Professional Documents
Culture Documents
Phrase Rhythm As Form in Classical Instrumental Music
Phrase Rhythm As Form in Classical Instrumental Music
SAMUEL NG
Theories of form in tonal music have largely been predicated on thematic and! or harmonic criteria.
By contrast, this essay supports Edward T. Cone's argument that the essence of tonal form is, by and
large, rhythmic. By integrating concepts derived from James Hepokoski and Warren Darcy's Sonata
Theory on the one hand, and analytical tools developed by William Rothstein and David Temperley
on the other, this article demonstrates that phrase rhythm can construct stylistic norms, establish
sonata types, differentiate between levels of compositional choice, and create deformations through
the expansion or even rejection of normative procedures. Thus the interaction of rhythm, meter, and
hypermeter within phrase structure proves to be a primary factor in understanding tonal form.
Keywords: phrase rhythm, musical form, classical instrumental music, Sonata Theory
-- - -- -
I II
concepts.f The ultimate goal of this paper is to re-envision and a)
resolve certain conceptual dissonances between Cone's views on
temporality and Hepokoski and Darcy's Sonata Theory in order
to define aspects of phrase rhythm and form currently underde-
I- I
-- - I I- I I- I- I-
-- - --- - -- --
veloped in the theoretical literature. b)
--- - -- ----
I
c)
As Mark Evan Bonds has proposed, historical studies on
form may be divided into two fundamental categories: confor- I I I I I
mational and generative. 9 The conformational approach identi-
fies a set of common properties shared by a large number of EXAMPLE 1. Three basicphrase-rhythmic configurations of
works; thus the analysis of a specific piece can include a com- tight-knit themes (adaptedfrom Temperley [2003])
parison to a norm. The generative view, however, treats each summaries of the most preferred options in each zone; any of these
48
p
y" , '" \\ ,
~
··
~
Temperley's crucial contribution lies not in developing a tax- phrases of the closing theme, without overlapping the end of
onomy of two-phrase structures, but in observing the affinity the previous section, enter almost immediately after this
toward one of these two phrase-rhythmic configurations in spe- downbeat. Consequently, as the onsets of the phrases (which
cific areas of musical form. He argues that tight-knit opening are essentially of the same length as the four-bar hypermea-
themes are usually beginning-accented, whereas a' majority of sures) lag slightly behind the hypermetrical grid, hypermetri-
closing themes (i.e., not only themes within the closing zone in cal downbeats have shifted from the beginnings of phrases to
sonata form, but also those of an appended nature occurring their ends.
after a strong cadential close of a formal section) are end-ac- In this scenario, it is clear that the perception of the closing
cented. Example 2 reproduces Temperley's illustration of the theme as end-accented depends on hypermetrical context; ifwe
typical formula for creating an end-accented closing theme. detach the theme from the previous cadence and consider it in
Prior to the appearance of the theme in m. 50, an emphatic ca- isolation, we could easily hear it as beginning-accented by sim-
dential close (that of the previous formal section) articulates a ply shifting the hypermetrical grid one bar forward. However,
strong hyperbeat and establishes a hypermetrical context that
continues through the next section.l" The four-measure hearing of the cadence as hypermetrically strong. The continuation of the
hypermeter from the cadence should also not be taken for granted, since it
16 As Temperley points out, the hypermetrical downbeat at the cadential close is possible that the new theme restarts the hypermeter, as it does in the
is at times perceived through a metrical reinterpretation. In other words, beginnings of many secondary-theme zones. However, Temperley correctly
although a strict counting of hypermeasures in the passage prior to the points out that the textural change at the cadential discharge in Example 2
cadence may not lead to a strong hyperbeat at the cadence, the tonal force and its continuation in the closing theme encourage one to hear a consis-
of the cadence and the resulting structural accent often promote the tent hypermeter from the beginning of the closing zone.
54 MUSIC THEORY SPECTRUM 34 (2012)
as Temperley states,"? the schema illustrated in Example 2 downbeat at the cadence. Finally, Schema (d) is conceivable but,
(which he calls the "closing-theme schema") occurs frequently I assume, rare, as sustaining an entire hypermeasure in the ab-
in the closing zone of sonata form to facilitate the perception of sence of a new melodic phrase creates an awkward rhythmic and
its thematic content as end-accented; conversely, the creation of rhetorical gap that would likely obstruct musical flow.21
end-accented themes, which relies on the hypermetrical context The description above delineates a three-level hierarchy of
of the closing-theme schema, is frequently associated with clos- phrase-rhythmic scenarios at the beginning of the closing zone:
ing rhetoric. From these observations, Temperley proceeds to Schemas (a) and (b) at the first level, (c) at the second, and (d) at
suggest that the closing-theme schema in Example 2 may be the lowest. For the analyst, this hierarchy provides not only an
regarded as a norm for end-accented themes in the closing abstract conformational system, but also a meaningful norm con-
zone, and that deviations from this schema may be regarded as ducive to a generative investigation into unusual compositional
fundamental departures from the norrn.l'' choices. As Hepokoski and Darcy explain, "[N]ot to activate a
Temperley's assertion and its corollaries may easily be trans- first-level-default option ... would require a more fully conscious
lated into statements from Sonata Theory. At first-level default, decision-the striving for an effect different from that provided
materials in the main body of primary- and secondary-theme by the usual choice."22 In other words, the presence of a low-level
zones are beginning-accented, while the preponderance of end- default should prompt the analyst to pay close attention to the
17 Temperley (2003, 135). 21 I am not aware of any example of Schema (d) in the literature.
18 Ibid. (139). 22 Hepokoski and Darcy (2006, 10).
19 Hepokoski and Darcy (2006) discuss the issue of elision between the sec- 23 According to Lerdahl and Jackendoff's metrical preference rules (MPRs)
ondary-theme and closing zones in Chapter 8. Their discussion, however, (1983, 75-96), both the tonic chords (which are structurally important to
does not address the hypermetrical implications of such situations. I will the harmonic prolongation of this passage) and the onset of subphrases in
investigate these issues in the present essay. mm. 121 and 125 provide strong incentive to hear (hyper)metrical accents.
20 Although a lower-level default here, Schema (c) appears frequently at the See their discussion of MPR 2 (strong beat early) and MPR 9 (time-span
juncture between the transition and secondary-theme zones. I will elabo- interaction).
rate on this point. 24 Rothstein (1989, 58).
PHRASE RHYTHM AS FORM IN CLASSICAL INSTRUMENTAL MUSIC 55
CADENCE
End-accented
(a) I
Hyperbeats: 2 3 4 1 etc.
Beginning-accented I
I
(b) Hyperbeats: 2 3 4 1 etc.
Example from the literature leethoven, Symphony No.2, movement 1, mm, 112-31
"Beginning-accented
Hepokoski and Darcy describe this process as deferring the "es- which secondary-theme rhetoric yields to closing rhetoric-is
sential expositional closure" (henceforth EEC)-the point at fundamentally unproblematic, despite the deferrals discussed
which the exposition is considered structurally closed. Normally, above. In many instances, however, the location of this struc-
this point of closure of the exposition is provided by the first tural point may be more contentious. What criteria should one
perfect authentic cadence (henceforth PAC) in the secondary- employ to pinpoint the moment of definitive transition from
theme zone. Yet, as in the case of Mozart's Symphony No. 38, the secondary-theme zone into the closing zone? Indeed, this
various melodic, harmonic, and rhythmic factors may cause the issue has caused much disagreement among scholars.V For
EEC to be deferred to a later PAC.25 The accrued anxiety from Hepokoski and Darcy, EEC deferrals are invoked in two sig-
repeated failures of the secondary-theme zone to conclude nificant situations: either secondary-theme materials refuse to
through an EEC is thus finally alleviated (and recompensed) subside after a PAC, or the cadential effect is attenuated be-
when Mozart's unusual rhythmic choice promotes an elevated cause the cadence is either temporally premature (i.e., it hap-
sense of conclusion in m. 129.26 pens substantially earlier than expected within the
In the Mozart example, it is tacitly assumed that the precise secondary-theme zone) or phenomenally too weak (e.g., it is
location of EEC-and therefore the exact point at
27 Caplin (2009b) argues, from the viewpoint of formal functions, that one
25 For further details regarding EEC deferrals, see Hepokoski and Darcy's should not automatically assume any emphatic cadence to be the divider
discussion (2006, Chapter 8). between secondary-theme and closing zones. For him, the harmonic and
26 My interpretation of Mozart's second-level default assumes the validity of rhetorical profiles of thematic materials are the key factors in pinpointing
hearingm. 129 as a point of structural close, and that the following mea- the divider between the two zones. Generally, he reserves the closing func-
sures therefore entail a change in formal-rhetorical space. For those who tion for a short passage at the very end of the exposition that is harmoni-
are skeptical about this assumption, I offer another argument in a later sec- cally static and prolongs the attained secondary key. Rothstein (1989), on
tion of this essay-one based on phrase-rhythmic characteristics of the the other hand, follows Heinrich Koch ([1793] 1983) in favoring the first
three cadences in the "post-transitional" materials-to complement Hepo- PAC in the secondary-theme zone as the marker of its conclusion. See also
koski and Darcy's observations regarding the failed cadences that precede Hepokoski and Darcy (2006, Chapter 7) for a discussion of historical views
m.129. on this issue.
MUSIC THEORY SPECTRUM 34 (2012)
Antecedent
I
97
n
2 3
~
p
flDl 1 1
p
1
(or 1
H ypermeter restarts
5 6 1
---------1-------+---------+-3 4 1)
n-n 1TI1i'1TI
EXAMPLE 4. [Continued]
articulated with a sudden drop in the dynamic level). But how would seem the inevitable choice for EEC. Hepokoski and
should one define the threshold at which melodic materials are Darcy grant this interpretation, albeit with a significant quali-
deemed too similar or cadences too early or too weak? Questions fier: "With its preceding drop to piano in m. 68-which requires
ofwhether the EEC is delayed, and if so, by how much, may be a crescendo recovery-and the evaporation of the literal domi-
open to debate; different analysts may determine different loca- nant from the bass, this PAC at m. 74 is still problematic." They
tions of EEC. thus conclude that the Waldstein presents "a difficult case to
The first movement of Beethoven's Piano Sonata in C Major, evaluate in the absence of additional, supporting evidence, such
Ope 53 ("Waldstein"), part of which is shown in Example 6, il- as a P-based C theme." To call m. 74 the EEC is thus for them
lustrates this conundrum. Hepokoski and Darcy reject the first a reluctant compromise at best. 3o
viable candidate for EEC in m. 42 (not shown in the example) Considerations of phrase rhythm, however, provide strong
because of the varied restatement of the secondary theme im- evidence for the legitimacy of m. 74 as the EEC. As Example
mediately following. They also reject the next potential choice, 6 shows, hypermetrical profile and phrase delineation around
the PAC at the end of the restatement in m. 50, on the grounds m. 74 contain all the ingredients for creating the end-accented
that the passage following this cadence "grows directly out of closing schema we have previously discussed. In a manner
[the restatement's] triplet figuration."28 In both instances, post- similar to mm. 121-29 in Example 4, grouping and harmonic
cadential gestures are considered too similar to pre-cadential factors in mm. 50-74 promote a lucid four-bar hypermeter in
ones to deliver the closing rhetoric. After the triplet figure at m. which downbeats consistently coincide with the beginning of
50, Beethoven launches "a long excursus" that builds up tension subgroups and/or prominent harmonic changes. As the ca-
through rhythmic acceleration and a Stillstand auf der dence in m. 74 falls on a downbeat of the ongoing hypermeter,
Penultima;29 the PAC in m. 74 that resolves all these tensions
Schema (a)
131
-= an
3 4 3 4 1
4 1
l111JJJJ
EXAMPLE 5. Recompositions ofMozart, Symphony No. 38 in D Major, K 504, L end ofexposition
it provides the first ingredient for the end-accented closing phrase-rhythmic schema is a co-first-level default in the clos-
schema; indeed, without evidence of hypermetrical readjust- ing zone. The marked shift from the preponderance of begin-
ment in m. 75, the accented cadence establishes the hyper- ning-accented (sub)phrases to end-accented subphrases at m.
metrical context of the following passage. Subsequent melodic 75 may thus be construed as a marker for the closing zone
materials, by entering slightly behind the hypermetrical through the activation of one of the first-level defaults seen in
grid, cause the hypermetrical accents to fall at the end of Example 3. The choice of the end-accented schema over its
phrases. As I proposed in my discussion of Example 3(a), this beginning-accented counterpart at the first level further
PHRASE RHYTHM AS FORM IN CLASSICAL INSTRUMENTAL MUSIC 59
Beginning-accented groups
47
P decresc.
3 4 1
fp
2 3
...- -----.
fp
1 r
J ",....- --........
I
rr--
EXAMPLE 6. [Continued]
PHRASE RHYTHM AS FORM IN CLASSICAL INSTRUMENTAL MUSIC 6r
distinguishes m. 74 from the cadence in m. 50, which was end-accented schema, which deforms the expected rhythmic
elided with the following phrase. All in all, a radical change in shape of the formal space in which it occurs.
phrase rhythm introduced at m. 74 strongly suggests a defini- Such is the case in the first movement of Beethoven's Piano
tive progression into the closing function. In this example, we Sonata in A Major, Ope 2, No.2. As shown in Example 7, the
see that the awareness of phrase-rhythmic norms within the transition arrives at the dominant of the secondary key in m. 42,
closing zone helps us make an informed interpretation regard- prior to the entrance of the secondary theme. Typically, this har-
ing the functions of cadences and their relationship to impor- monic goal would then be prolonged through a "dominant lock,"
tant points of formal closure. which often concludes with several "hammer strokes" before the
medial caesura. Instead of this common procedure, however,
DEFORMATION: END-ACCENTED THEMES IN ZONES OTHER Beethoven employs what Hepokoski and Darcy call a "caesura-
THAN CLOSING ZONES fill"-here, a connecting passage leading from the end of the
transition zone in m. 46 to the secondary theme zone in m. 58 via
At times, a composer may reject all default choices and opt a stepwise descending motion in the bass. 35 This caesura-fill
for something radically unusual. In Sonata Theory, Hepokoski serves several significant tonal and metrical purposes above and
and Darcy call such a situation a deformation. As they explain, a beyond its fundamental role as a rhetorical link. Tonally, the pas-
Caesura fill
VofE major
Leading to E minor?
47 ra II.
A ~ i ,...--~ ~
-.
- -. .,.
-- - -
\= ..............
111....1 I"""
...
-if"':' ~-,J-J" -,J-' ~-,J-y -,J-' ~-,J-
~ ~ tt.;P
2 3 4
) 2
. .
3 4 1 2 3
.....---..
4 1~
.. j) J J -J J
~
~ ~- I. ".....--......
, :
~~ - 1---- _ ....-"1 I I
espressivi
1 3
E major: 16
PAC
Closing zone (beginning accented)
passage from mm. 58-83 is a struggle for resolution both tonal on the passage's thematic character, which overrides concerns
and metrical. The end-accented groups, situated within a ner- about the immobility of the dominant (of the new key).
vous search for any single controlling key (let alone the expected The caesura-:fillhearing ofthe theme becomes much more pal-
one), sound as if they are trapped in a loop hopelessly lagging pable, however, when one incorporates phrase-rhythmic consider-
behind the hypermeter. As the sequence steps up its effort to ations. To begin with, observe that the medial caesura in m. 45
search for its tonal resolution (through harmonic acceleration in falls on a downbeat of the ongoing quadruple hypermeter (which
mm. 68f£), the end-accented rhythm of the passage also begins is projected by the grouping structure of the theme in mm. 37-45
to break down. The sequential repetition of the two-measure and prominent harmonic changes at mm. 41 and 45)-a telltale
groups in mm. 68-76 creates the effect of entrapment within sign that an end-accented passage may follow. As Example 8
the last two beats of the hypermeasure beginning in m. 66, thus shows, it turns out that harmonic and textural continuation be-
effectively dissolving the integrity of the end-accented four- yond the medial caesura preserves the ongoing hypermeter and
measure groups and four-bar hypermeasures. End-accented causes the following four-measure groups in mm. 45-57 to con-
rhythm further disintegrates when the sequence reaches a form to the end-accented paradigm in Example 3(a). This end-
standstill on viio~ ofE in mm. 74-83. Is this inherently ambigu- accented rhythm, which we have seen to be normatively associated
ous harmony going to restore the expected key of the major with the closing function, further enhances the post-cadential
Me End-accented groups
,...-----------------------------------,11
-1 p dolce
~~
~. l
2 3 4 2
Beginning-
accented
Darcy construct a taxonomy of transition types-as listed in extended to accommodate the phrase overlap so that no metrical
Example 10-based on thematic and rhetorical elements. reinterpretation is needed at the onset of the transition. With an
Integrating ideas from Caplin's work on formal functions, these elision only in the phrasal, but not the metrical, domain, the im-
transition types demonstrate diverse thematic relationships with pact of the transition is usually less forceful than in Example
the primary-theme zone and various degrees of developmental 11(a). Finally, the transition in Example 11(c) does not overlap
impulses within the transitional materials. In the following dis- the primary zone phrasally or hypermetrically. The launching of
cussion, I will complement these thematic-rhetorical types with the transition zone delivers the least impact among the three. In
three other transition categories based on phrase-rhythmic con- the following discussion, I will call these three rhythmic situa-
siderations. As we shall see, these rhythmic types interact with tions PO+MR (phrase overlap and metrical reinterpretation), PO
Hepokoski and Darcy's thematic-rhetorical types to form sub- (phrase overlap only), and DISJUNCT, respectively. By and large,
categories illuminating both normative and unusual procedures these three categories exhaust phrase-rhythmic situations be-
at the beginning of the transition zone. tween the primary-theme and transition zones.t"
Example 11 presents three phrase-rhythmic scenarios at the A substantial number of Hepokoski and Darcy's thematic-
juncture between the primary-theme and transition zones. In rhetorical types show strong propensities for particular phrase-
Example 11(a), the two zones are elided by a phrase overlap. At rhythmic scenarios outlined above. To begin with, Types 4, 7,
the same time, this thematic elision also causes the last beat ofthe and 8 in Example 10 are intrinsically associated with the
hypermeter in the primary-theme zone to be eclipsed by the first DISJUNCT rhythmic type. These transitions begin as the sec-
downbeat of the transition zone, thus resulting in what is usually ond part of a larger formal unit (e.g., the consequent of a period,
called a "metrical reinterpretation.Yl With overlaps in both the continuation of a sentence, or a hybrid) which rarely over-
phrasal and metrical domains, this situation typically launches the laps with the first part or reinterprets the ongoing hypermeter.
transition with great momentum. In Example 11(b), in a manner In fact, a phrase overlap is altogether impossible in Types 7 and
similar to that seen in Example 11(a), the two zones overlap 8 because these transitions begin in the middle of a phrase.:"
phrasally; however, the last phrase of the primary zone is On the other hand, Types 1 and 2 are particularly susceptible to
the PO+MR and PO configurations.t" The explorative and de-
modules; a structure-determining dialogue with the principle oflarge-scale velopmental nature of these two types is often accompanied by
rotation; and the need for a quasi-symmetrical tonal resolution in the last
sonata-space rotation" (344). From this nexus, Hepokoski and Darcy pro-
ceed to identify five sets of compositional choices. The criteria used in con- 44 It would seem that, aside from the three phrase-rhythmic situations out-
structing these types are based on the presence (or absence) of thematic or lined here, another (theoretical) possibility is to apply the end-accented
rhetorical functions. Types 1 and 2 are characterized by the absence of such paradigm in Example 3(a) at the beginning of the transition. However, I
major thematic sections as the recapitulation and development; Type 3 have not come across such usage in practice.
contains all thematic units in the so-called "textbook model"; Type 4 has 45 Hepokoski and Darcy (2006, 95) regard these types as exceptions to the
the thematic scheme of a rondo; and, finally,Type 5 mimics the alternation general rule that transitions should begin a new phrase.
between ritornello and solo in concerto movements. 46 As Hepokoski and Darcy (2006, 95, 101) assert, the difference between
43 See Rothstein (1989, 52-56). Types 1 and 2 is only a matter of degree of development impulse.
PHRASE RHYTHM AS FORM IN CLASSICAL INSTRUMENTAL MUSIC
(a)
1 2 3 4/1 2 3 4
(b)
2 3 4 1 2 3 4
(c) ~~
1 2 3 4 1 234
additional unit of fragmentation, which forces the cadential of the sentence, the consequent phrase thus necessitates a met-
function to "spill over" into the ninth measure.t'' In contrast, the rical reinterpretation when its last measure is eclipsed by the
opening theme of the Eighth Symphony in Example 12(b) is hypermetrical downbeat at the beginning of the transition. 51
clearly periodic in its conception. As stated above, periods (es- To be sure, not all Type 1 and Type 2 transitions belong to the
pecially as expository units in the primary-theme zone) are rhythmic types PO and PO+MR. Their appearance as rhythmi-
more resistant to the developmental impulses and expansions cally DISJUNCT, however, is much rarer, often motivated by
seen in Example 12(a); even when expansions do occur, the pre-
mium put on proportional balance between the antecedent and 51 The occurrence of a Type l/PO+MR transition is especially common in
the consequent limits the extent to which the end of the period Haydn's late symphonies when the primary-theme zone is periodic; see the
can be altered to accommodate phrase overlaps and to avoid first movements of Symphonies Nos. 93, 96, 99, 100, 102, 103, and 104. A
metrical shifts.i" In the Eighth Symphony, although the conse- related situation in which Type l/PO+ MR transitions often occur exists
when the primary-theme zone is set in a rounded binary form ending with
quent phrase is expanded through an inconclusive cadence in m.
a period. This situation happens, however, more commonly in last move-
8 and the so-called "one more time" technique (to borrow Janet
ments than in first movements; see the last movements of Haydn's Sym-
Schmalfeldt's term'") in mm. 9-12, the length of the four-mea- phonies Nos. 93, 94, 96, 97, 99, 101, and 102. On the other hand, certain
sure basic unit is preserved to maintain an overall symmetry. conditions of a periodic primary-theme zone seem to facilitateexpansion of the
Without the same flexibility available to the continuation unit consequent phrase at the expense of strict proportional balance. One of these
conditions applies when the period is more extended than the common 8+8
construction (such as when the phrases are longer than eight measures, or the
period consists of more than two phrases--e.g., the so-called "double period").
for definitions of prolongational and cadential progressions.) He then states In this scenario, the consequent phrase is more open to expansion because the
that an "extension of continuation function" is normally implemented by sense of balance is less susceptible to distortion by modifications of phrase
"[ajdditional units of fragmentation and a thorough sequential progres- lengths. Another condition that facilitates expansion exists when the phrases
sion," whereas the cadential function is often expanded by what Schoen- within the period are themselves sentences, in which the continuation unit is
berg (1976, 30) calls a "written-out ritardando." often expanded from an underlying prototype when undergoing development.
48 Other examples of the opening sentential theme "spilling" into an extra In examples of this category,an opening periodic structure may lead to a Type
measure can be found in Beethoven's String Qpartetin D Major, Ope 18, llPO transition if the expansion in the continuation unit of the sentence helps
No.3, IV; his Piano Sonata in C Major, Ope 2, No.3, I; and his Piano eliminate the need for metrical reinterpretation.The primary-theme zone of the
Sonata in E~ Major, Ope 7, I. first movement of Mozart's Symphony No. 39 in B Major, K. 543, illustrates
49 Of course, consequent phrases in periods are sometimes expanded. What I both of these conditions. The period, which consists. of twenty-nine measures,
assert is that, comparatively speaking, sentential forms are more prone to is subdivided into a 14+15 scheme. Substantiallylarger than the traditional 8+8
internal expansions than periodic ones, especially when internal expansions period, the slight expansion in the consequent phrase minimally disturbs the
create (sublphrases of irregular lengths (such as five-, seven-, or nine-mea- balance between the two phrases.At the same time, both phrases are themselves
sures), as in the case of Beethoven's First Symphony discussed here. The sentences: the antecedent is cast in a 4+4+6 format, while in the consequent
difference in the importance of development, and thus also in rhythmic phrase the six-measurecontinuation unit is expanded by one measure to accom-
plasticity (to borrow Frank Samarotto's [1999] term) between sentences modate for the phrase overlapwithout reinterpreting the hypermeter. Interest-
and periods is clearly delineated in Schoenberg's (1967,58) assertion that ingly; whileType llPO transitions are relativelyrare in Mozart's oeuvre,they are
sentences are "a higher form of construction than the period" because of featured in both the first and last movements of this symphony; highlighting the
their intrinsic differences regarding the role of developing variation. influence of Haydn's practice, on top of the "monothematic" treatment of the
50 Schmalfeldt (1992). secondary-theme zone in the last movement.
PHRASE RHYTHM AS FORM IN CLASSICAL INSTRUMENTAL MUSIC
F~-
I'
13 -e-
1\ ~ ~
tJ
I lCs;; er t I
i
p
" - - - - - - - - - --e- p
; - - r- ... ...
~~
... ...
~:
tJ
U ~~~"~~~,, ~ ~y ~~
4' 4' ":'"
U I CJ' CJ'
p P
I 2 I 2 I 2 I 2
l - :- - -
I
:- - -
I
:- -
I
:- -
I
:- - - - I ~-=?-
...........
r-
I
:- -
I
I
-
21
c
tJ !!~~!!~~ ! ~
~
...
TJ
-- T
........-I
I"'" ...
~ !' ~ =J:~T
!'
I <tr
-,j-.
VI 0 !' !'
:
I
rs;
I
I rs;
I
I
2
rs;
I
I
-
-.0... -e-
~
n.
P-
...
2
... -
-.----
.-I.
-:::
:
I
--=========================V
(continuation function extended
,,
by "additional unit of fragmentation") Transition
cadential function (PO)
~
~
l:
27
"
tJ r_ .
:
e 1&----------
if
I I I I
, if .If
.. I
11
'''"' - ....
- - - - :-
I
~ :-
} tJ V
CJ'
=J:~T ! ! ~ If ! ! T..:-J' V i r ~ ~:
), I 2
~
I 2
;.
I
: P- :- - :- : P-
I
V -====== .If
EXAMPLE 12(A). Beethoven, Symphony No.1 in C Major, Ope 21, L mm. 13-33
highly specificconditions within the primary-theme zone, particu- Hepokoski and Darcy's Type 3 transition exhibits phrase-
larly when the primary-theme zone is tonally "over-determined." rhythmic tendencies virtually opposite to those ofTypes 1 and
Hepokoski and Darcy use this term to describe primary-theme 2. Most Type 3 transitions begin by restating the opening mea-
zones that "consist of multiple modules, several ofwhich end with sures of a movement after the primary-theme zone's definitive
a I:PAC or IAC."52The later modules of an over-determined pri- close on a PAC. The clarity of this intent usually requires that
mary-theme zone tend to decrease in dynamics, usually coming to no overlap occur between the end of the primary-theme and the
full repose at the delayed conclusion. Since in most circumstances beginning of the transition. We could thus infer from this
the induction of a phrase overlap involves a preceding build-up of observation that the DISJUNCT rhythmic type is a first-level
momentum/tension and anticipation of resolution, the protracted default in Type 3 transitions.t" However, under certain condi-
harmonic stasis and deterioration in energy at the end of a tonally tions of the primary-theme zone, a Type 3 transition may begin
over-determined primary-theme zone render the rhythmic types with the PO, or even the PO+MR, scenario. One such condi-
PO and PO+MR largely unsustainable. In other words, a tonally tion obtains when the primary theme consists of two parts
over-determined primary-theme zone projects a trajectory that is having a loud-soft sequence. Example 13(a), which shows the
simply not conducive to the rhythmic subtypes typically associated opening of Mozart's Haffner Symphony, illustrates this situa-
with the independent transition.f' tion. After the more subdued second half of the opening theme
in mm. 6-13, a Type 3 transition beginning in m. 13 overlaps
52 Hepokoski and Darcy (2006,74).
53 For examples of DISJUNCT independent transitions following over-
determined primary-theme zones, see the first and last movements of 54 Familiar examples of this situation include the first movements of Mozart's
Mozart's Piano Sonata in F Major, K. 332. Piano Sonata in B~ Major, K. 333, and his String Quartet in G Major, K. 387.
68 MUSIC THEORY SPECTRUM 34 (2012)
Antecedent Consequent
II
pdolce
pdolce
2
1 2
Transition
(lAC) Consequent (repeated) (PO+MR)
I I
the end of the primary-theme zone to regain the energy level of duple hypermeter than in the previous example; m. 6, there-
the opening measure. Although the hypermeter of the opening fore, is projected as hypermetrically weak. Nevertheless, the
forte fanfare in mm. 1-5 is somewhat ambiguous (as indicated beginning of the sentence at m. 6 resets the hypermeter be-
by the two competing interpretations marked "Hearing N.' and cause of the 1+ 1+2 sentential structure in mm.6-9.
"Hearing B" in the example), the grouping structure of the Rothstein has discussed this particular situation using what
piano sentence in mm. 6-13 forcefully positions the hypermetri- he calls the "rule of congruence," which is an adaptation of
cal strong beats at the even-numbered measures (and the weak Lerdahl and Jackendoff's MPR 9 (time-span interaction). 56
ones at the odd-numbered measures, as shown in Hearing B), He explains that,
thus necessitating a metrical reinterpretation at the beginning [I]f a four-measure phrase is subdivided, melodically, either as
of the transition in m. 13. 55 2+2 or as 1+1+2, a listener should infer a metrical pattern in
Example 13(b), taken from the opening of Mozart's Piano which the first downbeat in each two-measure group is also the
Sonata in A Minor, K. 310, demonstrates a schema very similar downbeat of a two-bar hypermeasure. The first downbeat in the
to Example 13(a). The loud-soft sequence again comprises a first two-measure group will be the downbeat of a four-bar
five-measure gesture followed by a sentential construction. The hypermeasure. In plain English, a four-measure phrase is, gener-
ally, also a four-bar hypermeasure, especially if it can be subdi-
opening five-measure gesture articulates a more unequivocal
vided into two equal parts (as is true, for example, of phrases
subdivided 1+1+2).57
55 Hearing B of the piano sentence in mm. 6-13, which I argue to be the
preferred hypermetrical interpretation of the passage, may seem discordant
From this perspective, the primary-theme zone concludes at m.
with the durational/textural accents in mm. 7 and 9; nevertheless, the 9 on a hypermetrical weak beat, which is then reinterpreted as
placement of hypermetrical accents at the beginning of the subgroups at the first strong hyperbeat of an overlapping Type 3 transition.I''
mm. 6, 8, and 10 is again strongly motivated by Lerdahl and ]ackendoff's
strong-beat-early rule (especially because the rule often serves as the dom- 56 Lerdahl and]ackendoff(1983, 90).
inating factor at hypermetricallevels, as mentioned earlier in this essay). A 57 Rothstein (1995, 173).
comparable situation can be found in Rothstein's analysis of the opening of 58 The reading of a phrase overlap between the two zones here is somewhat
Beethoven's Piano Sonata in G Minor, Ope 49, No.1 (1989,59). Although more complicated than that in K. 385. The contentious point is that some
mm. 2 and 4 are dynamically and durationally accented, Rothstein posi- may read the primary-theme zone in the sonata as ending on V in m. 8.
tions the strong hyperbeats at mm. 1 and 3 so that the subgroups in mm. This reading would lead to the understanding of the transition as
1-2 and 3-4 are beginning-accented. Hepokoski and Darcy's Type 4. My interpretation of the passage follows
PHRASE RHYTHM AS FORM IN CLASSICAL INSTRUMENTAL MUSIC
~ subgroup onsets
I I I
Allegro con spirito
2 2 1 2
2 (1 2) 1
Transition
PO (+MR)
I
\~
~
--
f&4 f&4
110-
:
- ~
:- -
r r
-/-
I-~ e t"'
I. I. e t"'
~
f. ~
-#-
f· f
2 1 2 1 2 1
1 2 1 2 1 2/1
.- 1 1
I .... -0:-,.....-......
~
-
.....
.1-:
.. , - ,
~
r-
-&-
:- :- ~
:- :- -- ~
!
In both Examples 13(a) and (b), the PO and PO+MR rhythmic milieu. This intertextual orientation provides the mechanism
types serve to revitalize momentum and connect seamlessly to through which Sonata Theory is able both to embrace the mer-
the forceful restatement of the opening idea. In these instances, its of the conformational approach (in its comparison of the
the DISJUNCT scenario, which normally appears in Type 3 subject piece to a carefully reconstructed norm) and address
transitions, would have caused the primary-theme zone to end generative concerns (in its revelation of the particulars of a piece
on a low dynamic level, resulting in an awkwardly muted con- as a result of the comparison). Further, given that recent studies
clusion of a primary-theme zone which began with zest, and an on intertextuality in Western music have provided invaluable
incongruous, forced return to the opening fanfare at the begin- insights into the construction and perception of musical mean-
ning of the transition. ings, the profits gained by a dialogic formal theory on issues of
criticism and hermeneutics are unquestionable.P''
PHRASE RHYTHM AS BALANCE BETWEEN INTERTEXT A potential problem with the intertextual approach to form,
AND CONTEXT however, is that it tends to undermine attention to the teleology
of decisions within the composition under consideration.
Ultimately, all components of Sonata Theory-stylistic Whereas the particulars of a piece are evaluated in light of their
norms, levels of defaults, deformation, formal types, etc.-work relationship with the normative background, their contextual
together for the purpose of "uncovering and interpreting the rationale and consequences are sometimes inadequately ex-
dialogue of an individual piece with the background set of plored. Thus, a highly unusual element is primarily appreciated
norrns.P" The dialogic orientation of Sonata Theory thus pro- as an idiosyncrasy, possibly entailing certain hermeneutical in-
vides an intertextual perspective of musical form-one that sights, whereas the organic connection. between the eccentric
links each compositional decision in a given piece to all its choice and its local and global contexts within the piece often
counterparts outside the text, yet recognizes that those compo- seems only a secondary concern.P!
sitional decisions occur within a specific historical and generic
60 For recent representative studies in musical intertextuality, see Korsyn
Hepokoski and Darcy's reading of the primary-theme zone as ending with (2001) and Klein (2005).
a PAC in m. 9, thus making the transition an instance ofType 3. Primarily, 61 With their intertextual approach, which often de-emphasizes the functions
I believe that the continuous motion of the dotted-eighth and sixteenth of events within their contextual surroundings, Hepokoski and Darcy al-
rhythm in m. 8 makes it difficult to hear a cadential close on V in that most invariably resort to hermeneutical routes when explaining the pres-
measure. For a slightly different hypermetrical analysis of the passage, see ence of low-level defaults and deformations. For examples, see Hepokoski's
Wick (1991,88). analyses of Mozart's Overture to Idomeneo and Beethoven's Overture to Die
59 Hepokoski and Darcy (2006, 11). Ruinen von Athen in Hepokoski (2009a, 75-88).
MUSIC THEORY SPECTRUM 34 (2012)
Allegro maestoso
Transition PO (+ MR)
I I
5
f
2 2/1
The importance of internal logic and organic coherence is Investigations of phrase rhythm thus often help illuminate the
nevertheless indisputable, especially when interpreting the role of contextual logic of marked events vis-it-vis intertextual routes,
low-level defaults and deformations.P To illustrate the signifi- the latter of which help to situate the former within the even
cance of maintaining a sensitive balance between intertextual and broader schemes of stylistic and compositional traits.
contextual orientations, let us revisit the EEC deferral in Mozart's Let us now investigate how phrase-rhythmic analysis, cor-
K. 504 discussed above. Hepokoski and Darcy's main analytical roborated by voice-leading considerations, offers a concrete
insights reside in their interpretation of this passage as an in- explanation of the EEC deferrals in the first movement of
stance of EEC-deferral via retention of secondary-zone charac- Mozart's K. 504. Example 14 shows a voice-leading reduction
teristics and, subsequently, appreciation of Mozart's masterful and phrase-rhythmic analysis of the passage in question.
technique for delaying the appearance of an expected structural Hepokoski and Darcy have argued that this passage derives its
point. While this intertextual statement describes and assesses formal logic primarily through its expansion of an initial pe-
how Mozart ingeniously distinguishes one particular zone of a riod (rnm. 97-112) into a large sentence with an appendix.P"
sonata-form movement from its counterparts, it offers no con- Addressing the formal role of other contextual elements, the
crete explanation for the raison d'itre of the deferral with regard voice-Ieading/phrase-rhythmic analysis in Example 14 adds
to musical elements within the movement. In other words, how another critical dimension to the consideration of the
do the composer's structural (e.g., thematic, tonal, rhythmic, and
metric) and aesthetic (e.g., affective, proportional, rhetorical, and
expressive) concerns within this particular piece lead to the un-
usual choice? Or, conversely, how does that choice influence the music will be helpful. As Almen states, "[W]e can analyze musical units
movement's structure-and aesthetics? These potentially fruitful with respect to the features that are present or absent and also with respect
grounds for investigation remain uncharted. to the chain of bundled features that appear over time. The former
approach is called paradigmatic, and involves a consideration of those
It is my contention that phrase-rhythmic analysis helps strike a
elements that were chosen from among those that might structurally oc-
needed balance between intertextual and contextual orientations cupy the same position.... A paradigmatic analysis would focus on the
when incorporated into the dialogic framework of any theory of connotations resulting from the particular choices that were made ....The
form. Although the manner in which I have hitherto addressed latter approach described above was characterized by Saussure as syntag-
phrase rhythm in formal discussions may be described as paradig- matic and involves a consideration of relations between elements in a se-
matic, phrase-rhythmic analysis-with its focus firmly grounded in quence or a combination.... Looked at from this 'horizontal' perspective,
the linear trajectory of elements within the temporal domain- the distinctive features that characterize a signifying unit are grouped into
a 'bundle of features' that are themselves part of a temporal chain of fea-
is nevertheless first and foremost a syntagmatic enterprise.v'
tures." The dichotomy between paradigmatic and syntagmatic is parallel to
that between intertextual (dialogic) and contextual, as I establish in this
62 Numerous studies have shown the importance of examining the musical section. The intertextual orientation looks- at the same formal position
connections between a highly marked idiosyncrasy and its surrounding ele- across multiple works and considers the "connotations" of particular choices
ments. See Smith (2000) for a representative study. (which, as I have shown, are often translated into certain hermeneutical
63 The linguistic categories of paradigmatic and syntagmatic have been made implications). The contextual orientation, on the other hand, examines
well known to music theorists by recent works on music semiotics. While "bundles of features" on the horizontal axis within one composition for
a detailed explanation of these terms is therefore unnecessary here, Byron syntactical connections.
Almen's (2005, 46) summary of the two categories as they are applied to 64 Hepokoski and Darcy (2006, 152).
PHRASE RHYTHM AS FORM IN CLASSICAL INSTRUMENTAL MUSIC
/\
3 (ill
97
r_ . . -> I ~h<1 .....I I ~
. -. -- =---1l .l ~ I
--;.~
~)~: J .J -
~
. -- ..
'" ,lj,
- -- -
@)
:.
-
•
I6
fl
• III --.. • III
- r,------ III
:~
6 6 6 6 6 6 6 6 6 -----
1 2 3 4 1 2 3 4 1 2 3 4 1
•• //If
---r-.. ,--~
. III
.--7"
. r r-, .:-1--.
. ~
,..
L I "Y-J L
v
PO+MR PO
I I I I
/\ /\ /\ /\
2? I? I?) S
110~ ----J' ~h
'" ,lj, <,
.1.-
DISJUNCT
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - " r---
/\
\~ ,--
~
1-
)
~ •• III
6
'--------
6
i.
I .--'I~
6/
6
:
4
. }
J.. --
.........
2
--i-V
3 4
r-
---
I
1
i
1!
i
ii 6 V
EXAMPLE 14. Mozart, Symphony No. 38 in D Major, K 504, L mm. 97-129, voice-leading reduction
passage's large-scale coherence. To begin with, the analysis Not only does it slip into A
the minor mode (thus transforming
shows that the o~IaIJ trajectory toward m. 129 as the second- the local Kopfton into ~~' ~ut~he end of the consequent also
ary-theme zone's definitive conclusion is orchestrated by the fails to materialize the
A A
3 - A2-1 descent expectedA
toA close
A
the
large-scale registral transfer that prolongs the Kopfton and de- period. Neither the 3-2-1 in m. 110 nor the ~3-2-1 in the
lays the descent of the Urlinie.6s As shown in the voice-lead- first flute in mm. 110-12 succeeds in providing such closure.
ing reduction, a large-scale arpeggiation begins from the On the one hand, A A
theA former is analogous to m. 102, which
Kopfton C~S in m. 101, traversing E s in m. 113, As in m. 123, articulates the 3 - 2 -1 descent as a motion to an inner voice
and reaching C~6 in m. 125. The Kopfton c~s in m. 101 is also lacking proper harmonic support to close the period; on the
prepared by a local linear ascent from A 4 in m. 97. Intriguingly, other hand, the latter is detached from the main melodic
each of these tones in the deep middleground arpeggiation is strand by its higher register and is therefore heard more logi-
elaborated with local descents that define the tonal profiles of cally as a lead -in to the next phrase.
individual phrasesA and
A
thematic materials in the secondary- The following phrase, appearing in the form of a sentence in
theme zone. The 3-2 descent and interruption in the anteced- mm. 112-21, takes the overallarpeg§.iation to E s. From this pitch,
entA phrase
A A
in mm. 97-104 set up the expectation of a the phrase completes with a 5-4-3-~-1 descent to propose a
3 - 2 -1 descent and an authentic cadence in the consequent. definitive close of the secondary-theme zone. Yet the 5-Zug is
Mozart's realization of the consequent, however, is unusual. somewhat concealed by the cover tone As, which originates from
the failed attempt of the first flute to structurally close the sec-
65 By Kopfton and Urlinie I am referring to the head tone and the melodic ondary-theme zone in m. 112. When Mozart reinvigorates tran-
descent of the secondary theme area, not those of the entire movement. sitional materials in mm. 121-29-the last module of the
72 MUSIC THEORY SPECTRUM 34 (2012)
secondary-theme zone-the large-scale arpeggiation swiftly pro- that the phrase-rhythmic analysisofthe passage captures the pro-
ceeds from E 5 in m. 121 to A 5 in m. 123, and finally reaches C#6 cess by which the ball is finally caught at m. 129. Significantly,
in m. 125.66 From this Kopfton, the secondary-theme zone finally when the ball is caught at the end of a spectacular throw, our fas-
makes its definitive descent to the local tonic in m. 129, accom- cination comes not merely from judging it against the more com-
panied by an unadulterated PAC to provide unequivocal struc- monplace instances; rather, the action of the throw, the trajectory
tural closure. Considering voice leading alone, we see that the caused by the action, and the posture of the receiver required to
deferral of EEC to this particular point culminates a large-scale catch the ball all captivate us by their singular connection with one
melodic endeavor that provides the driving force for the entire another within one enthralling play.
secondary-theme zone. The impact of this melodic process is so A
strong that even the closing zone-which normatively prolongs 1 A FINAL EXAMPLE
attained at the EEC-momentarily suggests further ascent to E 6
in mm. 131-33. The illusion is short lived, however, since the to~ To conclude, I will offer a final example that brings various
voice succumbs to the gravity of the tonic in a hasty '5 -4-1- ~-1 ideas explored in this essay into a contextually informed inter-
descent immediately before the codetta. pretation of a highly unusual event. Example 15 shows the coda
More significant for our discussion is the way in which of the first movement of Mozart's Piano Sonata in D Major, K.
2 3 4/1 2 3
tJ
I""
P
,..
r
-..- --, ..-
- --...
- U' .0\.
4/1 2 3 4
~ -#- ; -;,- -;,- -;,- ~ -it-
1 ; ...
A
;. ___ --; .--1*-
~
....I
....I
= .....-.or
19
i
.IJ, .-
. -~ ~ _. ... - .. -
etr ~ ~.
- .
etr
J
.... -
,.,
....
.-
~ -- ..,
==
IIII-
....
......I
-- .......
1111"" I""
--=
-
I"" ~ ,..
~~ ~
4
~,: I}:
1
~ r~-,.. ---.....--1*- rt- «-,
2
----.
1 -- -----......
..... l_ r: ,..
-= -
EEC Closing (CO)
C 1 (end accented)
p f
2 3
Allegro
I
2 2
steer clear ofloop-like constructions and remains mostly faithful characteristics of the loop-thus vividly illustrates Almen's
to quadruple hypermererr? In the recapitulation (beginning at comic narrative archetype. This interpretation of the movement
m. 79), the rather literal transposition of materials from the sec- as comic in its overall narrative helps direct the analyst's atten-
ondary-theme and closing zones in the exposition implies that tion to an abundance of surface compositional details that con-
the distinctive phrase-rhythmic features of the opening loop- tribute to the playful character of the movement, such as the
metrical reinterpretations caused by phrase overlaps and the false recapitulation of the secondary theme in m. 58, the play
resulting triple hypermeter-have been banished from the entire with different registers in the secondary-theme zone in the reca-
movement between the opening statement and its delayed pitulation, and, most wittily, the deceptive move in m. 110,
return. In this observation, we see that the late recapitulation of where the closing material falls glaringly short of a triumphant
the primary theme is much more than a simple device to play conclusion parallel to that in m. 37.
with listeners' expectations: it realizes a highly distinctive narra- In this and all the previous examples, we have seen that
tive in which a truly "transgressive" element-to use Byron phrase rhythm provides promising ground for the construction
Almen's term73-is proposed right from the outset, but becomes of stylistic norms, establishment of types, differentiation be-
curiously suppressed until it is capriciously resuscitated under the tween levels of compositional choices, and creation of deforma-
most unexpected circumstances. The resulting narrative-one tions through stretching or even rejecting normative procedures.
that focuses on the fate of the unusual phrase-rhythmic Analytical decisions on these matters supply a rich source of
inspiration for further interpretations of structure, composi-
72 At only two locations is quadruple hypermeter temporarily problematized: tional practice, style, and expression. Phrase-rhythmic analysis
mm. 52-57 and mm. 66-74. In the former, repetition in mm. 56-57 results thus complements traditional theories of form-those based on
in an expansion from four measures into six. In the latter, hypermeter grad- themes, tonal structures, and formal and rhetorical functions-
ually breaks down as the Fortspinnung passage undermines larger metrical to give us an even more informative picture of form in music.
patterns than the notated measure, even shifting the surface meter in mm.
Heeding the implications of Cone's aphorism by rehabilitating
69-73 by a half-measure until a one-and-one-half-measure unit beginning
in m. 73.5 restores the notated meter and quadruple hypermeter at m. 75.
the marginalized elements of rhythm, meter, hypermeter, and
In both cases, deviations from quadruple hypermeter provide no explicit their interaction with phrase structure, I hope that this essay
reference to the opening loop and triple hypermeter. will help steer the study of musical form in a productive direc-
73 Almen (2008, 7). tion embracing all relevant elements and parameters.
MUSIC THEORY SPECTRUM 34 (2012)
WORKS CITED Reicha, Anton. 1832. Traite de Melodic: Abstraction faite de ses
Rapports avec I'Harmonie, suivi d'un Supplement sur l'Art
Alrnen, Byron. 2008. A Theory of Musical Narrative. d'Accompagner la Melodic par I'Harmonie, lorsque la Premiere
Bloomington: Indiana University Press. est Predominante Ie tout Appuye sur les Meilleurs Modeles
Bonds, Mark Evan. 1991. Wordless Rhetoric: Musical Form and Melodiques. 2 vols. Paris: A. Farrenc.
the Metaphor ofthe Oration. Cambridge [MA]: Harvard Riemann, Hugo. 1903. System der musikalischen Rhythmik und
University Press. Metrik. Leipzig: Breitkopf und Hartel.
Burnham, Scott. 2002. "Form." In The Cambridge History of Rosen, Charles. 1980. Sonata Forms. New York: W. W. Norton.
Western Music Theory. Ed. Thomas Christensen. 880-906. Rothstein, William. 1989. Phrase Rhythm in Tonal Music. New
Cambridge: Cambridge University Press. York: Schirmer Books.
Caplin, William E. 1998. Classical Form: A Theory ofFormal - - - . 1995. "Beethoven with and without 'Kunstgepriing':
Functions for the Instrumental Music ofHaydn, Mozart, and Metrical Ambiguity Reconsidered." In Beethoven Forum 4:
Beethoven. New York: Oxford University Press. 165-93.
- - - . 2009a. "What Are Formal Functions?" In Musical Samarotto, Frank. 1999. "A Theory of Temporal Plasticity in
Form, Forms and Formenlehre: Three MethodologicalReflections. Tonal Music: An Extension of the Schenkerian Approach to
Music Theory Spectrum, Vol. 34, Issue 1, pp. 51-77, ISSN 0195-6167, elec-
tronic ISSN 1533-8339. © 2012 by The Society for Music Theory. All
rights reserved. Please direct all requests for permission to photocopy or
reproduce article content through the University of California Press's
Rights and Permissions website, at http://www.ucpressjournals.com/
reprintinfo.asp. DOl: 10.1525/mts.2012.34.1.51