Darcy, "The Medial Caesura," 126-27. They found that medial
caesuras occur from 15% to 75% of the way through the exposition. A HC in the
tonic key occurs the earliest, in the 15-45% range; a HC in the subordinate key in
the 25-50% (rarely 60%) range; and a PAC in the subordinate key in the 50-70%
(rarely 75%) range.
1 8. Rosen, The Classical Style, 414-15.
19. Hepokoski and Darcy, "The Medial Caesura," 127-28.
20. Caplin often reads the real formal structure of a composition as operating at half or
twice the level of the notated meter. In fast duple or triple time, sixteen measures
of notated material, with cadences spread out twice as widely as the norm, may
function syntactically as an eight-measure formal unit (R=2N). Similarly, in slow
compound or quadruple time, cadences may occur every two notated measures; in
such a case, it is preferable to show analytically that these two measures are analThis
content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
5 8 Theory and Practice
ogous to a complete four-measure musical
as R=1/2N. See Classical Form, 35.
21. Charles Rosen, Sonata Forms, rev. edn. (New York: Norton, 1988), 236.
22. An ECP is a cadential progression that spans an entire phrase; see Caplin, Classical
Form, 45-47.
23. Cf. Rosen's reading cited above, in n. 21.
24. Caplin discusses the formal ramifications of such a decision in Classical Form,
'21 -29.
25. Although the dominant chord achieves its customary root position only on the second
beat of m. 4, a HC is clearly perceived here. Haydn withholds the Ct on the
downbeat of m. 4 to highlight the voice-leading similarities of mm. 3-4 and 7-8
(the PA-Ei bass motion that concludes the opening phrase is transformed into
F#-F^, resolving to E, in the ensuing phrase).
26. Caplin, Classical Form, 21 A, n. 21. See also 129, n. 21, cited below in n. 28.
27. See n. 20.
28. Caplin, Classical Form, 129. As both the "antecedent" and the "conse
with a HC, one is forced to "reinterpret the first [unit] as a main theme and
ond as a transition," as Caplin states.
29. Caplin, Classical Form, 135-38.
30. H.C. Robbins Landon, Haydn, Chronicle and Works, vol. 3 (Bloomingt
Univ. Press, 1978), 435.
31. Hepokoski and Darcy, "The Medial Caesura," 125.
32. One could also consider this material (mm. 29-31, repeated in mm. 3
an abbreviated Subordinate Theme 2, as it would comprise six "real"
using Caplin's R=1/2N formula, thus potentially forming a compressed
measure" theme. Regardless, its closing character is incontestable.
33. Caplin, Classical Form, 97.
34. Caplin, Classical Form, 89.
35. Alan Campbell, The Binary Sonata Tradition in the Mid-Eighteenth
Bipartite and Tripartite "First Halves" in the Venice XIII Collection of K
Sonatas by Domenico Scarlatti (M.A. thesis, McGill Univ., 2000), 43-4
36. Hepokoski, "Beyond the Sonata Principle," 141.
37. Caplin, Classical Form, 129.
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 59
38. Cf. Larsen, "Sonata Form Problems," 275-76; and Hepokosk
Medial Caesura," 118-19.
39. Hepokoski and Darcy, "The Medial Caesura," 1 19.
40. Cf. Charles Rosen's discussion of multiple false recapitulations in Haydn's
Symphony No. 55 (Sonata Forms, 276-80).
41. As Charles Rosen has noted (in words that I have paraphrased to conclude this
paragraph), Haydn uses a similar procedure (i.e., an oddly concluding recapitulation
followed by an extended rest, and then by a coda that brings back missing
recapitulation material) in the opening movement of his String Quartet in E'> Major,
op. 50, no. 3; see Sonata Forms, 158-61.
42. Hepokoski and Darcy, "The Medial Caesura," 126-27.
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
60 Theory and Practice
Appendix.
Haydn, Symphony No. 90: fourth m
£ (MT repeated)
TRANSITION (expansio
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 6 1
* p n i p n' p n ' p "•" n'n"•" Cj+. ii
(HC?)
A A+ +. A m A* +. A m fi fi
^ iS it^Jiu +. m Lr i [^¿'Uu +
■^ LT | ¿-T "H-T^i j j „ j -3|,,j- ]/j|;jrj^rriHC (Cil)
^gjF'f i[f ff if'ffji^^inmj ivii07 (G+)
(dominant arrivai)
■>-,] '<UUiï * i i - i - |i
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
62 Theory and Practice
54 MT incipit (-S TV)
faiS^icrxr 54 MT incipit
_.. . , .inn .rar] r32.n
71 Closing Section
PAC (Gil
77 DEVKLOPMKNT
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 63
94 L
•y
100 ^^^^ ^^^ L L
105 L
i
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
64 Theory and Practice
FALSE RL-CAP. (becomes model-sequence)
116^
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 65
RECAPITULATION
138 ^^^ ^^b (MT)
fcecrcED-i^
„r Í if
Ml incipit (- ST, measure 58IT.) ^i^-- -..-
150 r j i ^i^-- m -..- r^ i m r^
dP- 150 F- r j -p -i ..f r r i
^=^=^l -p r -i I trn ..f .
>>=r f f |f frp7|r
ài t 't c " iE-EtE-
1^,3 Closing Section (codetta)
i ¿ i , i r ff r-p-~T-J
^^^^p^JE^Ì^^=ÌEé^; i , i r ff
PAC (C+)
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
66 Theory and Practice
180 .
tg
205 f S k k
^^"( 205
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms
Proportion and Function in Sonata Form 67
" ~m ~m ~m ~é
218 __■_- -
■fiiü
«* ** * i
This content downloaded from
147.251.6.77 on Fri, 11 Aug 2023 22:05:32 +00:00
All use subject to https://about.jstor.org/terms