UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Span structure in Robert Schumann’s late works Wu, Pei-Hua 2002

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2002-0290.pdf [ 2.37MB ]
JSON: 831-1.0090314.json
JSON-LD: 831-1.0090314-ld.json
RDF/XML (Pretty): 831-1.0090314-rdf.xml
RDF/JSON: 831-1.0090314-rdf.json
Turtle: 831-1.0090314-turtle.txt
N-Triples: 831-1.0090314-rdf-ntriples.txt
Original Record: 831-1.0090314-source.json
Full Text

Full Text

S P A N S T R U C T U R E IN R O B E R T S C H U M A N N ' S L A T E W O R K S by PATTY PEI-HUA W U B.Mus., The University of British Columbia, 1998 B . A . , The University of British Columbia, 1999  A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S FOR T H E D E G R E E OF M A S T E R OF A R T S in T H E F A C U L T Y OF G R A D U A T E S T U D I E S School of Music  We accept this thesis as conforming to the required standard  T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A October 2001 © Patty Pei-Hua Wu, 2001  In presenting  this  degree at the freely  thesis  in  University of  available for reference  copying  of  department publication  this or of  thesis by  this  his  partial fulfilment  of  British Columbia,  I agree  and study.  for scholarly or  her  7)AUslO  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  'Xkc^miiU.  £  ,  >W\  requirements that the  I further agree  purposes  representatives.  may be It  thesis for financial gain shall not  permission.  Department of  the  is  that  an  advanced  Library shall make it  permission for extensive  granted  by the  understood be  for  allowed  that without  head  of  my  copying  or  my written  Abstract  Robert Schumann's music is frequently noted for its rhythmic complexity at the surface level, within apparently regular duple and quadruple phrase structures. However, not all of Schumann's music is based on regular phrase lengths or regular hypermetric patterns. The interaction between motivic relationships, different types of accents and harmonic structure in his later music creates a conflict between metric and group structures, and necessitates a more careful consideration of these two categories and of this interaction. This paper examines Schumann's use of these ways of subdividing music, and analyzes five of his later works from 1849 and 1853: 1) the Adagio from the Adagio and Allegro for Piano and Horn, Op. 70. 2) Phantasiestucke, Op. 73, No. 1. 3) Four Marches for Piano, Op. 76, No. 1. 4) Marchenerz'ahlungen, Op. 132, No. 3, and 5) Ges'dnge der Fr'uhe, Op. 133, No. 2.  This paper consists of two parts. The first chapter discusses and clarifies the definitions of the terms that are related to rhythmic structure, such as group, span, phrase, clause, measure, and chunk. The second chapter contains detailed analyses of the five late works by Schumann. The analyses demonstrate the irregular grouping lengths and span structure, as well as the ambiguous hypermeter and accentual structure in his late works. The focus on the span structure in each piece helps to enhance understanding of the music's overall structure. These analyses illustrate an interaction between various modes of partitioning the musical flow in each piece, as a way of showing a coherent rhythmic result that provides a positive picture of Schumann's achievement in these works.  TABLE OF CONTENTS  Abstract  ii  Table of Contents  iii  List of Tables  iv  Acknowledgement  v  INTRODUCTION  1  Chapter 1  Defining the. Terms—Phrase, Group and Span  2  Chapter 2  Analysis of Span Structure in Schumann's Later Works Op. 70, Adagio and Allegro for piano and Horn: Adagio Op. 73, Phantasiestilcke, No. 1 Op. 76, Four Marches for piano, No. 1 Op. 132, M'drchenerz'dhlungen, III Op. 133, Gesange der Fruhe, No. 2  19 27 33 38 45  Conclusion  51  Bibliography  54  List of Tables Table  Page  1. Hypermetric regularity preserved over the course of surface-meter deceleration  23  2. Bipartite division of the Adagio Movement and its expression in terms of spans  24  3. Symmetrical design of spans in mm. 2-37 of Phantasiestilcke, Op. 73, No. 1 :  29  4. Musical form and measure divisions of Phantasiestilcke, Op. 73, No. 1  30  5. A n elaboration of plagal cadence in the Coda, mm. 57-65  31  6. Overall harmonic progression of Phantasiestilcke, Op. 73, No. 1 7. Symmetrical organization of hypermeter  33 41  8. Division of the piece by span in Gesange der Friihe, Op. 133, No. 2  50  ACKNOWLEDGMENTS  I should like to extend my heartfelt appreciation to my supervisor, Dr. William E . Benjamin, who has been a source of inspiration for me and has been a guiding spirit for this paper. His generous guidance and encouragement provided an impetus to me during the long period of time I spent in writing and correcting many drafts. Special thanks are extended to Dr. John Roeder, who offered many precious suggestions and comments for the final draft. Without their great assistance this thesis would not have been completed.  1  Introduction  Unlike the music of the Classical era, that of the 19 century longs for th  strangeness, wonder, and ecstasy. While music in Classical forms aims for structural unity and provides a distinct, often symmetrical, phrase-and-cadence structure, Romantic composers expanded and loosened the instrumental forms they had inherited from the eighteenth century. The enriching and elaboration of harmonies, melodies and rhythms is a distinct characteristic in Romantic music. A true Romantic, Robert Schumann composes music that brims over with impassioned melody, novel changes of harmony, and driving rhythms. However, the wandering of continuities and unclear music divisions, particularly in his late works, have also made the overall structure of much of his music difficult to grasp. This essay consists of two essential parts. The first part discusses and clarifies the definition of key terms, relating to rhythmic structure, that are to be applied in the analyses of the second part. Part II contains the analyses of rhythmic aspects of a few of Schumann's later works, composed in 1849 and 1853. Its purpose is to illustrate how a several different modes of grouping and partitioning the musical flow interact to produce complex, but purposeful and coherent, rhythmic results.  2  Chapter 1  Defining the Terms—Phrase, Group and Span  When trying to understand the form and the structure of an entire piece of music, one must ask oneself, "How do I divide the whole piece into parts?" " Where does one. part end and the next begin?" and "How do I relate these parts to each other and to the whole movement?" Although a full answer to these questions involves the complexity of the interior content of the music, the idea here is that a presupposition of analysis is that we divide the music into parts or subdivisions for a better understanding of the music's structure. Any continuity extracted from the whole of a piece of music is a part or subdivision. A "part" can be of any length or size. Periods, phrases, measures, even single events are all "parts" or "subdivisions" of the music in this general sense. There are various ways of subdividing a piece of music, some of which depend largely on melodic or harmonic relationships, others of which do not. But among these ways are two kinds of part or subdivision, which I will call "group" and "span". A "group" is a continuous succession of musical events, i.e., of musical content. The attack point of the first event is usually understood as the beginning of the group. Where a group ends is often much less clear, particularly in places where groups overlap or are followed by silence. The beginning of Schumann's Adagio movement, Op. 70, provides an example of "groups." As shown in Example 1, the first four measures of the music are grouped into a unit according to the uninterrupted melodic line played by the horn. The second group begins on the second beat of m. 4, the attack of the new musical event by the piano. Similarly, the beginning of the third group starts at the second beat of m. 6, with the  entrance of the horn melody. There is a possibility that group 2 and group 3 could combine into one group since the melodic line played by the horn in mm. 6-8 is an imitation of that of the piano melody in mm. 4-6, thus the sense of a new event succession is weaker in mm. 6-8. Yet even in m. 4 the sense of division is weakened by the fact that the piano enters on the same pitch class as the horn, over the same harmony, as if it was simply continuing the melody.  Example 1  Illustration of music groups in Schumann's Adagio, Op. 70, mm. 1-8  The precise point at which each group ends is typically vague, for three reasons. First, there maybe no clear cadence at the end of each group. That is, the harmonic, melodic and rhythmic factors approaching the ending of a group may not be coordinated in such a way as to clearly signal the ending of a group in a manner conforming to the prevailing style. Second, the final event of the group, for example, the final chord, maybe of indeterminate length; in other words, the pre-final events may not allow one to infer exactly how long the final event should last. Third, the event that initiates the new group may appear to begin too soon, seeming then to overlap the end of the preceding  4  group. This creates a feeling that the music is being cut off by a new group,, and makes the ending of the first group seem to fade out. The point of onset for a group may occasionally be unclear as well. For instance, some may argue that in Example 1, the second group should begin on the downbeat of m. 4 instead on the second beat, since that is where the G4 in the horn part that initiates the melody of the second group begins. However, I mark the second beat in m. 4 as the beginning point of the second group because it is the exact point where a listener is forced to give primary attention to a different instrument.  As can be seen, this is a persuasive, but hardly decisive, criterion.  This kind of subdivision has also been defined by Lerdahl and Jackendoff in their A Generative. Theory of Tonal Music (1983). According to these authors, grouping structure is the most basic component of musical understanding. They use the term "group" as the generic term that refers to the units such as motives, themes, phrases, periods, sections and the piece itself. In their theory, the grouping structure of the music 1  is heard in a hierarchical fashion, as shown in Example 2a, where small units combine together becoming a bigger group in the next level. However, this kind of grouping system is best applied to thematic sections in Classical style, such as in the Mozart example. Such music contains clear motives, and clear phrases with definite cadences and phrase groups that are often symmetrical and made clear by cadence hierarchies. A comparable hierarchic grouping structure for the excerpt from Schumann's Op. 70 in Example J is much more complex, as shown in Example 2b. Level 1 shows many overlapped groups at the level of the music's motives. As well, the groups at each level  'Fred Lerdahl and Ray Jackendoff. A Generative Theory of Tonal Music (Cambridge: MIT Press, 1983), p. 12.  5  contain varying numbers of bars. Going back to Example 2a, level 3 of the grouping structure shows two four-bar groups, and each group is what we clearly know as a "phrase", namely, a group that ends with a cadence. However, not every group at level 3 of a hierarchical grouping structure is a phrase. In Schumann's Adagio movement in Example 2b, the first group at level 3 of the grouping structure ends around the downbeat of m. 4. This group cannot be considered a phrase because it does not end with a clear cadence. Recall that a group may end simply because the music changes to something else, without having a cadence.  Example 2  Comparison of hierarchical grouping structure  a) 1*' movement of Mozart's Sonata, KV 331, mm. 1-8  fry ^ t  11  1U  'II fl  [_JJ  li 1  •  -  *  1  p.  1  -1  M—1  0 m  r  f J f ' f ^ • r r f fr  1 , 1  *  -1  I ^ T IEL  Y—  ff-  ff  1  —1  1  1  ^ L  * L  b) Beginning of Schumann's Adagio, Op. 70, mm. 1-8  VL'titilhom in F.  Pianoforte.  1  ^ P f g C-rU 1 1' = LJJ i' P J J " * f—if— • j^. Jf f f 1f — i — i — i — i — f — '' ' r =  !  h  i i 1  1  F=  i - -JO  , .II  6  As to the ending of the group that ends around the downbeat of m. 4 of Example 2b, some may want to claim that there is in fact a cadence there, a half-cadence to be specific. However, the approach to the V in the bass is very much weakened by the II4/3 in the second half of m. 3, and there is no real agogic accent on the horn G4 in m. 4. More fundamentally, we have no way of knowing, from various stylistic cues, that the phrase is going to end. While it is reasonable to infer that it ends after the fourth downbeat, with a dominant harmony, it could as well have ended with a I in the second half of the fourth bar, or in a fifth bar. In the Mozart case, all kinds of subtle cues let us know how and, much more specifically, where the phrases will end. A phrase may also contain number of sub-phrases. The difference between a phrase and a sub-phrase is that sub-phrases always lack cadences. The norms of a style come into play here. In Classical style, phrases typically extend over four or more bars, while sub-phrases extend over two, as in level 2 of the grouping structure in Example 2a. Groups at level 2 of Example 2b cannot be referred to as sub-phrases because a subphrase is a sub-division of a phrase and has to be occur within a phrase. Since some groups are phrases while others are not, groups may be placed into two categories. In the phrase category are pitch-structural groups that stretch over several measure-downbeats and are clearly end-demarcated by a cadence. The "cadence" is the terminal segment of a phrase in which melodic, harmonic and rhythmic aspects are coordinated to indicate clearly that the segment of music is coming to an end. /  A more precise definition of "a phrase" is given by Peter Westergaard in his textbook on tonal theory. A phrase, 1. establishes one set of pitches and then  7  2. moves to a second set of pitches in such a way that a. we expect those pitches b. we have some sense of when they are about to occur, and c. once they have occurred we know the phrase has gotten where it's going and that no further pitches are needed to complete that phrase.  Westergaard's definition contributes an important emphasis on the pitch structure of a phrase and at the same time focuses on the sense of movement and timing that listeners have. His definition formalizes the distinctions I made in discussing the difference between the Mozart and Schumann examples. Having considered the meaning of "phrase," we shall now look at the second category of group at the same level as phrases. I use the term "clause" for those groups of phrase-like length that are not phrases. A "clause," in its musical meaning, is a pitch structural group that stretches over several downbeats without having a clear cadence, that is. without meeting every or all of the parts of Westergaard's second criterion. The main difference between a "phrase" and a "clause" is therefore at their ends. A group like that stretching over the first four downbeats in Example 2b is a clause. Example  3 presents a tree diagram that shows relations between different types of  divisions. It summarizes and clarifies the terms that we discussed so far, and as well as other key terms relevant to further discussion. As shown in the diagram, there are two major categories or types of musical subdivision, group and span.  Both phrases and  clauses are groups at approximately the 3-measure to 10-measure level, and the main distinction between the two lies in the presence or absence of a cadence. A "sub-phrase" is clause-like, but generally shorter and located within a phrase. While it is true, in a  2  Peter Westergarrd, An Introduction to Tonal Theory (New York: Norton, 1975), p. 311.  8  sense, that a phrase is close-ended and a clause open-ended, this sense has to do with the pitch structural expectations spoken of by Westergaard. In the more general sense I explored above, both phrases and clauses end more or less imprecisely. In extremely clear cases, like Example 2a, we may want to claim that we know exactly where a phrase ends, but in many (perhaps most) cases, this is not true of phrases. This is most evident i f ^ one considers the final phrases and cadences of movements, the artistic effect of which often depends precisely on not knowing exactly where they end.  Example 3  Tree diagram showing different types of divisions  Music Divisions and Partitions  Group Successions of affiliated musical event with vague boundaries  Phrase With cadence  Clause No cadence  Span The Duration of time between two specific points and the events contained therein  Measure Equal span at some level  Chunk Unequal Span at some level  9  We have examined the system of groups on the left side of the tree diagram. We shall now go on to the right side, where additional terms that are important in the subdivision of music are found. At the top of the right side is the term "span". It refers to a stretch of musical time, or to the interval of time or duration between two specific time points, measured in contextually countable units such as beats and measures, as well as to the events occurring in that lime. A span structure thus involves the segmentation of the time within a piece of music, by cutting through it at specific points, so that each span begins at one point of cutting and extends to the next without including it.  3  A time-span structure can be constructed independently of the grouping structure of some music, and it can be either of long spans or of short spans, hence at various levels. As with groups, there are two types of spans at various levels. Whereas the two types of group are distinguished on the basis of content—phrases "tell" us how and roughly where they will end, clauses don't—the two different kinds of spans, although they necessarily have content—even if only silence—are distinguished by comparing their lengths. An uninterrupted series of equal-length spans is a measure-level, and its spans are called measures. Everyone knows what measures are and why they are important in analysis. Less often appreciated are series of spans of comparable but unequal length. I will call these chunk-levels, and their individual spans, chunks. A chunk-level becomes apparent on the basis of accent. That is, the spans of a chunk level must begin with comparably weighted accents, accents that seem to us of comparable salience. It is also true that we prefer chunking when motivically parallel segments are analogously placed in chunks at the same level, but motivic parallelism is largely a  3  William E. Benjamin. A Theory of Musical Meter in Music Perception, Vol. 1, No. 4 (1984), p. 360.  10  criterion of group, not span, structure. Of course, the criterion of comparable, but not equal, length is also important. Example 4 below shows the main difference between measures and chunks.  Example 4  Distinction between measures and chunks  "Measure-level"  "Chunk-level"  There are also some other terms that are relevant to analysis of time-spans. A meter is a hierarchical system of measure-levels in which each level is comprised of measures that subdivide measures at the previous level. Example 5 illustrates this definition of meter. It presents the beginning measures of the first piece of Schumann's March for the piano, Op. 76 (1849). Three measure levels are shown under the music, designated a, b, and c respectively. Considering any of these in isolation we may speak of the beginning point of its measures as pulses. As soon as two or more levels are considered at once, however, we have a meter. The 4/4, or common time, meter that appears in the score tells us that we should take all three of the analyzed measure levels seriously and that they should be heard together, as one system. When this is done, the pulses at level c (and, precisely speaking, the measures they initiate) become beats and the pulses at level b become strong beats. The pulses at level a are called downbeats.  11  Because standard notation tells us almost nothing about measure at levels higher than level a (represented by bar lines), we have no terms to describe their measure beginnings.  Example 5  Illustration of "Meter"  l>'|,  g if J _j if  m.l  Measure-level c Measure-level b Measure-level a  4  111  ~~ r f T T -  -*  m.2  m.3  7 2 3 4 I 7 2 3 4 | i 2 J 4 x x x x x x x x x x x x X  X  X  X  X  X  X  X  X  Many authors have discussed musical accent and attempted to separate accents into different types. Benjamin in his article "A Theory of Musical Meter" (1984) discusses many types of accents and classifies them in three broad categories: gap accents, accents of image shift, and accents of climax. A gap accent occurs where there is a contextually large difference between successive values in some genres that is being followed (intervals, durations, loudness levels, etc.). A n accent of image shift occurs where what has been heard as stable for some time (a pitch, harmony, orchestration, etc.) is replaced by something else (a new pitch, harmony or orchestration). A n accent of climax occurs at a peak (or trough) of a progressively shaped pattern of activity in some  12  dimension (such as pitch or duration). Other authors have a completely different way of classifying accents. For example, Lehrdal and Jackendoff believe that only some accents have physically measurable characteristics (phenomenal accents), while others are tied to a metric or structural interpretation. Whatever one's view of accent, it is clear that measure-levels must, to a considerable degree, be substantiated by accent. At the same time, comparably salient accents may occur in the music without having an equal time-span between each. In other words, accentual time span structures do not necessarily coincide or correspond with measure-levels. Where they do not, they constitute, in effect, a chunk-level. Another term that is important in the analysis of time-spans is hypermeasure.  4  A  hypermeasure is any measure larger than the notated measure. The term refers to a bringing together of measures in which the measure itself serves as a beat. In other words, hypermeasures are supra-measure units that are perceived as if they were measures, because they exhibit a regular alternation of strong and weak beats analogous to that of single measures. For example, Example 6 illustrates the presence of ' hypermeater in a passage from the first piece of Schumann's Ges'dnge der Fruhe, Op. 133. Example 6b presents a durational reduction or time-span reduction of the passage in Example 6a. This type of reduction reduces the rhythmic values of a piece or passage by some constant factor, in this case 4:1 (a whole note equals one beat of a hypermeasure). The passage in Example 6a, according to its accentuation and repeated segments, is divided into three four-bar measures, with the beats in each measure  The term hypermeasure was first introduced by Edward T. Cone in Musical Form and Musical Performance (NY: Norton, 1968).  4  13  numbered "12 3 4." A regular pattern of four-bar units is created in the passage and each four-bar unit becomes a hypermeasure, as shown in the durational reduction in Example 6b.  Example 6  Illustration of Hyper measures and Hypermeter  a) Schumann, Gesiinge der Friihe, Op. 133, No. 1, mm. 19-31  b) Durational reduction of the above passage  T-fif  1 =p=R* < j J ,(V l>'—A : i irr / t * * : ((..PI  1  • *  T — — —4  h  • *  1  }jjf -^&&  t^-r^ 1  7~?  7  Hi;  f  .fj \  Hypermeter, deriving from the term hypermeasure, therefore means the meter at levels above the notated measure. It refers to the combination of measures on a metrical  14  basis. It includes both the recurrence of equal-sized measure spans and a regular repeated pattern of alternation between strong and weak measures. The numbers marked in the measures of Example 6a are therefore to be read like beats in a bar of common time. Hypermeter may sometimes correspond to a phrase structure or may not. As defined earlier in this essay, a phrase is a group with a cadence at the end. For instance, the first hypermeasure (mm. 19-22) in Example 6a corresponds to a phrase since the group draped over its four notatecl-measure downbeats ends with a cadence. The second hypermeasure (mm. 23-26), however, includes no cadence towards the end of its content. Therefore it does not correspond to a phrase structure. One might speak of it as corresponding to the clause indicated in the example. William Rothstein, in his book on phrase rhythm (1989), also makes a comparison between hypermeter and phrase structure, which he considers to be the two main components of musical rhythm at levels larger than the single measure: Hypermeter and phrase structure are both hierarchical in nature. One can speak of levels of hypermeter—larger and smaller hypermeasures, with the former containing the latter; one can refer as well to levels of phrase structure—larger and smaller phrases, the former similarly containing the latter. Large phrases are often known by other names: periods, sections, and ultimately whole movements or pieces all represent levels of phrase structure. 3  Rothstein and others fail to note, however, that phrase structure is a partial aspect of group structure, which also includes clauses, and that hypermeasure-levels are only one aspect of span structure, which also includes chunks. While large-scale rhythm is, in a sense, a counterpoint of groups and spans, it is in another sense an interplay of four subdivisional components: chunk, measure, phrase and clause.  5  William Rothestein. Phrase Rhythm in Tonal Music (NY: Schirmer Books, 1989), p. 13.  15  Hypermeasures typically extend through four notated bars, and phrases often stretch over four notated downbeats. A hypermeasure is a span rather than a group because it is the content of the time-span between two specific points, often notated measure downbeats that are four notated measures apart. Groups, it will be recalled, are comprised of affiliating musical events and usually have no precise boundaries. While regularity of group length (phrase symmetry) is usually found in Classical thematic structures. Romantic music, such as many works of Schumann, contains a more complex grouping structure that very often diverges from a regular pattern. One of the aspects that cause the irregular grouping structure is the overlapping of groups. Example 7 is the second piece of Schumann's Gesange der Fr'uhe, Op. 133.  The main thematic  motive that begins in m. 17 in the alto part is interrupted by the same motive starting in the bass of m. 1 8. Another entry of the motive, before the previous one is finished, comes in at the downbeat of m. 19 in the very top part. One can speak here of two bar groups (groups stretching over two notated downbeats) beginning at successive measure downbeats, and thus overlapping by one bar.  Example 7  Illustration of overlapping groups in Schumann's Gesange der Fr'uhe, Op. 133, No. II, mm. 17-19  )  16  It is clear when one would speak of an overlap of groups. Cases such as that discussed in the preceding example, where motives in different voices overlap, are of one kind. Of another kind are cases in which one or more musical events in a single voice or texture affiliate in two directions at once. Thus, in Example 8, from the first of the Four Marches, Op. 76, the initial Bb's of m. 14 affiliate in two directions. One could therefore speak of two groups overlapping in these Bb's. The first of these groups cadences on the Bb's, and is therefore a phrase. But something more is going on here. Partly because of the hypermetric conventions of the march genre, we feel that the cadential harmony on the downbeat of m. 14 ought to last more or less a full bar and, consequently, that the span (and group) that begun in m. 14 really ought to have started a bar later. In other words, we feel that the surface span structure, in which the music is partitioned at the start of m. 14, masks an underlying span structure (a hypermetric level) of four bar spans, in which there would be a second m. 14, at the start of which a new span (and group) would begin.  Example 8  Illustration of elision in Schumann's March, Op. 76, No. I, mm. 11-15  17  While an overlap simply means the overlapping of groups (either phrases or clauses), we use the term elision to describe overlapping of spans, as shown in the diagram below in Example 9. The numbers indicate the beats of hypermeasures. Elision occurs in the time-span where two brackets across each other. The last beat of the first hypermeasure (which lasts a whole measure long) thus is also the first beat of the second hypermeasure.  Example 9  Illustration of elision  Elision  1 / I 2 I 3 \4=1I 2 I 3 I 4 l  In many cases, as Lehrdal and Jackendoff point out, the effect of an elision is not that a sub-span is shared by two spans, but rather that a terminal event(s) of the first timespan is obscured or overwritten by a simultaneous event in the second. This is the case 6  in Example 8, where the cadential event of the group in the first time span, which ought to have been Bb4 in the upper register, is overwritten by Bb3, the appropriate initial event of the next span (and group). In effect, elision involves our perceptions of span length (hypermeter) and group shape in a subtle, style-conditioned interplay.  6  Fred Lerdahl and Ray Jackendoff, A Generative Theory of Tonal Music (Cambridge: MIT Press, 1983),  p. 55-62.  18  In conclusion, different types of music subdivision can be implied by various musical dimensions, such as motivic relationships, accent, harmonic structure, and stylistic conventions of various genres. In Schumann's later music, there is an unusual richness in the interaction among these different types, as I shall now demonstrate in several extended analytical examples.  19  Chapter 2  Analysis of Span Structure in Several Later Works  This section of my essay consists of structural analyses that focus on the group and span structure in several of Schumann's later works. The works chosen are mostly from the years 1849 and 1 853, the most productive years in the final stage of his life, during which Schumann composed with good spirits throughout. The various types of music division and the specific terms discussed in the first section are applied in each analysis. The melodic and harmonic continuities are carefully examined in each work, with a view to identifying such higher-level rhythmic features as hypermeter and elision. Through the structural analysis of each piece, the complexity and regularity or irregularity of its inner structure should become much clearer..  Op. 70, Adagio and Allegro for Piano and Horn: Adagio Liui^siim, mit iiuiigem  Ausdruck.  1849 was probably Schumann's most productive year for chamber music. Instead of adding to his list of quartets and quintets, he turned to new musical territory: music for one instrument with piano. Schumann seems to have been especially attracted by the  20 horn in this year and was eager to experiment with the tone color of the instrument in many different contexts. The Jagdlieder, Op. 137, which uses four horns as an accompaniment for male chorus, the Concertsliick for four horns and orchestra, Op. 86, and the Adagio and Allegro for horn and piano, Op. 70, all date from 1849. The Adagio and Allegro exemplifies a form used by Schumann and other early nineteenth century composers in concerted works or showpieces for solo instrument and either piano or orchestra. Shorter than the four-movement sonata or the three-movement concerto, it nonetheless provides for a character contrast and allows the soloist to demonstrate lyricism as well as technique. The Adagio of Op. 70 is full of deeply expressive passages that integrate melodies of the two instruments into one continuous flowing line. However, this integration and continuous passing of melodies from one instrument to the other make the group structure of this movement difficult to understand. Clear group boundaries are hard to perceive because, in most cases, one instrument has already started a new musical idea before the other is finished. The chart in Example 10 shows the spans of the whole movement. Instead of following only the melodic phrase structure, which is frequently unclear and lacks conciseness, as mentioned before, the diagram outlines spans created by strong accents at particular measure downbeats (notated or not). Any sustained regularity in spans may be perceived as hypermeter, and deviations from this regularity may be examined to determine if they are the result of elision, or if they represent an untransformed chunklevel.  21  Example 10  Span Structure of the Adagio movement, Op. 70.  O-J rt / ,  4/412  s~Jt  3 4  2 / 4 1 2 1 2 1 2 3 4  4/4  1 2 3 4 5/4  2/4  2  1 2 3 4 5  1 2  The music is mostly based on a hypermetric level of four-measure spans with occasional deviations, as outlined in Example 10. Each quarter note in the chart represents a measure of common time in the score. The first span is in mm. 1-4, where the horn plays the main melodic line above the piano. The melody shifts to the piano in m. 4, where an anacrusis leads to the downbeat of m. 5, the first bar of a new two-bar span.  7  This is immediately answered by the clarinet in mm. 6-8, implying another two-  bar span, mm. 7-8. This is very typical for Schumann's writing, in which one instrument  'instead of hearing the piano melody of m.4 as a pick up to the 5' ' bar, it is also possible to hear the entrance ofthe piano melody at m. 4 as the beginning of the second span. This creates an elision in m.4: 1  4/4 1 2 3 4 2/4 12  ^ 12  22  often imitates the other. At m. 8 there is an elision of one bar, at which point the music returns to the quadruple hypermetric level in mm. 8-15. The beat in m. 15 is again elided, forming a live-measure span in mm. 15-19. This is the first time, and the only time in the piece, that the music deviates from the duple or quadruple hypermeter basis. After another hypermeasure of four bars, the opening music returns in a transformed version at m. 24, creating the same structure of measure spans as in the beginning of the piece.  Example 11  Fluctuating of meter in Adagio movement, mm. 25-40.  \  For most of the time, the music proceeds in accordance with the notated 4/4 meter. However, there is a fluctuating of meter in the music starting in m. 30, as shown in Example 11. The passage uses agogic and dynamic accents, such as the sforzando marking in m. 31, to create irregular downbeats and induces a temporary departure from the original 4/4 meter. The meter at m. 30 thus becomes 3/2 and the hypermetric quarter note in Example 10 now represents six quarter notes in the score. At m. 36, the meter slackens once more, becoming 4/2, so that eight quarter notes equal one beat of the hypermetric reduction. One may speak here of a metric deceleration at the surface level, and it is interesting that hypermetric regularity is preserved over the course of this deceleration, as Table 1 shows. The music returns to original 4/4 meter at m. 38 and maintains regular four-measure hypermetric groupings to the end of the Adagio movement.  Table 1  Hypermetric regularity preserved over the course of surface-meter deceleration.  Measures:  27  Measure spans:  ^—^ 1 2  28  29 ^ 1  30  33  ^ 2  1 . 2  — 3  I  ,  1  Increasing of duration between hyperbeats  -»  24  Going back to Example 10, one sees a total of 17 spans of various lengths in the music.The movement is basically divided into two parts, each of which begins with a form of the opening theme. Table 2 below shows the bipartite division of the Adagio movement and its expression in terms of spans. As shown on the table, the whole movement is made of two parts plus a coda (7+7+3). Both section A l and A2 contains an equal number of spans. Despite the return of the opening theme in the coda, the section essentially brings the movement to an end by sustaining the tonic harmony throughout.  Table 2  Bipartite division of the Adagio movement and its expression in terms of spans.  Sections  Measures  Number of Spans  Al  mm. 1-23  7 spans  A2  mm. 24-49  7 spans  Coda  mm. 50-61  3 spans  The range of harmonic activity in this movement is very limited. In distinction from his earlier practice, even in short works, Schumann in this movement uses a more conservative way of writing tonally, reminiscent of Baroque music. He does not apply many chromatic harmonies or any unusual modulations in the music. Example 12 shows a reduction of the main harmonic activity in the bass. The harmonic motion, for the better part of the piece, is an arpeggiation of pre-dominant harmonies (IV, II, VI).  25  26  Although there are a few places where the music shows arrives on tonic harmony, such as mm. 16-17 and mm. 24-25, the resolutions to these tonic chords are generally weak. As a result, there is a II harmony prolonged from m. 12 through m. 31 as diagrammed in Example 12. One may also argue for the importance of the dominant harmonies that appear in mm. 1 5 and 27. However, these two chords can be analyzed as IV/II, allowing one to hear the supertonic harmony throughout the span. Overall, the music remains in the orbit of the tonic in that all tonicizations are brief and unsupported by distinct cadences.  27  Op. 73, Phantasiestilcke, No. 1 Zart qud mil Ausrtruck.(J »HO.) Oarititillo ia A..  Piaaoforte.  r- fj ^  r~fi  1-7  "  The Phantasiestilcke, Op. 73, also composed in 1849, is a set of three short character pieces for clarinet and piano. In these three pieces, Schumann allows for the solo part to be played, alternatively, on the violin oV violoncello. As early as Op. 12, Schumann used this title for collection of highly charged character pieces in ternary form, in which song-like elements are mixed with more complex textures. This cycle is in the same vein. There is a unity to the work, in which both the second and third movements recall some of the melodic motives from the first piece, and there is only limited contrast among the movements. While the first is meditative, the second light, and the third fiery in its character, all three pieces have similar rhythmic and textural traits. In all three movements, the clarinet part is mostly playing in common time while the piano accompaniment seldom breaks off its triplet figures. As in many of Schumann's late works, the group structure in these pieces is also somewhat indefinite. The melodies in the piano and clarinet interfuse each other, forming irregular groups, as in the. Adagio examined above.  28  Example 13  Spans in Phantasiestilcke, Op. 73, No. 1.  >=J  H  5  5/4  (6/4  1 2 3 4  5  2  6)  i9  3/4  1 2  3 4  ^  >*  3  1 2 3  1 2  4/4  1 2 3 4  3 4  12  1 2 3 4  1 2  1 2  A  33 5/4  5  4/4  2? 1 2 3 4 5  4 5  ( 6/4 1  4/4  1 2 3 4 1 2  2 3 4 5 6 )  Coda 5' 8/4  12  3 4  5 6 7 8  ( 4/4  6/4  1 2 3 4  T*4  12  5 6 3  4 )  4/4 1 2 3 4 (5)  29  As shown in the chart of spans in Example 13, thefirstmeasure of the piece is seen as a pickup measure to the second bar and thefirsthyperbeat of thefirstchunk coincides with the downbeat of m. 2. According to the chart, thefirstchunk is five bars long, extending from m. 2 through m. 6. Starting in m. 7, the music moves in four-bar measures until m. 19 where there is a shift to three-bar measures. Four-bar measures return in mm. 25-32, but after two regular spans, the span beginning on m. 33 is extended to five bars (by the melodic extension in m. 36). It is interesting that within this particular section (mm. 1-37), the number of bars in each chunk or hypermeasure -  decreases and gradually increases again as the music proceeds, as shown in the Table 3 below. The result is a symmetrical structure in the organization of the measure spans within the section.  Table 3  Symmetrical design of spans in mm. 2-37  Measures numbers;  mm. 2-6  7-18  19-24  25-32  33-37  Spans:  5 bars  4 bars  3 bars  4 bars  5 bars  Stalling in m. 37, the opening music returns. The repeated section of music extends to m. 55. An eight-bar span is found in mm.51-58, as shown in the chart in Example 13. Even though another strong accent is created at m. 55, both the melodic and harmonic movements seem to suggest a measure span that extends the preceding four measures. First the melodic line continues at m. 55, and mm. 51-56 are also unified as a  30  prolongation of V. At the same time the strong cadential effect of mm. 55-57 makes it possible to hear mm. 51 -58 as an eight-bar span containing an eight-bar phrase. The downbeat of m. 59 should be understood as the beginning of the Coda because of the return of the principal motive in the solo instrument. The last section of the music (mm. 8  59-69) is divided into two spans (6 measures + 4 measures). The very last beat of the music is placed in parentheses on the chart, to indicate an extra bar that is only an extension of the tonic harmony at the end. The form of the piece can be grouped .by looking at the organization of measure spans and thematic motives in the music. A durational reduction of the whole piece is shown in Example 14, with each measure of the music represented by a quarter note in the reduction. The musical form of the movement is outlined in Table 4 below:  Table 4  Musical form and measure divisions of Phantasiestilcke, Op. 73, No. 1  Sections  Measures  Measure spans  A  mm. 1-21  21 bars (5+4+4+4+3)  B  mm. 22-37  16 bars (3+4+4+5)  A  mm. 38-58  21 bars (5+4+4+8)  mm. 59-69  11 bars (6+5)  Coda,  Notes m. 1 treated as a  pickup  8 = (4+2+2)  It is possible to hear m m . 51-56 as a 6-bar group and m. 57 as initiating a new 4-bar group. The latter is most tangible via the 4-bar phrase leading from A to D in the bass in these bars. While this interpretation obscures the arrival o f the coda at m. 59, it restores the 4-bar hypermeter towards the end o f the piece. This restoration may be heard as complementing the stability o f the regained tonic.  31  It is interesting that, in the A section, the beginning of each span of the reduction contains a descending third motive in eighth-notes, bracketed in the durational reduction in Example 14. This motivic parallelism reinforces the span structure. The motive is rhythmically altered starting in the B section. Even though there does not seem to be a clear contrast between the A and B sections, there is enough evidence to call section B a "Development". It begins with the minor-second motive from the opening piano part. The motive is answered by the clarinet and then is expanded in the next few measures (mm. 22-26), where the minor-second motives keep appearing, in the clarinet in m. 24 and in the top part of the piano throughout. While this section develops a motive from the A section, it also contains new materials, such as the continuous arppegio figures played by the clarinet starting in m. 29. As shown in the durational reduction in Example 14, the music stays mostly in the home key-A minor. The overall harmonic movement begins with the tonic and modulates to III (C major) at the end of the A section, in m. 19-21. The pre-dominant harmonies VI and II are reached, respectively, in mm. 29 and 35. A dominant harmony at m. 37 leads the music back to the A section, which finally concludes with a perfect cadence at mm. 56-57. The Coda presents an elaborate plagal cadence, as shown in Table 5.  Table 5  An elaboration of plagal cadence in the Coda, mm. 57-65.  i  Measures: Harmonic Progression:  m. 57 I ->  m. 60  m. 62  m. 63  m. 65'  IV -> (16/4) -> IV -> I  32  Example 14  Durational reduction in Phantasiestucke, Op. 73, No. 1.  ?  JL  g f t t * (  f t  ~ y  44-  ttr=£  1  T V  8 Si  4-  3  4  _—  9>y  ^r  \—t  U L f l _i—rj  (.=^4  — —ji; ;  7 r  •('  ,  *t *rjr* 1 f M£\>  ; n  i4-  tf*  —'i  "  ..  i  f/  f  i  J  — i  •—'  — i = f bzz~2^_  •  /  B  Although there are chromatic harmonies throughout the music, Schumann maintains the overall harmonic structure at the simplest level. The overall harmonic progression of the movement, shown in Table 6 below, is as fully paradigmatic of conventional minor tonality as possible.  Table 6  Overall harmonic progression of Phantasiestilcke, Op. 73, No. 1.  Sections: Harmonic Progressions:  A  I -> III  B  VI ^ II ^ V  A  I -» V -> I  Coda  IV -» I  Op. 76, Four Marches for piano, No.l Mit grosster Energie.  The Four Marches for piano, published in 1849, have their source, at least in part, in a great extra-musical inspiration, the Dresden insurrection of 1848. Although he sympathized with the insurrection's goals, Schumann tried to avoid the violence, and fled to the tiny village of Kreischa, where he withdrew from horrible reality and buried himself in composition. He was excited politically and wanted to express his feelings in  34 music. The Four Marches were composed on his return to Dresden in June, as shown in his remarks in a letter to his publisher: You receive here with some marches—not the old Dessauer type—but, rather, republican. It was the best way I could find to express my excitement—they have been written with realfieryenthusiasm. 9  All four pieces are written in a traditional ternary form (A-B-A), with middle sections a little more flowing in style though not in big contrast to the main sections. The key sequence of the four pieces, E-flat, G minor, B-flat and E-flat, suggests that Schumann intended them as a cycle. Although some may consider that the marches are of greater extra-musical significance than musical value, credit is due to Schumann for the abundant rhythmic subtleties that overshadow the prosaic 4/4 march measure. In thefirstpiece of the Marches, the mixture of triplets and duplets intensifies the movement and creates heroic-militant gestures that move vigorously throughout the piece. The main interest, however, lies in the span organization. Atypically for a march, which is usually in four-measure spans, the spans in this piece are very irregular, as outlined in the chart in Example 15. Thefirstspan clearly begins with the strong downbeat in thefirstmeasure, and the music is immediately answered by a three-bar span in mm. 4-6. While the fourth bar of thefirstspan is elided and also becomes the downbeat of the second span, there is no sense of elision at m. 7, where the third span begins. The hypermeter expressed by the spans changes to 2/4 starting in m. 7.  10  At m.  11, the hypermeasure returns to a 4/4, with the last measure elided (m. 14) to the next  Joan Chissell, The Master Musicians Series: Schumann. N.Y.: Farrar, Straus and Giroux, Inc, 1967. p. 86 One may also hear a larger four-bar hypermeasure in mm. 7-10.  9  l0  35  Example 15  Spans in Marches for piano, No. 1  O  4 / 4 1 2 3 4 3/4  1 2 3  2/4 (4/4  CL-L^  4/4  f  2 3 4  1 2 3 4 ) 2 / 4 1 2 1 2 1 2  4/4  1 2 3 4  1 2 3 4 2/4 1 2  2 / 4 1 2 1 2 1 2 3/4  1 2 3  "/—^ /—^ 2 / 4 1 2 1 2 1  1 2 3  3/4 1 2  3  Coda 3/4  1 2  3  4/4 1 2 1 2  2 ^ 1 2 1 2  mm. 53-86 = mm. 1-34  ^ 1 2  ^ 1 2  36  two-bar span. The two-bar spans continue in mm. 16-19, where the music shows a parallelism with mm. 7-10. The music finally reaches an interlude of regularity at m. 20, where the spans are based on a four-bar unit until m. 27. A downbeat is created at m. 28 where the music contains material very much like that of m. 7 or m. 16, thus forming another two-bar span. At m. 29, however, the music begins another new musical idea. The new idea plays itself out in a four-bar phrase (mm. 29-32). A final three-bar span begins in m. 32, with another elision. As shown in the chart in Example 15, the B section (mm. 35-52) basically contains spans of two-bar length, forming a 2/4 hypermeter throughout the section. In the beginning of the B section (measures 35-40), there are two different meters showing on the chart, indicating that there is a hemiola of chunks occurring in the passage. It is clear that the passage (Example 16) contains two-bar melodic and rhythmic chunks. However,  Example 16  Effect of hemiola in mm. 35-42.  37  the left-hand part has a clear motion to Bb on the downbeat of m. 38, which divides the passage harmonically into two three-measure spans. The rest of the B section maintains regular duple hypermeter. A series of elisions is found in mm. 45-48, where the second beat of one hypermeasure is also the beginning of another. A dominant pedal is sustained throughout mm. 48-52 and leads into the reprise at m. 53. The Coda begins in m. 86 over the elided final bar of the A section, and winds its way through a quotation from the second movement of Fantasia, Op. 17, as shown in the Example 17. The section contains four-bar spans, with m. 89 being elided, as diagramed in Example 15. This final section basically prolongs the dominant, tonic and subdominant harmonies of the key. Moving forward with the greatest energy and drive, the piece finally concludes with a bright majestic sonority at the end.  Example 17  Use of quotation from Op. 17 in the Coda  March, Op. 76, mm. 87-88  Fantasia, Op. 17, mm. 210-211  Overall, there are many spans in this movement that begin with strong chords, accented sforzando or double forte,- on a notated downbeat, such as in mm. 1, 4, 7, 9, 14, and so on. There are many elisions of measures and overlapping melodic phrases throughout the piece, and the use of hypermetric hemiola is extraordinary. The same  38 phrase structure does not occur twice in the movement except where the A section is repeated. Phrases rarely end in the expected way and the result is a certain overall waywardness. For example, although the same musical material is used in mm. 24-27 as in mm. 4-6, mm. 4-6 are a three-bar span while mm. 24-27 are a span of four bars. Similarly, in the beginning of the piece, the material at mm. 4-5 is repeated at m.6. But instead of beginning a new span parallel to that of mm. 4-5, m. 6 forms the third bar of a three-bar span, the chunk content of mm. 4-5 being compressed into approximately five beats (m. 6 - m.7, beat 1).  Op. 132, Mdrchenerz'dlilungen, III Huliijrs Tunipu, mit znrtcm Ausdruck.  Ruhiges Tempo, mit sartero Ausdruck. J = 54. _T- -  -£H  • a ••  *- 0 0  1  ~  —  The M'drchenerzahlungen, Op. 132, or Fairy Tale Narrations, were composed as a set of four pieces for piano, clarinet and viola. The work was written in 1853, the year before Schumann's attempted suicide and total mental breakdown. As compared to those written inl849, many of the works of 1853 represent something of a decline, perhaps  indicative of oncoming illness. These very late works perhaps suggest a tired mind, though a mind still capable of creating moments of a strangely poignant kind of beauty, as in the third of the Mcirchenerz'dhlimgen, marked "Ruhiges Tempo, mit zartem Ausdruck" ("A restful tempo, with delicate expression"). Schumann's typical imitative way of musical writing is well represented in this movement, where the clarinet and viola create a dialogue above the piano. The unceasing accompaniment of swaying sixteenth notes in the piano imparts a fine natural flow to the music.  Example 18  Spans in Mdrchenen'dhlungen, Op. 132, III  40  The large-scale rhythmic structure of the piece can also be best understood through the structure of spans shown in the chart in Example 18. The piece begins with a pick-up bar to m. 2. A hypermeter of four-bar spans is established throughout the music, but there are a few places where the spans deviate from the regular four-bar pattern. For instance, the downbeat of m. 18 begins a six-bar chunk, where the clarinet and viola imitate each other with the beginning motive of the opening theme. The music returns to its regular quadruple hypermeter in mm. 24-39. At m. 40, a four-downbeat phrase begins. This overlaps with a three-bar clause on the downbeat of m. 43. The result is a 6/4 chunk. Starting in m. 46, the music returns to the regular four-bar span. Another six-bar chunk occurs at m. 58. The last two spans, starting in m. 64, are four-bar hypermeter. Overall, the movement shows a regularity of four-measure spans throughout, with only occasionally changes of hypermeter. The graph below shows this design. The portions that protrude from the horizontal line indicate the places where the hypermeter changes from its regular 4/4 pattern.  Measures:  1-17  18-23  24-39  40-45  46-57  58-62  64-70  Establishment of, and deviation from, hypermeter  It is also interesting that the deviations from regular hypermeter result in a symmetrical design, as shown in Table 7. The 4/4 hypermeter alternates with changes of meter throughout the movement. The six-bar chunk in the middle acts as an axis of the overall form.  41  Table 7  Measures: 1-17  4/4  Symmetrical organization of hypermeter  18-23  24-39  40-45  46-57  58-62  64-70  6/4'  4/4 4/4  6/4  4/4 4/4  6/4 6/4  4/4  t  jr  J  As in the pieces discussed thus far, the balance between sections of the form is shown by counting hypermetric downbeats and the downbeats of chunks. Such downbeats are marked with numbers in each hypermeasure of the durational reduction as shown in Example 19. From the durational reduction (where a dotted half note is represented by a quarter note), we can see that there are total of 14 hypermetric or chunk downbeats in the movement (14 = 6 + 6 + 4). Each sequence of six downbeats may be considered a section, with the. last four serving as coda. At the ends of both principal sections the melodies played by the two instruments end together and the bass closes with a perfect cadence (mm. 26-27 and mm. 52-53). Both sections are thus harmonically closed. It is interesting that both sections I and II contain the same number of hypermeasures. This once again shows Schumann's sureness of control over the larger levels of form. A summary of the overall form is outlined in Table 8.  43  Table S  Formal organization of hypermetric downbeats  Sections  Measures  Number of Hypermetric Downbeats  I  mm. 1-27  6 hypermetric and chunk downbeats  II  mm. 28-53  6 hypermetric and chunk downbeats  Coda  mm. 54-70  4 hypermetric and chunk downbeats  The durational reduction in Example 19 also shows the overall harmonic activity of the music. In this movement, the harmonic development remains clearly within the tonic key G major. The tonicizations of other keys are rarely stabilized and those keys are all diatonic. As shown in Table 9 below, each section forms a harmonic unity that begins with the tonic and ends with a perfect cadence. At the end of the first section, the tonic is decorated with a flat seventh that carries over into the beginning of section II, blurring the sectional articulation. The coda executes a pair of cadences and extends the final tonic harmony.  Table 9  Overall harmonic movements in each section  Section I (mml-27):  I -> III  VI -» II  V  I  Section II(mm. 28-53):  I 7-> VI - M V -> V-> 1^7 -> V -> I  Coda (mm54-70):  IV-> V -> I -> V-> I  b 7  44  It is also interesting to look at the melodic and harmonic movements through the Schenkerian approach. As indicated in the reduction, thefifthdescent of the melodic line begins at the seventh hypermeasure at m. 28 in the music. It descends by steps and finally reaches the first scale degree at m. 53. The melodic line in the last part of the durational reduction emphasizes scale degrees 1, 3 and 5, suggesting an extension of the tonic harmony. There is a lot of melodic chromaticism in the piece, especially in the piano accompaniment, such as D#-E in m. 1, G#-A and E#-F# in m. 2, G-F# in m. 3 and so forth. These chromatic motives can be found almost in every measure of the music. The use of chromatic figures is also reflected in the durational reduction in Example 19, particularly in the movement from D# to E that occurs prominently throughout. For example, it appears in the alto part of both m. 1 and m. 2, and also occurs in an imitative way in m. 3 of the reduction where the alto answers the bass. The motive is also inverted, moves in the opposite direction from E to D#, as in thefifthand sixth measure of the reduction.  45  Op. 133, Gesange der Fr'uhe, No. 2 Belwbt, merit zu. rasch. h, iso.  The Songs of the Dawn, also written in 1853, was Schumann's last completed pianoforte composition. It is a set of five pieces depicting the approach and advance of the morning, which works more as an expression of feeling than as painting. These pieces are highly differentiated. Each has its own distinct character. Thefirstpiece is a quiet elegy of the utmost solemnity. And while the third piece is full of animation, with a dotted rhythmic patter persisting throughout—presumably designed to depict morning in its fullest glory—number four has a soothing natural flow that concludes with a poignantly lovely final section. The second of the five pieces, however, is probably the most peculiar. With considerable ambiguity of span and group structure, the music may appear to lack direction. The melody and its accompaniment are intertwined in an extreme way, even for Schumann. There is also some confusion in the mood of the piece: while attempting to sound triumphant, thefrequentlyjarring turns of phrase also have a labored, even fearful quality. As compared to the previous music example (the third of Mdrchenerz'dhlungen, Op. 132) which was written in the same year, the second of Morning Songs shows a much more complex and irregular span structure. As the chart in Example 20 shows,  46  there are total of 13 chunks in the movement. The downbeat of the first chunk begins at m. 2, if one sees the first measure as a pickup to the next bar. The music basically shifts back and forth between two-measure and three-measure chunks throughout, with only two incidences (mm. 19-22 and mm. 32-35) where a four-measure chunk occurs. In many cases, the downbeat of a chunk also marks the beginning of a statement of the main thematic motive of the piece, as in mm. 2 , 5 , 7 , 17, 18, and 25. There are a few occasions where chunking is the result of a change of character at a downbeat. For example, in mm. 14 and 27, the music changes texture by using staccato chords. This immediately creates a span of its own.  Example 20  3  3 / 4 1 2  2/4  1 2  3  4/4  Span structure in Morning Songs, Op. 133, II  2 / 4 1 2  1 2  3  4  lObars  4/4  1 2  3  4 bars  4  ||  3/4  fT^3  2 / 4 1 2 1 2  2/4  f~2  1 2  3 / 4 1 2  3 / 4 1 2 3  3  2/4  5 bars  1  2  47  A series of group overlaps is found in mm 17-22. While the alto voice in m. 18 continues a two-bar group begun in m. 1 7, the thematic motive begins in the bass at the downbeat of m. 18 and continues for two bars. The same thing happens a bar later, where the top voice in the music contains the entry of part of the theme and marks the downbeat of a four-measure span that occurs for the first time in the piece. Throughout the movement, the music is dominated by its opening thematic motif, each entry of which is transformed in an artfully continuous way that avoids exact repetitions. Example 2 1 shows different forms of the opening thematic motive. It is introduced in three different versions within the first nine measures of the music: in m. 2, it appears in the alto voice, and it is voiced in m. 5 in the same part.. The subject then occurs again in another form in the bass of m. 7. It reappears, for the first time, i n octaves in mm. 10-11. A s shown in the Example 2 1 , the motive always retains its ; , rhythmic pattern. It is the intervallic relationships that Schumann alters to avoid the direct repetitions.  Example 21  MMZ.  /?,  Various forms of the opening thematic motive  2S~  On  It  3—Sf  TT}.  ft r—r  J 71  1£  izsE  48  It is also interesting that Schumann incorporates quasi-Baroque writing techniques in the organization of the thematic motives. The stretto in mm. 17-19 is an example. There are also a few places that Schumann uses melodic sequences in the music. For example, a sequential pattern based on the beginning part of the thematic motive is found in the top voice of mm. 19-21. A similar pattern is found in mm. 25-26, where the bass contains a short sequence based again on the rhythmic motive from the opening theme. All thematic motives based on the same rhythmic pattern are diagrammed in the Example 22. As in the opening theme, most of these motives begin with an ascending fourth except in mm. 25 and 26. The descending fifth (D-G) in the beginning of the motive in m. 25 is a transposed and inverted version of the simultaneously occurring, paradigmatic rising fourth.  Example 22  Motives based on the rhythmic pattern of opening theme  49  What is quite remarkable about this music is its almost total avoidance o f phrases in favor of clauses. To each chunk there corresponds a clause. For example, the first clause extends from m. 1 to the downbeat of m. 5, the second from the downbeat o f m. 5 through m. 6, the third from the downbeat of m. 7 to the downbeat of m. 10, and the fourth from the downbeat o f m. 10 through most o f m. 1 1. A s can be seen, many o f the clauses overlap in at least one attack. They are clauses, however, and not phrases, because they do not cadence but simply flow into one another. One might argue about some places in the music. For example, some might hear a cadence in mm. 4-5, but it would be difficult to argue for a cadence in mm. 6-7. 9-10. or 1 1-12. In fact there are only two unequivocal cadences in the whole piece. One is the overlapped imperfect authentic cadence at mm. 24-25. The second is the final cadence at mm. 31-32, where another group overlaps, on the downbeat of in. 32, producing glaring parallel fifths. There does not seem to be a clear form to this music. In fact, except for the first piece, all movements of the Songs of the Dawn seem to lack conciseness. Nevertheless, by looking at the span structure and chunk content, an overall form may be discerned.. A s discussed earlier, the music basically alternates between two-bar and three-bar chunks. It can be grouped into longer chunks (refer back to the chart in Example 20). The music may thus be seen as divided into five distinct sections including a coda at the end. A sectional outline is provided in Table 10. Both sections A l and B l begin by stating the principal motive, and each starts with a pickup-bar to the downbeat. Sections A 2 and B2, starting in mm. 12 and 27 respectively, show a parallel change in music texture. According to Table JO, sections A l and B l both contain ten-bar spans while sections A 2 and B2 both consist of five-bar  50  spans. This surprisingly creates a symmetrical whereby the music is clearly divided into two parts (A1+A2 and B1+B2), and each contains an equal number of measures (15 measures). A coda, four bars in length, brings the music to the end.  Table 8  Division of the piece by span in Gesange der Friihe, Op. 133, No. 2  Sections  Measures  Al  mm. 2-11  A2  •mm. 12-16  5 bars (2+3)  Bl  mm. 17-26  lObars (2+4+2+2)  B2  mm. 27-31  5 bars (3+2)  Coda  mm. 32-35  4 bars  Measure Span lObars (3+2+3+2)  The harmonic movement further indicates a clear division of the form that confirms this larger organization of spans in the music. In both section A (mm. 2-16) and B (mm. 17-31), the music begins with dominant harmony, having an applied dominant as a pickup in the previous measure, as shown in the bass of mm. 1-2, and mm. 16-17. As noted above, there are only two places where the cadences appear clearly in the music. In the coda, the tonic harmony is prolonged with plagal chromaticism. As in the third movement of the Marchenerzdhlungen, Op. 132, discussed above, Schumann applies many chromatic harmonies in this particular movement. This creates unstable and dissonant sonorities throughout the piece, which therefore acquires an edge of turmoil within the prevailing mellowness.  51  Conclusion  Schumann's later works, as can be seen from the analyses, often lack clear formal markers and an obvious group structure. The melodic lines in different parts are interwoven in registers throughout, phrases and clauses frequently overlap, and hypermeter is disrupted by elisions. The lack of cadences for long stretches of some pieces—the avoidance of phrases—is another disruptive feature. Therefore, an analysis that focuses on spans (hypermeasures and chunks) can be very helpful in enhancing the understanding of the music's structure. It is important to distinguish between groups and spans. Groups can be determined with the help of careful methods of melodic and harmonic analysis, such as Schenkerian analysis. These approaches allow an examination of the tonal motion that creates group structure. While such group structure demonstrates the coherence of musical passages on the basis of the melodic, harmonic, and rhythmic content, a span structure takes accentual features as primary criteria, buttressing them with motivic data. The life of the music may lie in the counterpoint of these two modes of sub-division. Hypermeter,  derived from hypermeasure, arises where there is regularity in the  length of spans. Schumann is often accused of replying too heavily on two-bar or fourbar hypermeter." As we have seen, however, some of his late works avoid hypermeter almost entirely, proceeding in chunks. In all the works examined, there are significant  " Neal, Jocelyn. Schumann's Hypermeter and Phrase Rhythm: Beyond the Duple Norm. Paper presented at the Annual Meeting o f the Music Theory Society o f N e w Y o r k State, Stony Brook, N e w Y o r k . A p r i l 1996.  52  departures from an established hypermeter, whether by a shift to chunks of nonconformant length, or through the use of elision.  Example 23  Durational reduction of opening of Marchenerzahlungen, Op. 132, III.  a) First measure as a pick-up bar  T 3  ' ) - ) ,rjj * [ i f *f—cr t i '  fj 7 1  p  LH  TTT  ^rff  —  Z>j Firs/ measure as the downbeat  ^ 4 "fjlf tiff tr  ' I  „i- -TTrmh /—  =tf—  1  i  i  - i — i — =  o 1  .—.  if  Many of Schumann's pieces from 1849 and 1853 begin with a pick-up bar to the first hypermetric group. In some of these cases, one may instead hear the first measure of the music as the downbeat of the first hypermeasure. It is clear that in the Adagio and the first of the Marches, the melody and all other parts start at in the very beginning. This creates a hypermeasure starting right at the first measure. However, in some cases, such as in thefirstmovement of Phantasiestilcke, Op. 73, if one hears thefirstmeasure as the  53  beginning of the first span, the melody played by the clarinet would appear on the second beat of the span and create a syncopated effect. Similarly, if the first measure of the third of the Marchenerz'cihhingen, Op. 132 is understood as the beginning of the first hypermeasure, instead of as a pick-up bar, a syncopated effect is created. Example 23 presents two durational reductions. One understands the first measure as a pick-up bar, and the other understands it as the hypermetric downbeat. The result is obtained when one sees the first measure as the downbeat of the first hypermeasure, as in Example 23b, is somewhat unconvincing. Not only is the soloist's beginning accent made into a syncopation, but there is no change'of harmony across the second hypermetric barline, between m. 5 and m. 6. In such cases, a durational reduction can be very helpful in deciding how to scan the spans. Schumann keeps the harmonic movement in these late works very simple. Although there is much chromaticism in the music, particularly in the music from 1853, the tonality is clear throughout, and always expressed in terms of strong harmonic progressions. The most distinctive large-scale harmonic feature is an emphasis on predominant harmonies, as illustrated in each example in Chapter 2. The form of these late pieces is often hard to read based on thematic cues alone. However, when balances and symmetries in the span structure are taken into account, a clear binary or ternary form usually emerges. The degree to which Schumann was able to control these balances and symmetries - intuitively or otherwise - testifies to the strength of his musical mind in the waning days of his career.  54  Bibliography  Bedford, Herbert. Robert Schumann: his life and work. London: Greenwood Press, 1971. Benjamin, William E. "A Theory of Musical Meter." Music Perception, Vol. 1, No. 4 (1984): 355-413. Chissell, Joan. The Master Musicians Series: Schumann. London: J. M. Dent & Sons Ldt, 1967. . Schumann Piano Music. Seattle: University of Washington Press, 1972. Cone, Edward. T. Musical Form and Musical Performance. New York: Norton, 1968. Fuller-Maitland, J.A.. Robert Schumann: 1810-1856. London: Sampson Low, Marston & Co., LTD, 1884. . Schumann's Pianoforte Works. London: Oxford University Press, 1929. . Schumann's Concerted Chamber Music. London: Oxford University Press, 1927. Lerdahl, Fred. & Jackendoff, Ray. A Generative Theory of Tonal Music. Cambridge: MIT Press, 1983. Rothstein, William. Phrase Rhythm in Tonal Music. New York: Schirmer Books, 1989. Schachter, Carl. "Rhythm and linear analysis: aspects of meter." The Music Forum, Vol. VI (1987): 1-59. Walker, Alan. Robert Schumann: the man and his music. London: Barrie & Jenkins Ltd., 1972. Westergaard, Peter. An Introduction to Tonal Theory. New York: Norton, 1975.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items