Penderecki

MURPHY.
FRA Page 184 Thursday, February 8, 2007 1:25 PM
A MODEL OF MELODIC
EXPECTATION FOR SOME NEO-
ROMANTIC MUSIC OF PENDERECKI
SCOTT MURPHY
1 PROVIDES a piano-score reduction of the first ten measures to

E XAMPLE
Pendereckis First Violin Concerto (1977). This music functions as
the orchestral expositions slow introduction: the music that follows this
excerpt introduces a quicker tempo, running sixteenth notes, and a new
pedal point. Scholars generally agree that this concerto and the opera
Paradise Lost (1978) are the two earliest major works from the com-
posers neo-Romantic period, although some perceive a foretaste of
this new compositional direction in the short orchestral work The Awak-
ening of Jacob (1974).1 Several features of this periodsometimes co
existing or interacting with Pendereckis older avant-garde tech-
niquesconspicuously harken back to the common-practice style in gen-
eral and the nineteenth century in particular: tertian harmonies,
Wagnerian triadic progressions, Brucknerian creatio ex nihilo, traditional
PNM451.book Page 185 Wednesday, January 31, 2007 10:57 AM
Melodic Expectation in Penderecki 185
orchestration and instrumental doubling, continual developing variation,

and protracted but not completely unfamiliar formal designs. Some of
these can be heard in the ten measures of Example 1: the relentless F
pedal point supports an interspersed D -major harmony in measures 15
and a shadowy tonicization of F minor, and wisps of meandering melodic
material slowly emerge from this tonal adumbration in measures 610.
EXA MP L E 1: P ENDEREC K I, V IOLIN CONCER TO (NO. 1 ), MEASUR ES 110
One particular facet of the twilight of Romanticism that is found in

Pendereckis music is the adoption of tonal materials such as tertian har-
monies but a neglect of common-practice syntax. Predominant
dominanttonic progressions, expected resolutions of dissonances or
active scale degrees (such as the leading tone), standard cadential formu-
las and the like are infrequent or completely absentalthough more
recent works, such as his Credo (1997) and De profundis (1998), give
these features a larger role to play. The excerpt from the Violin Concerto
offers a modest example of this neglect: each one of the four related
melodic fragments in measures 610 ends unresolved within the implicit
F-minor tonality. The endings on B in particular, although Penderecki
clearly treats them as points of arrival, hang suspended in the weak tonal
field; a tonal hearing that is guided by the F pedal point would expect a B
to resolve perhaps back up to C or to continue a descent to A or A .
Such tonal suspension is not without ample precedent in the music of the
common-practice period; one can find corresponding moments in, for
example, the music of Mozart. But Pendereckis concerto and other
186 Perspectives of New Music
stylistically similar works differ from common practice in that they unre-
mittingly frustrate such tonal expectations during their entire span.
And this continual frustration, both of tonal expectations and perhaps
also of the listener who brings them alone, is arguably part of the musics
aesthetic: the neo-Romantic facade ostensibly promises tonal successions
that are never delivered. Yet tonal expectations are not the only pitch-
class expectations one may bring to this music. This study proposes the
hypothesis that the highly constrained melodic style of Pendereckis neo-
Romantic music avails another set of pitch-class expectations that, while
not derived directly from tonal practice, are analogous to a rudimentary
set of tonal expectations, and thus may be understood as a substitution
for common-practice pitch-class expectations. Part I defines this melodic
style from a pitch-class perspective, focusing on the two characteristics of
pitch-class diversity and interval-class concentration. Part II then outlines
the new expectancy model, and establishes an analogy between this
model and certain models of tonal expectation. Part III applies the model
to analyses of passages of Pendereckis music in this neo-Romantic style,
and Part IV introduces graphical and tabular means of surveying
Pendereckis melodic style and the complexity and richness of the expect-
ancy model that this style engenders.
I. PENDERECKIS MELODIC STYLE(S)
A feature of Pendereckis neo-Romantic period is often referred to, in

some form or other, as the reintroduction of melody.2 However, in the
context of Pendereckis earlier music, this reintroduction may be bet-
ter understood as a whittling down, shearing off several of the com-
posers earlier avant-garde techniques, and resulting in a more focused
and simplified presentation of a single-line pitch succession of which the
low string lines of Example 1 are paradigmatic. First, Pendereckis micro-
tonal practices that dominated his earlier music take a back seat to the
exclusive use of the twelve chromatic pitch classes (henceforth, pcs) dur-
ing this period.3 Second, whereas most individual melodic lines are easily
obscured by many other interwoven, concomitant melodic lines in the
dense, polyphonic textures that frequent Pendereckis earlier music, his
neo-Romantic music serves up a good helping of monophony. Third,
and perhaps most simply, many of Pendereckis single-pitch lines in his
earlier music are very fast, blurring any sense of a discrete pitch succes-
sion, while his neo-Romantic music includes many more moderately
paced melodies.
MURPHY.FRA Page 187 Thursday, February 8, 2007 2:36 PM
EXA MP L E 2: P ENDER ECK I, V IOLIN CONCER TO, MEASUR ES 1122,

CONTR ABASS
Yet Pendereckis neo-Romantic melodies are hardly reversions to com-

mon practice: the whittling down metaphor is suggested because, at
the core, two significant and generalizable features of Pendereckis melo-
dies from his earlier music initially survive the shift into the neo-
Romantic period. The first will be called pitch-class (pc) diversity, and the
second interval-class (ic) concentration.4 Generally defined, a melody is
pc-diverse to the degree that pc repetitions are fewer and farther apart
than would occur by chance. A melody is ic-concentrated to the degree
that the non-zero ordered pc intervals between its adjacent pitches are
distributed more unevenly over the six interval classes than what would
occur by chance. But before fleshing out these concepts with more pre-
cise definitions and examples, the issue of what does and does not consti-
tute a melody in Pendereckis music requires some attention. For the
purposes of this study, a melody is defined as a totally ordered succession
of single chromatic pitches, relying on silence and other conventional
means of segmentation to demarcate its beginning and ending. Further-
more, Penderecki occasionally builds his melodies by starting with a
single note or a handful of successive pitches, restating this kernel several
times, and affixing one or more notes to the end of some or all of the
restatements of the melody that carry over into the next statement.
Often, these restatements are separated by silence, but even those that
are not clearly project their process of accretion, usually by preserving
registral position. Example 2, which provides the bass line for the section
in the Violin Concerto that immediately follows the music of Example 1,
displays one application of this procedure: the four-note repeated
sequence DD EF first accrues to DD EFBB A in measures
156 and then immediately accrues to the full melody of DD EFB
B AA GF . . . . (The horizontal lines below Example 2, which will

be explained shortly, do not concern this accretion.) Therefore, in taking
this practice into account, a melody under scrutiny should be purged of
any interior repetitions, which are either immediate repetitions of a pc, or
immediate repetitions of the same ordered sequence of pcs.5
EXA MP L E 3: P ENDERECK I, V IOLIN CONCER TO, MEASUR E 273
Given this more narrow definition of a melody, pc diversity is relatively

easy to come by in relatively short, randomly generated melodies. For
example, any four-note pc succession without interior repetitions is 73%
likely to be non-duplicative, that is, where all the pcs are different. This
percentage lowers to just under chance (49%) for five-note melodies. But
from there, the odds drop off considerably: 27% for six notes, 13% for
seven, and only 5% for eight. In comparison, when examining
Pendereckis melodic practice as a whole, the maximal lengths of non-
duplicative successions of pcs without interior repetitions are on average
nine notes, demonstrating a deliberate and efficient composing-out of
the composers familiar harmonic clusters. Penderecki scholar Ray
Robinson comes close to this same figure in his assessment of the com-
posers overall melodic practice: from Strophes [1959] to the Violin Con-
certo No. 2 [1995] the linear material developed is based on a freely
atonal scale model. There are times . . . where Penderecki is writing
serial music; there are others when he uses eight or nine units of the
scale as the basis of constructing a freely atonal melodic line.6 The mel-
ody played by the second cellos in measures 67 of the Violin Concerto
(Example 1) offers a fair example of this diversity: no pc is repeated in its
seven notes. The bass line of Example 2 has pc repetitions, but they are
considerably spread out, as demonstrated by the analysis below the
music. Each horizontal line represents a non-duplicative pc succession
within the bass line that begins and ends with the two pcs at the lines
two endpoints. (Note that some of these successions begin with a pc that
ends an earlier succession; to save space in these cases, lines representing
such successions share endpoints.) These thirteen successions constitute
the set of non-duplicative pc successions where no member of the set is
completely embedded within any other: their average length is just under
nine.7 Example 3 provides a similar analysis for a solo line from later in
the concerto: the average length of the five successions is just over nine.
Ideally, a thorough statistical analysis of this repertoire would confirm or
correct this hypothesis; for now, a general familiarity with this music will
need to suffice. This kind of pc diversity that characterizes Pendereckis
overall melodic practice will be called 9-pc diversity: non-duplicative suc-
cessions of either six or twelve notes are more rare than those of nine.
However, it must be emphasized that the number of nine is merely a sug-
gested average, not a necessary minimum nor a limiting maximum, for
the lengths of Pendereckis non-duplicative pitch successions exhibit a
substantial degree of variance around this mean.
Furthermore, this average may vary when one focuses on a particular
piece or other smaller cross-sections of this repertoire. For example, some
of Pendereckis works, particularly from his early music, adopt twelve-
tone serial techniques; therefore, one expects pc diversity to be higher
when limiting ones scope to such pieces. However, in much of his music,
many lines begin as if they might exhaust the aggregate, but then a non-
trivial pc duplication occurs before all twelve pcs have been presented:
examples abound in the Capriccio for Oboe and Strings (1964), Sonata
for Cello and Orchestra (1964), Capriccio for Siegfried Palm (1968),
Second String Quartet (1968), and others. Example 4 reprints a sam-
pling of melodic runs from De natura sonoris No. 1 (1966) that are
highly representative of this time period. Although there are some non-
duplicative successions of twelve notes among these runs, there are many
more non-duplicative successions that fall short of twelve. This type of
writing takes Penderecki farther away from the modernism of Vienna and
Darmstadt and closer to the modernism of New England: Ruggless pc
diversity that falls short of consistent aggregate completion has been dis-
cussed,8 and the same has been noted in Ivess music.9 The image of
broad-brush10 strokes that Adrian Thomas uses to describe
Pendereckis sonorism of the 1960s thus also serves as an appropriate
metaphor for Pendereckis pc diversity: the composers priority to cover
all ofor enough ofthe pitch-class aggregate with one coat of paint
after another comes across clearly, even if some strokes overlap or he
misses a couple of spots.
Along with 9-pc diversity, the other feature that characterizes
Pendereckis melodies from his neo-Romantic period as well as from the
EXA MP L E 4: P ENDERECK I, DE NATUR A SONOR IS NO. 1, R EHEAR SAL 6
earlier music of the 1960s and early 1970s is interval-class (ic) concentra-
tion. One does not need combinatorial analysis to understand that a rela-
tively large sample of randomly generated melodies, with or without
interior repetitions, will have a relatively flat distribution of the six non-
zero unordered pc intervals (i.e., interval classes) between adjacent notes.
The only bias would be a reduction of the number of tritones (ic6) by
one-half, since this interval class only contains one ordered pc interval.
Compared with this flat distribution, a great majority of Pendereckis
melodies noticeably concentrate on two or three interval classes,
although the degree of this concentration and the interval classes that are
concentrated upon vary among pieces and stylistic periods. The semitone
unquestionably serves as the fundamental building block of all of
Pendereckis melodic motions, but the melodic interval classes that
relieve the semitonal motion change from the sonoristic period to the
neo-Romantic period. Before the mid-seventies, most of his melodies
weight the minor third and (somewhat less so) the major second consid-
erably more than the major third, perfect fourth, and (somewhat less so)
the tritone. The melodic runs in Example 4 from De natura sonoris no. 1
display intervallic preferences very typical of the melodies from the 60s
and early 70s. This bias is also reflected in most of Pendereckis serial
music: the twelve-tone rows from Psalms of David (1958), Dimensions of
Time and Silence (1960), and the St. Luke Passion (1966) focus almost
entirely on the small interval classes of 1, 2, and 3.11
With the arrival of Pendereckis neo-Romantic period, however, ic
concentration becomes even more distinctive: his melodies now strongly
gravitate toward the semitone and the tritone interval classes, or what
will be called ic1/6 concentration. Adrian Thomas hears ic1/6 concentra-
tion as a hallmark of this period, beginning with the Violin Concerto:
While retaining some earlier methods, the concerto marks out future
territory in its discursive, narrative structure and in its focus on two inter-
vals, the semitone and tritone. The semitone has been the unmistakable
cornerstone of Pendereckis vocabulary throughout his career, and unre-
solved chains of semitones, commonly separated by a tritone, have been
the most conspicuous aspect of his melodic style since the mid-1970s.12
The excerpts from the Violin Concerto discussed thus far exemplify this
practice: the melodies in Examples 1 and 2 use semitones and tritones
exclusively, and the melody of Example 3 intersperses a single minor sev-
enth. Wolfram Schwinger notes that, by the Te Deum of 1980,
Penderecki was by now expert in the construction of melodies based on
small intervals, particularly the minor second and the augmented fourth
(the tritone).13 Of course, the melodies of Examples 1 and 2, and even
that of Example 3, are textbook examples; Penderecki does not exclu-
sively use ic1/6 concentration in constructing all of his neo-Romantic
melodiesthe minor third and major second still make a fair number of
appearancesjust as all of Pendereckis melodies do not exclusively
present nine pcs before repeating one of them.
EXA MP L E 5: P ENDER ECK I, TH E DE VILS OF LOUDON, FIV E MEASUR ES

BEFOR E R EHEAR SAL 4
Ray Robinson, in pinning down Pendereckis general stylistic traits

that transcend particular style periods, claims that this mutual preference
for semitones and tritones holds throughout his career.14 This study
refines this claim by asserting that Pendereckis melodic use of ic1/6
concentration began sporadically and gradually increased during the six-

teen years prior to the Violin Concerto, only to emerge as dominant with
the beginning of the neo-Romantic period. During the 1960s,
Pendereckis melodic lines generally do not spotlight the tritone any
more than the other four non-semitonal interval classes, particularly the
minor third. In the early 1970s, works such as the First Symphony
(1973) and the Magnificat (1974) include extended melodic passages
featuring ic1/6 concentration, although these are partially eclipsed by
the numerous other passages that do not. Moreover, some of the few
exceptions to this gradual process are programmatically motivated: for
example, the tritone finds its way into many of the melodic lines in Dies
Irae (1967). Ic1/6 concentration also plays a dramatic role in
Pendereckis first opera The Devils of Loudon (1969): one of the clearest
examples is provided in Example 5, which sets a text whose English trans-
lation is the blood that will flow between us will make us one.
Thus, what sets the use of ic1/6 concentration in early neo-Romantic
works like the Violin Concerto and Paradise Lost apart from its use in
these earlier instances is that, like 9-pc diversity, it permeates the entire
work as a melodic modus operandi. This also sets the early neo-Romantic
works apart from more recent music. Although Thomas is correct that
ic1/6 concentration has been the most conspicuous aspect of his
melodic style since the mid-1970s, it is certainly does not overwhelm-
ingly dominate the melodic writing in most of the music from 1980 on as
it does in the Violin Concerto and Paradise Lost. Rather, it becomes one
stylistic thread that the composer weaves into a new compositional fabric,
much as the avant-garde techniques that dominated the music of the
early sixties became woven into the synthetic music that followed. Yet
this feature of ic1/6 concentration, coupled with 9-pc diversity, is still
Pendereckis most distinctive melodic fingerprint, even if it is but one of
many stylistic marks presented by a works melodic content. Therefore, in
a piece like his Capriccio for solo tuba (1980), where the abundant
motivic repetition, scherzo and waltz rhythms, and significant octatoni-
cism might make it difficult for even connoisseurs of twentieth-century
art music unfamiliar with this music to name its composer, the few pas-
sages that unambiguously project 9-pc diversity and ic1/6 concentration
disclose Pendereckis identity.
II. MELODIC EXPECTATIONS AND TONAL ANALOGIES
Pendereckis consistent use of 9-pc diversity and ic1/6 concentration,

whether in entire works like the Violin Concerto or Paradise Lost or in
isolated passages from earlier or later works, not only signifies a particular
melodic style, but it also affords someone familiar with the style the abil-
ity to make reasonable predictions about how some of these melodies
might continue. For example, given a six-note melodic incipit CBFE
D A in the context of Pendereckis neo-Romantic style, the claim of
ic1/6 concentration predicts that D , G , or A will probably follow,
since these are the three pcs that ensue when adding the intervals of a tri-
tone, a descending semitone, and an ascending semitone respectively to
the final note A. These kinds of predictions give rise to what will be called
ic1/6 expectations. Alternatively, the claim of 9-pc diversity predicts that
one of the six other pcs that has not yet been stated {C , D, F , G, G ,
A } will probably follow. Admittedly, there is also a fair but presumably
smaller chance that a C will follow, since C occurred first in the melody,
whereas the odds of D coming next are much more slim (unless it is
commencing an interior repetition). As mentioned earlier, the nine-note
average is far from a precise limit of pc diversity, and to use it as such
would be misconstruing Pendereckis melodic style. However, without
more precise statistical information, the predictions based on 9-pc diver-
sity are reluctantly but necessarily simplified by 1) restricting their appli-
cation to pc successions no more than eight pcs long, and 2) only
predicting pcs that have not been stated within such successions. These
kinds of predictions give rise to what will be called 9-pc expectations. The
stylistically informed person bringing these two sets of expectations to a
melody in medias res could be the composer himself deciding what note
to write next, or it could be a performer or an expert listener anticipating
the next pitch. The monophonic and unhurried melodic presentations
that occasionally turn up in Pendereckis neo-Romantic music make this
last scenario far more feasible.
As discussed at the outset, although Penderecki revisits many aspects of
the nineteenth-century style in his neo-Romantic music, he slights
common-practice tonality by not meeting a great majority of the expecta-
tions brought by a tonal ear. Therefore, what remains untapped in
Pendereckis music are familiar experiences like the relative tension inher-
ent in a dominant triad or the stability of a tonic triad, the surprise
wrought by an unexpected harmonic progression, or the satisfactory con-
clusion offered by a familiar cadence. One could interpret their absence
as the composers deliberate restraint from stylistically regressing any fur-
ther. However, a consistent employment of 9-pc diversity and ic1/6 con-
centration, while not creating expectations that imitate those of tonality,
nonetheless may create expectations specifically within the domain of
pitch class that are analogous to, and potentially standing in for, those of
tonality, facilitating a communication of familiar messages to the per-

former, listener, and analyst through a new modernist medium.
However, the familiar tonal messages that will be analogized in this
study do not emanate from what conventional wisdom deems particularly
rich models of tonality, such as Heinrich Schenkers, but rather from
admittedly simplified models of tonality: local event-to-event expecta-
tions of diatonic harmonies as they have been codified by prose and/or
flowcharts in textbook introductions to tonal harmony, or by more
recent quantitative measures.15 One may first question how analogizing
such a rudimentary set of expectations with Pendereckis neo-Romantic
music adequately captures the richness of the latter. In one sense, it
inherently does not, in that these tonal harmonic expectations are limited
to a single parameterpitch classand to Markov first-order probabil-
ities. But, in another sense, such expectations achieve an appreciably high
degree of complexity within these limits, as demonstrated by four basic
conditions shared by most models of basic diatonic harmonic expecta-
tions. To illustrate these conditions, Example 6 employs graphs that
depict an abstract elements expectancy profile. The horizontal axis
arranges all the possible continuations of the element partially ordered
from high to low by the degrees to which these continuations are
expected, shown on the vertical axis. The following four conditions
should be met by any set of expectations proposed as analogous to a text-
book model of diatonic harmonic progression:
EXA MP L E 6: GRA P HIC DE PICTIONS OF THE FOUR CONDITIONS FOR AN

A NA L O G Y TO TONAL EXPECTATIONS
1. The dominant condition: there is at least one highly determinate ele-

ment that has a single most probable and strongly expected continuation. A
fundamental aspect of tonal harmonic syntax is that one firmly expects
dominant harmony to be followed by tonic harmony. Dominant har-
mony is therefore analogous to any element whose number of likely con-

tinuations is, for all practical purposes, narrowed to one. Example 6a
represents the dominant condition with a single spike for the sole highly
determinate continuation.
2. The tonic condition: there is at least one highly indeterminate element
where all of its continuations are expected to more or less the same degree.
Kostka and Payne purport that any chord may follow tonic harmony.16
This results in a flat expectancy profile for the tonic triad as shown in
Example 6b, which means, somewhat paradoxically, that all continuations
are equally expected and that no continuation is particularly expected.
These dual interpretations befit the two customary formal positions of
tonic: since tonic is the least partial regarding what in particular comes
next, a formal unit that opens with tonic is full of possibilities; and since
tonic is the least frustrated when nothing in general comes next, a formal
unit that ends with tonic is an appropriate conclusion. Tonic harmony is
therefore analogous to any element whose number of most likely continu-
ations is the maximum number or, by the same token, zero.
3. The communal gradation condition: the set of elements enjoys a vari-
ance among the degrees to which elements resemble dominant or tonic ele-
ments. These definitions of tonic and dominant elements are idealized
categories, existing as two poles of a continuum of expectancy profiles
whose length is measured by the number of most probable continuations:
from one at the dominant pole, to all (or, equivalently, none) at the tonic
pole. It is significant that the expectations of tonal harmony do not cluster
toward one end of the continuum or the other, resulting in either an
overly deterministic system where every harmony has a singular most
probable continuation, or an overly chaotic system where any harmony
could proceed to any other. Rather, many expectancy profiles of tonal har-
mony fall in between these two extremes. For example, the II chord, as a
pre-dominant harmony, is largely a dominant element as more broadly
defined here, in that its expectancy profile is more honed. However, since
most textbook models bundle the dominant and leading tone harmonies
together into a dominant function, the focus of pre-dominant harmonies
is not quite as sharp as that of dominant harmony, resulting in something
more like the shaded graph of Example 6c than the graph of Example 6a.
Furthermore, the mediant and submediant triads come closer to the tonic
pole: according to Robert Gauldin, they are less focused and more vari-
able17 in what harmony is expected to follow, and better symbolized by
the non-shaded graph of Example 6c. Ultimately, this continuum permits
any element on it to be described relative to another element: one as more
determinate and one as more indeterminate; graphically, one as a steeper
stair-step and one as a gentler stair-step; analogously, one as more domi-

nant and one as more tonic.
4. The individual gradation condition: most elements, besides completely
tonic elements, enjoy a variance among the degrees of expectation for each of
the elements that are likely to follow it. As the communal gradation condi-
tion makes fuzzy the crisp distinction between dominant and tonic cate-
gories, this condition makes fuzzy the crisp distinction between
completely expected and completely unexpected. In other words, this
condition prefers those expectancy profiles with more stair steps. For
example, Walter Pistons classic summaries of normative harmonic pro-
gressions such as IV is followed by V, sometimes I or II, less often III or
VI succinctly capture this concept of individual gradation that most
would ascribe to non-tonic tonal harmonies.18 The shaded graph of
Example 6d offers a simple example of a profile that better meets this
condition than those of Examples 6a and 6c. These intermediate degrees
of expectation permit a qualitative range in how certain continuations
may be interpreted: compare the deceptive near miss of VVI with
the retrogressive contravention of VIV.
How well do the two sets of expectations associated with Pendereckis
neo-Romantic melodic style measure up as analogous to tonal expecta-
tions according to these four conditions? The set of ic1/6 expectations,
applied to any set of melodies, baldly fails to meet any of them. All melo-
dies have the same expectancy profile: a binary delineation between the
three pcs that are each equally probable to follow, and the nine pcs that
are each equally probable not to follow. The number of expected pcs is
too high to meet condition #1, too low to meet condition #2, and too
fixed to meet condition #3; and the binary segregation fails condition #4.
Therefore, although ic1/6 expectations embrace neither complete deter-
minism nor complete chaos, they also completely eschew any singular
instance of these two states, producing a rather humdrum experience as
measured by these four conditions.
The set of 9-pc expectations, applied to the set of non-duplicative mel-
odies that do not exceed eight notes, fares only slightly better. A melody
of a single note amply meets the tonic condition, since any one of the
other eleven pcs have an equal chance of following. This makes formal
sense in the analogy with tonality: pc diversity associates every beginning
of a Penderecki melody with tonic and its potential to head in the highest
number of directions. However, this degree of tonicity, as measured by
the number of probable continuations, never returns anywhere near this
high again to conclude (as proper with the tonal analogy) a longer mel-
ody, for as a melody adds note after note, the number of probable contin-
uations, and thus its association with tonic, steadily and inexorably
decreases. This evinces a communal gradation, but one distinguished

only by the melodys length, not by its content. In other words, two mel-
odies of equal length will have the same expectancy profilehardly a
nuanced model of melodic expectation. Furthermore, a melodys most
probable continuations are equally probable, undermining individual
gradation. 9-pc diversity fails the dominant condition outright: there is
no melody where a single pc continuation predicted by 9-pc diversity
clearly sticks out as more probable than the others.
In light of these disappointments, perhaps a model for melodic expec-
tation in Pendereckis neo-Romantic music should be based on observa-
tions more detailed than these two relatively simple summations of the
composers melodic style as provided by existing scholarship. For
example, ic1/6 expectations could be fine-tuned to accommodate the
observed tendency for tritone leaps to occur less often than semitonal
intervals, or the observed tendency for a tritone leap to be followed by
the ordered pc interval that preceded it. Or the two simplifications earlier
imposed on 9-pc diversity could be lifted: certainly this would effect an
individual gradation, particularly for longer melodies, as pcs stated six to
eight notes earlier would be expected more than those stated more
recently, but less than those not stated yet. However, there is no immedi-
ate need for this, for a model that comes much closer to an analogy to
tonal expectations as embodied in the proposed four conditions can be
simply and elegantly achieved by considering the sets of 9-pc expecta-
tions and ic1/6 expectations working in tandem. That is, a pc is expected
to follow a melody only if it meets both 9-pc expectations and ic1/6
expectations. The first and most obvious product of their collaboration is
the demotion of one of the probable continuations of a melody as pre-
dicted by ic concentration: namely, the melodys penultimate note. In the
case of CBFED A, although ic1/6 concentration expects D to the
same degree as G and A , 9-pc diversity does not expect D at all, since
it was just presented. Thus, the combination of these two features leaves
only two pcs, G and A , as the two equally probable continuations.
However, this collaboration does not seem to affect communal variance:
all melodies, save those singletons without a penultimate note, would
experience this same reduction and be left with two equally probable
continuations. Yet, this is not always the case. While many melodies have
exactly two continuations that meet both criteria, some melodies have
exactly onethus fulfilling the dominant conditionand a few have
nonethus fulfilling the tonic condition. This creates a threefold com-
munal gradation that holds even among melodies of the same length, as
shown in the following analyses.
III. ANALYSES
Wolfram Schwingers examination of the opening of the Violin Concerto

(Example 1) is as follows: The second group of cellos gives out the first
thematic line (bar 6), a chromatic cantabile theme, rising and falling with
a tritone leap A flatD in the middle. This is a brooding theme. Cellos
repeat it, a tone higher and without its last note, whereupon violins (at
bars 9 and 10) keep the notes but change the rhythm in the metamor-
phosis, central to the character of this work (Schwinger 1989, 179).
Example 7a lays out the pcs of Schwingers cantabile theme: F G
A DD CB. The number above each pc is the number of pcs, as
allowed by 9-pc diversity and ic1/6 concentration, that may immediately
follow the melody at that point, and the tree graph that grows down
from the pc sequence indicates the immediate pc continuations not cho-
sen. After the three-way option following the first note, the melody
entertains two probable continuations for its second through fourth
notes: G could continue to A or D , A could continue to D or A, and
D could continue to D or E . However, when the melody executes the
switchback to D , the ground covered earlier greatly decreases the possi-
bility that a tritone leap will follow the D or C, as this would duplicate a
pc already used. The melody therefore has but one probable pc continua-
tion from the D to the C, and from the C to the B, locking it into a
semitonal descent. However, when the melody arrives at and finishes
with the note B, two probable continuations once again avail, returning
to the ambivalence presented earlier in the melody. The progression from
two probable continuations for the beginning, down to one for the
middle, then back up to two for the end creates a less determinatemore
determinateless determinate succession that is analogous with IVI, or,
at the least, a more tonicmore dominantmore tonic progression
using the broader definitions of these terms introduced earlier.
Recall that Schwinger takes care to notice that the statement of the
cantabile theme up a half step in measures 89GG AE DD is
without its last note. This pc sequence is analyzed in Example 7b. This
seemingly unremarkable omission eliminates a resolution to the final I of
the cantabile themes IVI progression, creating an ending analogous to
a half cadence: there is one continuation, far and above the others, that is
most expected to come next, yet the melody concludes nevertheless. Fur-
thermore, the endings of these three melodic statementsthe cantabile
theme in measures 67, its trimmed version in measures 89, and its full
restatement in measures 910realize the IVI progression of the cant-
abile theme on a larger scale.
a. Measures 67, cello 2
3 2 2 2 1 1 2
F# G Ab D Db C B F
C Db A Eb Bb
F
I V I
b. Measures 89, cello 2
3 2 2 2 1 1
G G# A Eb D Db C
C# D Bb E
F#
I V
EXA MP L E 7: EXP ECT A N CY ANALYSES OF THE OPENING OF THE V IOLIN

CONCER TO
EXA MP L E 8: V IOLIN CONCER TO, MEASUR ES 239

3 2 2 2 2 2 1
B Bb A Ab G C# D( D#)
C E Eb D F# C
F
I V
EXA MP L E 9: EXP ECT A NCY ANALYSIS OF V IOLIN CONCER TO, MEASUR ES

239
Therefore, although the B naturals are tonally determinate regarding pc

expectations generated by an F-minor perspective as discussed earlier, they
offer the opposite effect regarding expectations generated by the combi-
nation of Pendereckis two idiosyncratic melodic features. An inverted sit-
uation regarding the play between tonal expectations and the expectations
generated by Pendereckis two melodic features occurs right before the
soloists first entrance. The slow introduction gives way to a poco pi mosso
section that builds in volume and textural density for twelve measures to a
climax, followed by a return to the slow introductions F bass pedal and a
simpler texture. Example 8 displays the music that brings the orchestral
exposition to a close. The most two prominent melodies here are a mezzo
forte viola line in seven notes and a piano bass clarinet line that descends
by semitones with a single interior repetition. The two lines form a chro-
matic 76 suspension chain, then the violas leap into a C dissonant
against the bass clarinet B bass pedal, then resolve it to D. This faint but
sustained minor-third consonance ushers in the soloist. However, from
the perspective of expectations generated by Pendereckis two melodic
features, the violas D is relatively determined and thus unstable. As
shown in Example 9, all of the pitches leading up to the D have two prob-
able continuations; however, the D can only continue to D and stay true
to both melodic features. As proposed earlier, a melody with a single most
probable continuation is analogous to dominant. Thus, according to 9-pc
and ic1/6 expectations, the violas melody ending the orchestral exposi-
tion strongly anticipates a D , in contrast to the slow introduction that
concluded with less anticipation. And indeed, the soloist begins with the
anticipated D , but only after a pregnant pause.
A return to Example 3 provides an attractive example of an incipit in
the concerto that has no probable continuations at allwhat will be
called a terminal melody. Here, the seven notes B AE DD GA
3 2 2 2 2 2 0
Bb A Eb D Db G Ab
E Ab E Ab C F#
B
I I
EXA MP L E 10: EXPECTANCY ANALYSIS OF V IOLIN CONCER TO,
BEGINNING S OF MEASUR ES 273 AND 563
begin a melodic passage in the solo violin accompanied by faster and

slower melodies in the orchestral strings. As shown in Example 10, this
seven-note incipit paints itself into a corner, to reuse an earlier metaphor.
A continuation that honored ic1/6 concentration would reuse a pc (G,
A, or D), and a continuation that honored 9-pc diversity would engage
an interval outside of interval class 1 and 6 (ics 2, 3, or 4). Therefore, in
considering the expectations of both features, there are no pc continua-
tions that are especially anticipated, making the A that ends this seven-
note succession strongly analogous with tonic. Or, to put it another way,
all eight pc continuationsthe five pcs predicted by 9-pc diversity {C, E,
F, F , B} and the three pcs predicted by ic1/6 concentration {D, G, A}
are more or less equally probable. To follow the incipit with more mat
erial, Penderecki must violate one of the two features. In the excerpt of
Example 3, he violates ic1/6 concentration by ascending a minor seventh
to G , and then continues a strict maintenance of pc diversity and ic1/6
concentration up to the E .
EXA MP L E 11: P END ER ECK I, V IOLIN CONCER TO, BEG INNING OF

MEA S U RE 563 (THIR D AND FINAL CADENZ A)
However, unlike the lingering Bs in the opening of the concerto,

Penderecki does nothing else to articulate the tonicity of the A in this
Vivo passage beyond the change of bow. Yet later in the concerto,
Penderecki brings back this material at a softer dynamic and a slower
tempo. The concerto boasts three cadenzas, of which the opening of the
third and last (measure 563) is shown in Example 11. He begins with the
same seven-note incipit, but this time, the A is followed by a caesura (as
is the E when Penderecki starts over and runs through the first ten notes).
The slower tempo and monophonic texture may afford an opportunity for
an expert listener familiar with Pendereckis melodic style to project an
ascending semitone, a descending semitone, and a tritone from this lin-
gering A , and to recognize that all three resultant pcs have already been
presented.19 To call the music stuck at this point is a bit excessiveit
has already been emphasized how these features are not totally binding
but to say that this incipit offers relatively little bias about what will follow
next is appropriate. Therefore, from this point of view, it makes sense to
synchronize the caesura with the tonic quality of this specific moment:
nothing in particular is strongly expected to follow.
3 2 2 2 2 2 2 2 1 0
G F# C B Bb E Eb D Ab A
C# F C# F A F A Db
G#
I V I
EXA MP L E 12: EXPECTANCY ANALYSIS OF EXAMPLE 5
Another example where the end of a terminal melody coincides with a

conclusion articulated by one or more other components is the excerpt
from The Devils of Loudon introduced earlier (Example 5), a ten-pitch
melody that adheres strictly to 9-pc diversity and ic1/6 concentration.
Example 12 provides an analysis, showing that any eleventh pitch must
violate one of these two features. This point of least relative certainty in
what pc will followagain, what has been analogized with the tonic
scale degreeaccompanies the end of the sentence and Jeannes
melodic line. One might contest that, as stated earlier, 9-pc expectations
are applicable only to pc successions no longer than eight notes, and this
ten-note melody clearly exceeds this mark. However, the analytical sub-
ject need not be the entire ten-note melody, but merely the last six
notes, since all the probable ic1/6 continuationsE , A , and B
occur within this final span.
IV. THE PENDERECKI TREE
The above examples provide a few instances of communal gradation

among the expectations of melodies, even among melodies of the same
length, when the features of 9-pc diversity and ic1/6 concentration are
considered in tandem. But does this variety permeate the expectations
generated by all possible melodies that adhere to Pendereckis melodic
practice? The process of answering this question can begin by enumerat-
ing the set of non-duplicative melodies that use only ordered pc intervals
1 (ascending semitone), 6 (tritone), and e (descending semitone): what
will be called Penderecki melodies that form the Penderecki set. Since trans-
position and/or inversion do not change a melodys expectancy profile,
the Penderecki set may be partitioned into melody classes equivalent under
these canonic operations. However, rather than present the Penderecki
set first in tabular form, the set may instead assume a more compact form
that has already been adopted on a small scale so far. An extension of any
one of the trees in the analyses of Examples 7, 9, 10, or 12 to show all of
the Penderecki melodies that began with the same pc would result in a
tree isomorphic to Example 13, the Penderecki tree.20 As in the trees of
the earlier analyses, the vertices represent pcs, the edges represent inter-
vals 1, 6, and e, and every left-to-right path through the tree represents a
single Penderecki melody. However, in the Penderecki tree, every left-to-
right path that begins with the trees rootthe leftmost, zero vertex
corresponds bijectively with a Penderecki melody class. Therefore, each
of the 283 vertices of the Penderecki tree represents (as its terminus) a
melody class, and the number of vertices in each column represents the
number of equivalence classes of a particular cardinality. Because of its
size, the tree has been divided over two pages: the first page presents the
semitone (ST) branch, which indicates all melody classes that begin with
a half step, and the second page presents the tritone (TT) branch, which
indicates all melody classes that begin with a tritone.
Transpositional and inversional equivalences allow for a number of
simplifications of the Penderecki tree. First, the root can be any pc and
the tree would look exactly the same: consider the arbitrary choice of 0 as
the root as putting the Penderecki tree in normal form. Second, the
part of the tree that continues a descending-semitone two-note incipit is
an exact inversion of the part of the tree that continues the ascending-
EXA MP L E 13: T HE P ENDER ECK I TR EE (THE SEMITONE BR ANCH)

EXA MP L E 13 (CONT. ) (THE TR ITONE BR ANCH)
semitone two-note incipit; this is annotated with the words inverse of

01 . . . . Likewise, the part of the tree that continues a <6e> three-note
incipit is an exact inversion of the part of the tree that continues a <61>
three-note incipit; this is annotated with the words inverse of 061 . . .
on the TT branch. To save space, these two parts of the tree have been
pruned. Transposing and/or inverting a melody so that it begins with
012, 017, or 067 is to put the melody in prime form. For example, the
prime form of the melody B AE DC GA from Examples 3 and
11 is 0178932. As one traces the pitches of the prime form through the
tree, the number of right branches from each vertex in the melody repre-
sents the number of pc continuations from that particular point. Prime
form 0178932 (and thus all Penderecki melodies in the same melody
class) enjoys binary branching until the last note, which rests on a termi-
nal vertex, symbolized by the vertexs octagonal shape.
Since Penderecki employs 9-pc diversity instead of 12-pc diversity, his
melodies rarely climb near to the top of the tree. The simplification of
9-pc expectations invoked earlier is represented by the dotted vertical line
that indicates the point farthest from the trees root at which melodies
are less likely to cross than not. Therefore, among the 283 melody
classes, the 111 melody classes that are longer than one note and shorter
than nine notes are of primary analytical concern. These 111 melody
classes are tabulated in the Appendix. Following Allen Fortes nomencla-
ture for unordered set classes, each melody class is assigned a two-
number label (abbreviated LB in the Appendix) and a prime form (PF) in
the two leftmost columns.21 For example, B AE DC GA is a
member of class 7-15 with a prime form of [0178932]. This two-number
label can be identified on the tree with little trouble: the final vertex of
this melody is in the seventh column from the left, and the fifteenth ver-
tex from the top in this column. (Since the tree is divided into two parts,
one must remember to start on the first page when counting from the
top down.) In the next column of the table, each prime form is also char-
acterized by its series of intervals between adjacent pcs, what Robert
Morris calls INT1, simplified to INT.22 The next column provides the
number of most probable pc continuations for every melody class. The
remaining columns will be discussed shortly.
Although the Penderecki tree maintains some structural redundancies
beyond the two aforementioned that were pruned,23 these are overshad-
owed by the trees asymmetric complexity, an overall testament to the
synergistic union of the two relatively simple melodic features of pc diver-
sity and ic1/6 concentration. The table also facilitates an analysis of these
tandem expectations according to the four conditions outlined earlier.
The Appendix column entitled # PC CONT displays the number of
most probable pc continuations of a melody in a melody class. Setting
the three continuations of the root aside, this column reveals a three-
tiered communal gradation that cuts across the cardinality classes: six
melody classes are terminal with zero continuations (most tonic), forty
melody classes have one continuation (most dominant), and the
remaining sixty-five melody classes with two continuations fall in
between the most tonic and most dominant poles on the
communal-gradation continuum. It is curious that, among these 111
melody classes, the six terminal melody classes6-16, 7-15, 8-10, 8-25,
8-34, and 8-41are by far the most rare. In other words that continue
the paintbrush analogy, there are very few ways the composer can paint
himself into a corner so quickly that the wet surface truly pins one in.
Three of these six terminal melody classes7-15, 8-10, and 8-41end
with melody class 6-16; this can be seen by finding 6-16s interval series
of <6116e> at the end of each of their interval series. This promotes 6-16
to an extremely special and nearly singular status among these 111 mel-
ody classes, not completely unlike the singular status tonic harmony has
among the seven diatonic Stufen. The melodies from Examples 5 and 11,
which terminated with textual punctuation and a caesura respectively,
both end with 6-16.
To reiterate, the general heterogeneity and asymmetry of the

Penderecki tree correlates with the general heterogeneity and asymmetry
of tonal expectations, and the model of expectations that combines 9-pc
diversity and ic1/6 concentration meets the first three proposed condi-
tions deemed necessary to be analogous with a tonal model. What of the
fourth condition, that of individual gradation? Regardless of whether the
model generates an expectation of zero, one, or two pcs, there is still a
binary distinction between those pcs that are expected and those that are
not. One could differentiate among those in the latter group by, for
example, claiming that pcs generated by either 9-pc or ic1/6 expectations
are more expected than those generated by neither. For example, in the
case of the previously discussed incipit of CBFED A, this would
assign G and A as the most probable continuations, the D , C , D, F ,
and G as less probable continuations, and the C, E, F, and B as even less
probable. One could also refine this simplified version of 9-pc diversity by
incorporating a probability distribution around the mean of nine that
reflects a more thorough statistical analysis of the frequency of certain
lengths of non-duplicative pc successions. At the least, this would reflect
Pendereckis melodic practice by considerably reducing the expectation
of the penultimate D .
However, a focus on a differentiation among those that are the most
expected is closer to the priorities of tonal expectancy models. Such
models, such as those of diatonic harmony presented in introductory
textbooks, do not fret over the comparison of highly anomalous situa-
tions, such as whether a IV chord or II chord is more expected to follow
a V chord. Rather, they sift through the high end, teasing apart the
degrees of expectations for normative VI progressions versus deceptive
VVI progressions, or predominant IVV progressions versus plagal
IVI progressions. In the case of Pendereckis tandem melodic expecta-
tions, a high end suitable for teasing apart is when a melody has two
equally possible pc continuations, such as the G and A for CBFE
D A. What could make one of these pcs more expected than the other?
Since a seven-note melody falls short of the nine-note average of
Pendereckis pc diversity, it is assumed, in the absence of contrary infor-
mation, that after either the G or A comes next, the melody will con-
tinue to observe both 9-pc diversity and ic1/6 concentration for at most
two more notes. From this point of view, the continuations to G and A
are very different. If the G comes next, there are five more ways to press
on toward an eight- or nine-note well-formed melody. Yet, if the A
comes next, the seven-note melody is thereby terminal: no probable con-
tinuations are available. This can be seen easily enough on the Penderecki
tree: the 3 that concludes the prime form path [017893] is followed by
two branches, yet the branch with six vertices is considerably heavier
than the branch with one vertex.
The three rightmost columns in the table located in the Appendix deal
with this kind of bias for melodies. The column labeled # MEL CONT
shows for a given melody the number of longer melodies, up to a length
of nine, that the given melody begins. For CBFED A, which is in
melody class 6-9, there are seven such longer melodies. The left column
under the BIAS heading partitions a non-zero number of longer
melodic continuations into one or two subsets, parenthetically identified
by the melody class one note longer that begins each of the longer
melodic continuations in the subset. For example, for a melody in class 6-
9, these seven options for continuation are partitioned into the six that
begin with melodies in class 7-14 and the one that begins with a melody
inor, in this case, that belongs toclass 7-15. The final column
expresses the discrepancy between the size of the subsets as 1 (the size of
the smaller subset / the size of the larger subset). Thus, all values will be
between 0 and 1, where 0 represents a melody that is completely unbi-
ased, 1 represents a melody that has a single most probable continuation,
and the numbers in between flesh out the continuum between no bias and
complete bias between two possible pc continuations. In the case of a
melody in class 6-9, the size of the smaller subset is 1, and the size of the
larger subset is 6; therefore, the value is 1 (1/6) or .83. This relatively
high number represents the relatively strong expectation that one pc will
follow compared to the next most likely. The continuum from 0 to 1 now
correlates exactly with the continuum from most tonic to most domi-
nant, since terminal melodies are also assigned a value of zero.
These biases provide a higher-resolution picture of the moment-to-
moment expectations in the previously analyzed melodies. Example 14
revisits the cantabile theme from the opening of the Violin Concerto,
where the label above a certain point in the melody indicates the melody
class to which the succession up to that point belongs, and the value
between 0 and 1 below this point indicates the bias of this melody class.
The graph below charts the change of bias from most tonic (0) to
most dominant (1). The melody starts on tonic with negligible bias,
then the bias increases significantly with the leap to the D, forming mel-
ody class 4-2. Consulting the table in the Appendix, melody class 4-2s
fairly strong expectation (.53) is to be extended with an ascending semi-
tone to melody class 5-3, since it begins just over twice as many longer
melodies between five and nine notes as melody class 5-4 does. However,
the cantabile theme descends a semitone, taking the path less expected.
This demonstrates how this expectancy model accounts for the earlier
observation of the tendency in Pendereckis melodies for a tritone leap to
2-1 3-1 4-2 5-4 6-7 7-11
F# G Ab D Db C B
.02 .07 .53 1 1 .33
EXA MP L E 14: BIA S ED ANALYSIS OF V IOLIN CONCER TO, MEASUR ES 67,

CEL L O 2 (COMPAR E WITH EXAMPLE 7 A)
be followed by the ordered pc interval that preceded it. The bias num-
bers basically predict this phenomenon: no melody class with 2 pc con-
tinuations whose INT ends with <x, 6> has a biased preference for a
melody class whose INT ends with <x, 6, x>. After its channeled contin-
uations to C and B that fully invoke the determinism of dominant, the
melody rests on B, which is not completely tonic (.33), but offers the
closest value to tonic since the beginning of the melody.
Example 15 revisits the caesura-interrupted opening of the final
cadenza. This seven-note succession stays relatively close to tonic; the
small bias at the E (.38) is for the following D, which is also predicted
by the fact that the tritone leap was preceded by a descending semitone.
In fact, each point in this melody continues with the predicted pc
through the penultimate G, where the bias becomes much higher
between the expected continuation to the open-ended F and the unex-
pected continuation to the terminal A . This bias provides a strong
dominant-like expectation that leads to tonic when the melody becomes
stuck on the A , a dramatic change in bias analogous to an authentic
cadence that was not apparent in the analysis of Example 10. Of course,
this is one of multiple places where the analogy falls short: although two
successive moments in a melody can be analogous to dominant and tonic
because of their degree of determinism, the pc highly expected at the
dominant moment is not necessarily the one that creates the succeeding
tonic moment. In this model, dominant and tonic are not entities per se,
but states of expectation.
2-1 3-2 4-3 5-5 6-9 7-15
Bb A Eb D Db G Ab
.02 .38 .35 .27 .83 0
EXA MP L E 15: BIA S ED A NALYSIS OF V IOLIN CONCER TO, BEG INNING S OF

MEA S U RES 273 A ND 563 (COMPAR E WITH EXAMPLE 10 )
One final observation: of all the possible interval classes that can be
paired with Pendereckis ubiquitous interval class 1, this one-dimensional
continuum between tonic and dominant is only possible by combining
the half step with the neo-Romantic interval class 6. Since the tritone
inverts into itself, ic1/6 concentration avails only three ordered pc inter-
vals, and, after the first melodic interval, pc diversity limits the number of
possible pc continuations to two, since an interval followed by its inverse
would result in premature pc duplication. Therefore, with the exception
of the ternary branching from the root, the Penderecki tree is (at most)
binary, which affords the single dimension: how strongly is the melody
expected to go this way as opposed to that way?
***
In Pendereckis neo-Romantic music, there are a few moments, in addi-

tion to those previously discussed, when a conclusion of a phrase or sec-
tion as articulated by silence and/or a pronounced change in texture,
instrumentation, dynamics, tempo, etc. coincides with either a terminal
melody or, more generally, a melody ending dominant to tonic.24
However, dwarfing this phenomenon is the much more frequent situa-
tion in this repertoire when the first means of articulating closure is not
accompanied by the second, and vice versa. Since observations made by
this expectancy model do not strongly correlate with an independent
variable such as the musics formal division, the model does not support a
general theory of closure in Pendereckis neo-Romantic music.
However, this failure does not discount entirely, or even significantly,
the utility of the model presented herein. While the model may not be
able to predict whether a point in a melody in this particular repertoire is
the melodys conclusion, it generally succeeds at predicting what pc will
come next (if indeed the melody continues) at those moments when the
bias is high. So, whereas tonal pc expectations are largely left unrealized
in this music, expectations generated by 9-pc diversity and ic1/6 concen-
tration are much more profitable in their return of realizations. Further-
more, this set of expectations achieves an intricate balance between
complete chance and complete determinism that has been shown to be a
significant component of tonal expectations. Since Pendereckis neo-
Romantic music co-opts conventional tonal syntax to much less of a
degree than other aspects of common-practice music such as tertian har-
monic language and traditional orchestration, this study encourages the
hypothesis that the support of pc expectations generated by the proposed
model stands in for a support of tonal pc expectations.
On the one hand, to argue that this music engages pc expectations that
are analogous to tonal pc expectations might be understood as reinforce-
ment of the view that Pendereckis move toward neo-Romanticism was
truly a step backwards, a view shared by the composer himself. In his
book, Labyrinth of Time: Five Addresses for the End of the Millennium,
Penderecki uses the metaphor of a labyrinth to describe and defend his
return to more traditional idioms.25 Speaking on behalf of artists at the
end of the century, he states, We find ourselves in a labyrinth. We start
down many paths, we back up and return with a highly unclear sense of
our objective.26 He then notes that the current condition of art does
indeed suggest a dead-end. In all domains of art, we can see how ideas
and materials are being used up. . . . We have reached a point where
opening the door behind us is the summit of creativity.27 In another
address, Pendereckis tone is more somber: I cannot think of another
period in the history of music as marked by decline as our own. No new
paths of development can be seen. Artists are exploiting old ideas and
turning back to the past.28
On the other hand, unlike the literal return to common-practice fea-
tures such as tertian harmonies and traditional orchestration, it has been
proposed here that Pendereckis return to tonal pc expectations is ana-
logical. Therefore, the neo in Pendereckis neo-Romanticism may
signify more than Heraclitian anachronism and recontextualization. It
may also signify a means of creating expectations that, while analogous to
tonal expectations in their balance of chance and determinism, stem from
the combination of two melodic characteristics that, prima facie, hardly

support common-practice tonality. Ic1/6 concentration employs the two
rarest interval classes in the diatonic set, and 9-pc diversity offers one
means, akin to Schoenbergs conception of the twelve-tone system, to
avoid a bias toward a central pitch.29 As Mieczlsl/aw Tomaszewski puts it,
Penderecki did not hide his enthusiasm for the music of the end of the
19th century. He believed that this thread which was suddenly cut short
has enough vitality to be continued; of course it was to be done on
another plane, in full awareness of what happened later. He said yes to the
return of melody, but one which did not forget the purifying lesson given
to it by dodecaphony and serialism.30 The two spatial metaphors of
Pendereckis labyrinth and Tomaszewskis plane fit well together.
Penderecki may have retreated from the dead end in many obvious
ways, but, in at least one subtle way, he accompanied this retreat with the
discovery of a completely novel expression of a time-honored concept
that lifted this neo-Romantic style to a new level, thus blazing a way up
and out of the labyrinth.
A PPENDIX : T HE P ENDERECKI S ET P ARTITIONED INTO M ELODY C LASSES
LB PF INT # PC # MEL CONT BIAS

CONT
2-1 01 1 2 111 55 (3-1), 56 (3-2) .02
2-2 06 6 2 108 54 (3-3), 54 (3-3) 0
3-1 012 11 2 54 28 (4-1), 26 (4-2) .07
3-2 017 16 2 55 34 (4-3), 21 (4-4) .38
3-3 067 61 2 53 32 (4-5), 21 (4-6) .34
4-1 0123 111 2 27 15 (5-1), 12 (5-2) .2
4-2 0128 116 2 25 17 (5-3), 8 (5-4) .53
4-3 0178 161 2 33 20 (5-5), 13 (5-6) .35
4-4z 0176 16e 1 20 20 (5-7) 1
4-5 0678 611 2 31 18 (5-8), 13 (5-9) .28
4-6z 0671 616 1 20 20 (5-10) 1
5-1 01234 1111 2 14 7 (6-1), 7 (6-2) 0
5-2 01239 1116 2 11 7 (6-3), 4 (6-4) .43
5-3 01289 1161 2 16 9 (6-5), 7 (6-6) .22
5-4 01287 116e 1 7 7 (6-7) 1
5-5 01789 1611 2 19 11 (6-8), 8 (6-9) .27
5-6 01782 1616 1 12 12 (6-10) 1
5-7z 01765 16ee 2 19 7 (6-11), 12 (6-12) .42
5-8 06789 6111 2 17 9 (6-13), 8 (6-14) .11
5-9 06782 6116 2 12 11 (6-15), 1 (6-16) .91
5-10z 06712 6161 2 19 12 (6-17), 7 (6-18) .42
6-1 012345 11111 2 6 3 (7-1), 3 (7-2) 0
6-2 01234t 11116 2 6 3 (7-3), 3 (7-4) 0
6-3 01239t 11161 2 6 3 (7-5), 3 (7-6) 0
6-4 012398 1116e 1 3 3 (7-7) 1
6-5 01289t 11611 2 8 3 (7-8), 5 (7-9) .4
6-6 012893 11616 1 6 6 (7-10) 1
6-7z1 012876 116ee 1 6 6 (7-11) 1
6-8 01789t 16111 2 10 4 (7-12), 6 (7-13) .33
6-9 017893 16116 2 7 6 (7-14), 1 (7-15) .83
6-10 017823 16161 2 11 7 (7-16), 4 (7-17) .43
6-11z1 01765e 16ee6 1 6 6 (7-18) 1
6-12z2 017654 16eee 2 11 5 (7-19), 6 (7-20) .17
6-13 06789t 61111 2 8 3 (7-21), 5 (7-22) .4
6-14 067893 61116 2 7 5 (7-23), 2 (7-24) .6
6-15 067823 61161 2 10 6 (7-25), 4 (7-26) .33
6-16 067821 6116e 0 0 0
6-17z2 067123 61611 2 11 6 (7-27), 5 (7-28) .17
6-18z1 067128 61616 1 6 6 (7-29) 1
7-1z1 0123456 111111 1 2 2 (8-1) 1
7-2z1 012345e 111116 1 2 2 (8-2) 1
7-3z2 01234te 111161 1 2 2 (8-3) 1
7-4 01234t9 11116e 1 2 2 (8-4) 1
7-5z3 01239te 111611 1 2 2 (8-5) 1
7-6 01239t4 111616 1 2 2 (8-6) 1
7-7 0123987 1116ee 1 2 2 (8-7) 1
7-8 01289te 116111 1 2 2 (8-8) 1
7-9 01289t4 116116 2 4 3 (8-9), 1 (8-10) .67
7-10 0128934 116161 2 5 3 (8-11), 2 (8-12) .33

CONT
7-11z4 0128765 116eee 2 5 2 (8-13), 3 (8-14) .33
7-12z3 01789te 161111 1 3 3 (8-15) 1
7-13 01789t4 161116 2 5 3 (8-16), 2 (8-17) .33
7-14 0178934 161161 2 5 3 (8-18), 2 (8-19) .33
7-15 0178932 16116e 0 0 0
7-16 0178234 161611 2 6 3 (8-20), 3 (8-21) 0
7-17 0178239 161616 1 3 3 (8-22) 1
7-18z4 01765et 16ee6e 2 5 2 (8-23), 3 (8-24) .33
7-19z5 017654t 16eee6 2 4 1 (8-25), 3 (8-26) .67
7-20z6 0176543 16eeee 2 5 3 (8-27), 2 (8-28) .33
7-21z2 06789te 611111 1 2 2 (8-29) 1
7-22 06789t4 611116 2 4 2 (8-30), 2 (8-31) 0
7-23 0678934 611161 2 4 2 (8-32), 2 (8-33) 0
7-24 0678932 61116e 1 1 1 (8-34) 1
7-25 0678234 611611 2 5 2 (8-35), 3 (8-36) .33
7-26 0678239 611616 1 3 3 (8-37) 1
7-27z6 0671234 616111 2 5 2 (8-38), 3 (8-39) .33
7-28z5 0671239 616116 2 4 3 (8-40), 1 (8-41) .67
7-29z4 0671289 616161 2 5 3 (8-42), 2 (8-43) .33
8-1z1 01234567 1111111 1 1 1 (9-1) 1
8-2z5 012345et 111116e 1 1 1 (9-2) 1
8-3z1 01234te5 1111616 1 1 1 (9-3) 1
8-4 01234t98 11116ee 1 1 1 (9-4) 1
8-5 01239te5 1116116 2 2 1 (9-5), 1 (9-6) 0
8-6 01239t45 1116161 2 2 1 (9-7), 1 (9-8) 0
8-7z2 01239876 1116eee 1 1 1 (9-9) 1
8-8 01289te5 1161116 2 2 1 (9-10), 1 (9-11) 0
8-9 01289t45 1161161 2 2 1 (9-12), 1 (9-13) 0
8-10 01289t43 116116e 0 0 0
8-11 01289345 1161611 2 2 1 (9-14), 1 (9-15) 0
8-12 0128934t 1161616 1 1 1 (9-16) 1
8-13z2 0128765e 116eee6 1 1 1 (9-17) 1
8-14z3 01287654 116eeee 2 2 1 (9-18), 1 (9-19) 0
8-15 01789te5 1611116 2 2 1 (9-20), 1 (9-21) 0
8-16 01789t45 1611161 2 2 1 (9-22), 1 (9-23) 0
8-17 01789t43 161116e 1 1 1 (9-24) 1
8-18 01789345 1611611 2 2 1 (9-25), 1 (9-26) 0
8-19 0178934t 1611616 1 1 1 (9-27) 1
8-20 01782345 1616111 2 2 1 (9-28), 1 (9-29) 0
8-21 0178234t 1616116 2 2 1 (9-30), 1 (9-31) 0
8-22 0178239t 1616161 2 2 1 (9-32), 1 (9-33) 0
8-23z2 01765et4 16ee6e6 1 1 1 (9-34) 1
8-24z3 01765et9 16ee6ee 2 2 1 (9-35), 1 (9-36) 0
8-25z4 017654te 16eee61 0 0 0
8-26 017654t9 16eee6e 2 2 1 (9-37), 1 (9-38) 0
8-27 01765439 16eeee6 2 2 1 (9-39), 1 (9-40) 0
8-28 01765432 16eeeee 1 1 1 (9-41) 1
8-29z5 06789te5 6111116 1 1 1 (9-42) 1
8-30 06789t45 6111161 1 1 1 (9-43) 1
8-31 06789t43 611116e 1 1 1 (9-44) 1
8-32 06789345 6111611 1 1 1 (9-45) 1

CONT
8-33 0678934t 6111616 1 1 1 (9-46) 1
8-34z4 06789321 61116ee 0 0 0
8-35 06782345 6116111 1 1 1 (9-47) 1
8-36 0678234t 6116116 2 2 1 (9-48), 1 (9-49) 0
8-37 0678239t 6116161 2 2 1 (9-50), 1 (9-51) 0
8-38 06712345 6161111 1 1 1 (9-52) 1
8-39 0671234t 6161116 2 2 1 (9-53), 1 (9-54) 0
8-40 0671239t 6161161 2 2 1 (9-55), 1 (9-56) 0
8-41 06712398 616116e 0 0 0
8-42 0671289t 6161611 2 2 1 (9-57), 1 (9-58) 0
8-43 06712893 6161616 1 1 1 (9-59) 1
NOTES
This investigation was supported by the University of Kansas General

Research Fund allocation #2301406 and was first presented as
Pendereckis Melodic Tree: A Preliminary Report at the American
Perspectives on Penderecki Conference in Houston, Texas in October
2004. I thank Brian Bondari and Lon Mitchell for their support of this
research, and Ray Robinson and Deron McGee for reading an earlier
draft of this article.
1. The first four essays in the inaugural volume of Studies in Penderecki

endeavor to categorize Pendereckis output in terms of stylistic peri-
ods. Tomaszewski (1998) calls the neo-Romantic period Time of
Dialog with the Rediscovered Past (19761985), which begins
with the Violin Concerto. Robinson (1998) also begins his period
Synchronization of a More Expressive Musical Language With the
Elements of the Modern Style (19751986) with the Violin Con-
certo, although he acknowledges that The Awakening of Jacob paves
the way for this new direction (40). Chl/opicka (1998) calls the
period The Sphere of Neo-Romanticism (19741980), and con-
centrates on how Paradise Lost and Symphony No. 2 (1980) exem-
plify this phase, but she includes the Violin Concerto in this phase as
well. Schwinger (1998) simply dubs this period Retrospectives,
which spans the time from the Violin Concerto and Paradise Lost to
the Te Deum (1980) and Symphony No. 2. Some also consider the
start of this neo-Romantic period as the primary stylistic shift in the
composers oeuvre. Thomas (2001) divides a discussion of the music
into two parts: music up to 1974, and music after 1974. In an inter-
view with Penderecki, Robinson (1983a) notes that the composer
himself singled out The Awakening of Jacob as the beginning of a sec-
ond period.
2. Schwinger (1980) describes this new period as an exploration of the
realm of melody (11), and suggests in Schwinger (1989) that the
Violin Concerto draws much more from the spring of strong
melodic inspiration (60). Robinson (1983b) finds that this period
owes much to post-Wagnerian chromaticism with its expressive
lines, its lyrical outpouring, and its dramatic highlights (7). Bylander
(2004) states that this new period replaces many of Pendereckis ear-
lier innovations with clearly defined, lyric melodies and traditional
orchestration (10).
3. This transition is immediately apparent by skimming three successive

orchestral works: The Awakening of Jacob (1974) incorporates quar-
ter tones or glissandi into nearly all of its melodic motions, but only
30 out of the 70 pages of Schotts full score of the Violin Concerto
(1977) call for these microtonal elements, and they are completely
absent in the Second Symphony (1980).
4. The term concentration is preferred to saturation from Morris
(1983/1984), because Morriss rows are saturated with a single set
type of a particular cardinality, whereas Pendereckis melodies con-
centrate on two or three interval classes (set types of cardinality 2).
5. This is not unlike an acknowledgment of certain liberties some serial
composers take with their realizations of a row.
6. Robinson (1998), 46, n. 3.
7. This method of analysis is very similar to, although not equivalent to,
the method used to measure pc diversity in Tenney (1977). Tenney
assigns a non-duplicative pc succession to every pc in a Ruggles mel-
ody; the length of this succession is measured as the number of non-
duplicating pcs that lead up to and include the pc. Although this is a
sound method for computing an average pc diversity, it slightly
undervalues how long, on average, a given non-duplicative pc succes-
sion will continue to be non-duplicative, which is more in accordance
with the present interest in melodic expectancy.
8. Cowell (1930), 42: Ruggles writes at least seven or eight different
notes in a melody before allowing himself to repeat the same note,
even in the octave. Seeger (1932), 582: Then again, there are only
twelve semitones to the octave and everyone knows that is a pitiful
number, especially when you only have eight octaves, and, at least
theoretically, no melody [of Ruggles] may repeat a given tone unless
nine or ten others have intervened. Tenney (1977) notes a gradual
increase in the average length of non-duplicative pc successions: from
5.7 in the works from 19191923 to 8.89 in Organum (1944).
However, Slottow (2003) recognizes that Ruggles was quite flexible
in his application of pitch-class nonrepetition. Regardless of the aver-
age number of intervening notes, the actual number fluctuates
widely, and sometimes the method is suspended altogether in the
interests of concentrated motivic repetition, (6) which, to use a con-
cept introduced in this study, may be teased apart from the applica-
tion of pitch-class nonrepetition as an interior repetition.
9. Finally, a principle underlying many of Ivess structural models, self-

generating patterns, and transformational relationships is that a vari-
ety of pitch classes should be involved. In many cases, this concern
translates to some method of completing the aggregate, inviting pro-
vocative but usually unproductive comparisons with twelve-tone
music. The comparisons are more apt with pre-twelve-tone music,
where pitch-class saturation and equalization are common methods
of denying tonal incursions. Lambert (1997), 15.
10. Emanations, with its two string orchestras tuned a semitone apart,
thus paved the way for Anaklasis, Fonogrammi, and Tren [Thren-
ody], works which use novel graphic notation and what was to
become a characteristically broad-brush approach to musical mater
ials and their development. Thomas (2001), 306.
11. Indeed, with the exception of the preference to realize these intervals
as small registral intervals, one is strongly reminded of many of
Weberns rows that have the same interval-class concentration. The
second run in the first three horns (cr.) of Example 4 matches the
rowand its overlapping T5 extensionfrom Weberns Variations
op. 30.
12. Thomas (2001), 307.
13. Schwinger (1989), 230.
14. Robinson (1998), 46.
15. Lerdahl (2001) offers one set of quantitative measures for tonal
expectations; see pp. 1736 in particular.
16. Kostka and Payne (2004), 109.
17. Gauldin (2004), 121.
18. Piston (1941), 17. He ascribes this three-fold individual gradation to
tonic as well!
19. Isaac Stern, the dedicatee of the concerto, plays the beginning of this
third cadenza at = 106, and pauses about one second for this first
caesura in the works first recording on 10 January 1978 (Minnesota
Orchestra; Stanislaw Skrowaczewski, conductor).
20. This tree graph was drawn using Graphviz, an open source graph
visualization application available at www.graphviz.org. I thank Lon
Mitchell for bringing Graphviz to my attention.
21. Within each cardinality class, the equivalence classes are ordered by a
base-3 interpretation of the interval series of its prime form, where
10, 61, and e2.
22. Morris (1987), 107.
23. Two melodies X and Y in different melody classes will have the same
tree structure for their continuation if the unordered content of X
can be transposed or inverted into the unordered content of Y, and
the same transformation maps the last pc of X to the last pc of Y. For
example, if one traces the incipits 0176 and 0671 on the Penderecki
tree, one will see that the two branches that continue from these
points have the same structure, even though the pc content is differ-
ent. Here, I7{0167} = {0167} and I7(6) = 1. Two such melodies, and
the two melody classes to which they each belong, may be called Z-
related, if one is interested in transferring the spirit (but not the let-
ter) of Fortes original relation to the Penderecki tree. These Z rela-
tions are shown in the Appendix: any two melody classes with the
same zn are in a Z relation.
24. For example, in The Awakening of Jacob, the section between
rehearsal numbers 5a and 6 features the first and almost exclusive use
of ic1/6 concentration in the work, in spite of the glissandi that fill
in many of the intervals. Rehearsal 6 undoubtedly begins a new sec-
tion, as the strings suddenly drop out and the sustaining winds begin
to fan out into a cluster. The flute line, which begins six measures
before rehearsal 6 as the highest and the slowest moving in the dense
nine-voice texture and ends as the highest, is a terminal melody that
ends with 6-16. The double bass line, which is the lowest sounding
line throughout this section, is also a terminal melody that ends with
6-16. However, the contrapuntal density of this section makes it
much more difficult to attend to these individual lines.
25. Incidentally, a maze or labyrinth without closed loops is isomorphic
to a tree graph.
26. Penderecki (1998), 21.
27. Ibid., 24.
28. Ibid., 39.
29. According to Schoenberg (1984), The construction of a basic set of
twelve tones derives from the intention to postpone the repetition of
every tone as long as possible. I have stated in my Harmonielehre
that emphasis given to a tone by a premature repetition is capable of
heightening it to the rank of tonic. But the regular application of a

set of twelve tones emphasizes all the other tones in the same man-
ner, thus depriving one single tone of the privilege of supremacy. It
seemed in the first stages immensely important to avoid a similarity
with tonality. (246)
30. Tomaszewski (1998), 23, emphasis mine.
REFERENCES
Bylander, Cindy. 2004. Krzysztof Penderecki: A Bio-Bibliography.

Westport, Conn.: Praeger.
Chl/opicka, Regina. 1998. Stylistic Phases in the Work of Krzysztof
Penderecki. In Studies in Penderecki 1, 5163. Edited by Ray
Robinson and Regina Chl/opicka. Princeton: Prestige.
Cowell, Henry. 1930. New Musical Resources. New York: Knopf.
Gauldin, Robert. 2004. Harmonic Practice in Tonal Music. New York:
W. W. Norton.
Kostka, Stefan and Dorothy Payne. 2004. Tonal Harmony, With an
Introduction to Twentieth-Century Music. Boston: McGraw-Hill.
Lambert, Philip. 1997. The Music of Charles Ives. New Haven: Yale Uni-
versity Press.
Lerdahl, Fred. 2001. Tonal Pitch Space. New York: Oxford University Press.
Morris, Robert. 1983/1984. Set-Type Saturation Among Twelve-Tone
Rows. Perspectives of New Music 22 (FallWinter 1983/Spring
Summer 1994): 187217.
. 1987. Composition with Pitch Classes. New Haven: Yale Univer-
sity Press.
Penderecki, Krzysztof. 1998. Labyrinth of Time: Five Addresses for the
End of the Millennium. Chapel Hill: Hinshaw Music.
Piston, Walter. 1941. Harmony. New York: W.W. Norton.
Robinson, Ray. 1983a. Krzysztof Penderecki: An Interview and Analysis
of Stabat Mater, The Choral Journal 24, no. 3 (November): 716.
. 1983b. Krzysztof Penderecki: A Guide to His Works.
Princeton: Prestige.
. 1998. Pendereckis Musical Pilgrimage. In Studies in
Penderecki 1, 3349. Edited by Ray Robinson and Regina Chl/opicka.
Princeton: Prestige.
Schoenberg, Arnold. 1984. Style and Idea: Selected Writings of Arnold
Schoenberg. Edited by Leonard Stein. Translated by Leo Black.
Berkeley: University of California Press.
Schwinger, Wolfram. 1980. Krzysztof Penderecki: List of Works. Mainz: Schott.

. 1989. Krzysztof Penderecki: His Life and Work. Translated by
William Mann. London: Schott.
. 1998. The Changes in Four Decades: The Stylistic Paths of
Krzysztof Penderecki. In Studies in Penderecki 1, 6581. Edited by
Ray Robinson and Regina Chl/opicka. Princeton: Prestige.
Seeger, Charles. 1932. Carl Ruggles. The Musical Quarterly 18, no. 4
(October): 57892.
Slottow, Stephen. 2003. Wayward Compositional Procedure in the
Music of Carl Ruggles. Institute for Studies In American Music News-
letter 33, no. 1 (Fall): 67, 15.
Tenney, James. 1977. The Chronological Development of Carl
Ruggles Melodic Style. Perspectives of New Music 16, no. 1 (Fall
Winter): 3669.
Thomas, Adrian. 2001. The New Grove Dictionary of Music and Musi-
cians, s.v. Penderecki, Krzysztof. Edited by Stanley Sadie and John
Tyrrell. London: Macmillan.
Tomaszewski, Mieczlsl/aw. 1998. Pendereckis Dialogs and Games with
Time and Place on Earth. In Studies in Penderecki 1, 1332. Edited
by Ray Robinson and Regina Chl/opicka. Princeton: Prestige.
Murphy, Scott, "A model of melodic expectation for some neo-Romantic music of Penderecki",
Perspectives of new music 45/1 (s.l., United States: winter 2007), 184-222.
Copyright 2007 by Perspectives of New Music. All rights reserved. Content compilation
copyright 2016 by Rpertoire International de Littrature Musicale (RILM). All rights reserved.
RILM Abstracts of Music Literature with Full Text contains electronic versions of previously
published journals reproduced with permission. The RILM collection is owned and managed by
Rpertoire International de Littrature Musicale (RILM), 365 Fifth Avenue, New York NY 10016,
USA.
As a RILM user, you may print, download, or send articles for individual use as authorized under
the terms and conditions of this site, as well as under fair use as defined by U.S. and international
copyright law. No content may be otherwise copied or posted without the copyright holders
express written permission.
To view the entire list of journals included within the RILM Abstracts of Music Literature with Full
Text collection, please visit http://rilm.org/fulltext/.

Penderecki

Uploaded by

Copyright:

Available Formats

Penderecki

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Penderecki

Uploaded by

Copyright:

Available Formats

What are some of the neo-Romantic features found in Penderecki's music?

What are some of the neo-Romantic features found in Penderecki's music?

How does Penderecki use tonality in his neo-Romantic works?

How does Penderecki use tonality in his neo-Romantic works?

MURPHY.

FRA Page 184 Thursday, February 8, 2007 1:25 PM

1 PROVIDES a piano-score reduction of the first ten measures to

Melodic Expectation in Penderecki 185

orchestration and instrumental doubling, continual developing variation,

EXA MP L E 1: P ENDEREC K I, V IOLIN CONCER TO (NO. 1 ), MEASUR ES 110

One particular facet of the twilight of Romanticism that is found in

186 Perspectives of New Music

I. PENDERECKIS MELODIC STYLE(S)

A feature of Pendereckis neo-Romantic period is often referred to, in

Melodic Expectation in Penderecki 187

EXA MP L E 2: P ENDER ECK I, V IOLIN CONCER TO, MEASUR ES 1122,

Yet Pendereckis neo-Romantic melodies are hardly reversions to com-

188 Perspectives of New Music

B AA GF . . . . (The horizontal lines below Example 2, which will

EXA MP L E 3: P ENDERECK I, V IOLIN CONCER TO, MEASUR E 273

Given this more narrow definition of a melody, pc diversity is relatively

Melodic Expectation in Penderecki 189

190 Perspectives of New Music

EXA MP L E 4: P ENDERECK I, DE NATUR A SONOR IS NO. 1, R EHEAR SAL 6

Melodic Expectation in Penderecki 191

EXA MP L E 5: P ENDER ECK I, TH E DE VILS OF LOUDON, FIV E MEASUR ES

Ray Robinson, in pinning down Pendereckis general stylistic traits

192 Perspectives of New Music

concentration began sporadically and gradually increased during the six-

II. MELODIC EXPECTATIONS AND TONAL ANALOGIES

Pendereckis consistent use of 9-pc diversity and ic1/6 concentration,

Melodic Expectation in Penderecki 193

194 Perspectives of New Music

tonality, facilitating a communication of familiar messages to the per-

EXA MP L E 6: GRA P HIC DE PICTIONS OF THE FOUR CONDITIONS FOR AN

1. The dominant condition: there is at least one highly determinate ele-

Melodic Expectation in Penderecki 195

mony is therefore analogous to any element whose number of likely con-

196 Perspectives of New Music

stair-step and one as a gentler stair-step; analogously, one as more domi-

Melodic Expectation in Penderecki 197

decreases. This evinces a communal gradation, but one distinguished

198 Perspectives of New Music

Wolfram Schwingers examination of the opening of the Violin Concerto

Melodic Expectation in Penderecki 199

a. Measures 67, cello 2

b. Measures 89, cello 2

EXA MP L E 7: EXP ECT A N CY ANALYSES OF THE OPENING OF THE V IOLIN

EXA MP L E 8: V IOLIN CONCER TO, MEASUR ES 239

200 Perspectives of New Music

EXA MP L E 9: EXP ECT A NCY ANALYSIS OF V IOLIN CONCER TO, MEASUR ES

Therefore, although the B naturals are tonally determinate regarding pc

Melodic Expectation in Penderecki 201

begin a melodic passage in the solo violin accompanied by faster and

EXA MP L E 11: P END ER ECK I, V IOLIN CONCER TO, BEG INNING OF

202 Perspectives of New Music

However, unlike the lingering Bs in the opening of the concerto,

EXA MP L E 12: EXPECTANCY ANALYSIS OF EXAMPLE 5

Another example where the end of a terminal melody coincides with a

Melodic Expectation in Penderecki 203