AGAWU_Começo Seção Central Final

8/10/2019 AGAWU_Comeo Seo Central Final

1/11

CHAPTER2 Criteria for Analysis I 51

Beginnings, Middles, Endings

Aristotelian poetics bequeathed an enduring representation of the structure of

any expressive, temporally bound utterance: it has a beginning, a middle, and anending. Tis formal model or paradigm has an immediate, intuitive appeal. Inpoetry as well as in music, composers and listeners/readers regularly attend to themanipulation of or play with beginnings, middles, and endings. And throughoutthe history of thought about music, ideas relating to these functions exist. Mat-theson described the shape of a musical oratory in terms that recognize thesefunctions,17Koch developed a compositional theory of form featuring appendixesand suffixes,18 and Schenker located a denitive close of a composition in themoment in which 1 appears over I.19More recently, Dahlhaus has put forward atripartite structure consisting of an initial phase, evolution, and epilogue.20 My

own earlier study of classic music postulated a beginning-middle-ending para-digm,21while William Caplins inuential theory of formal functions recast ideaspertaining to beginnings, middles, and endings. Tus the formal unit or themetype known as sentenceconsists of a presentation phrasein which a basic idea isstated and repeated (beginning), followed by a continuation phrasefeaturing frag-mentation, harmonic acceleration, liquidation, and sequential repetition (middle)and a cadential idea(ending). Similarly, the fundamental harmonic progressionsthat dene the classic style are said to belong to one of three categories: prolon-gational, cadential, and sequential. Sequential processes are most characteristicof continuation phases or middles, cadences mark endings (including endings of

beginnings as well as endings of endings), while the stasis of prolongation mayinitiate the structure, prolong it at its middle, or close it.22

And yet, but for a handful of attempts, the beginning-middle-ending modelhas remained implicit in music-theoretical work; it has not come to occupy ascentral a place in current analytical thinking as it might. Tere are two reasonsfor this. One is that the model seems so obvious and banal (Craig Ayreys word) 23that it is not immediately clear how the analyst can explore its ramications in arigorous fashion. But only if one understands beginnings, middles, and endingssolely as temporal locations rather than as complex functionswith conventionaland logical attributes that operate at different levels of structure would the model

seem banal. A second, more signi

cant reason has to do with the unavoidable

17. Johann Mattheson, Der vollkommene Capellmeister, trans. Ernest Harriss (Ann Arbor, MI: UMI

Research Press, 1981; orig. 1739).

18. Heinrich Christoph Koch, Versuch einer Anleitung zur Composition, vols. 2 and 3 (Leipzig: Bhme,

1787 and 1793).

19. Schenker, Free Composition, 129.

20. Carl Dahlhaus, Between Romanticism and Modernism, trans. Mary Whittall (Berkeley: University

of California Press, 1980), 64.

21. Agawu, Playing with Signs, 5179.

22. William E. Caplin, Classical Form: A Teory of Formal Functions for the Instrumental Music of

Haydn, Mozart and Beethoven(Oxford: Oxford University Press, 2000), 35 and 24.

23. Craig Ayrey, Review of Playing with Signs, Times Higher Education Supplement3 (May 1991), 7.


2/11

52 PART I Teory

fact that, as a set of qualities, beginnings, middles, and endings are not locatedin a single musical dimension but cut across various dimensions. In other words,interpreting a moment as a beginning or an ending invariably involves a readingof a combination of rhythmic, melodic, and harmonic factors as they operate in

specic contexts. In an institutional climate in which analysts tend to work withindimensions as specialists, theories that demand an interdimensional approachfrom the beginning seem to pose special challenges. Tese difficulties are, how-ever, not insurmountable, and it will be part of my purpose here to suggest waysin which attending to beginnings, middles, and endings can enrich our perceptionof Romantic music.

For many listeners, the impression of form is mediated by beginning, middle, andending functions. Tchaikovskys First Piano Concerto opens with a powerful begin-ning gesture that, according to Edward T. Cone, dwarfs the rest of what followsadisproportionately elaborate opening gesture that sets the introduction offas an

overdeveloped frame that fails to integrate itself with the rest of the movement.24Some openings, by contrast, proceed as if they were in the middle of a process pre-viously begun; such openings presuppose a beginning even while replacing it witha middle. Charles Rosen cites the long dominant pedal that opens Schumanns Fan-tasy in C Major for Piano, op. 17, as an example of a beginning in medias res.25Andan ending like that of the nale of Beethovens Fifh, with its plentiful reiteration ofthe tonic chord, breeds excess; strategically, it employs a technique that might begured as rhetorically infantile to ensure that no listener misses the fact of ending.Ending here is, however, not merely a necessary part of the structure; it becomes asubject for discussion as wella meta-ending, if you like.26

As soon as we begin to cite individual works, many readers will, I believe, ndthat they have a rich and complex set of associations with beginnings, middles,and endings. Indeed, some of the metaphors employed by critics underscore theimportance of these functions. Lewis Rowell has surveyed a variety of beginningstrategies in music and described them in terms of birth, emergence, origins, pri-mal cries, and growth.27Endings, similarly, have elicited metaphors associated withrest and nality, with loss and completion, with consummation and transgura-tion, with the cessation of motion and the end of life, and ultimately with deathand dying. No more, we might say at the end of Tristan and Isolde.

How might we redene the beginning-middle-ending model for internal

analytic purposes? How might we formulate its technical processes to enableexploration of Romantic music? Every bound temporal process displays a begin-ning-middle-ending structure.Te model works at two distinct levels. First is thepure material or acoustical level. Here, beginning is understood ontologically as thatwhich inaugurates the set of constituent events, ending as that which demarcates

24. Cone,Musical Form and Musical Performance(New York: Norton, 1968) 22.

25. Rosen,Te Classical Style, 452453.

26. Donald Francis Tovey comments on the appropriateness of this ending inA Musician Talks, vol. 2:

Musical Textures(Oxford: Oxford University Press, 1941), 64.

27. Lewis Rowell, Te Creation of Audible Time, in Te Study of Time, vol. 4, ed. J. T. Fraser, N.

Lawrence, and D. Park (New York: Springer, 1981), 198210.


3/11


4/11

54 PART I Teory

reference at the beginning of a musical journey shares the material formsbut notnecessarily the rhetorical presentationof a comparable stability that is needed toground a dynamic and evolving structure at its end. It is also possible that endings,because they close offthe structure, subtend an indispensable function. From this

point of view, if we had to choose only one of the three functions, it would be end-ing. In any case, several of these functional permutations will have to be workedout in individual analyses.

It is not hard to imagine the kinds of technical processes that might be associ-ated with beginnings, middles, and endings. Techniques associated with each of aworks dimensionsharmony, melody, rhythm, texturecould be dened norma-tively and then adapted to individual contexts. With regard to harmony, for exam-ple, we might say that a beginning expresses a prolonged IV(I) motion. (I haveplaced the closing I in parenthesis to suggest that it may or may not occur, or that,when it does, its hierarchic weight may be signicantly less than that of the initiat-

ing I.) But since the beginning is a component within a larger, continuous struc-ture, the IV(I) progression is ofen nested in a larger IV progression to conferprospect and potential, to ensure its ongoing quality. A middle in harmonic termsis the literal absence of the tonic. Tis ofen entails a prolongation of V. Since suchprolonged dominants ofen point forward to a moment of resolution, the middle isbetter understood in terms of absence and promise: absence of the stable tonic andpresence of a dependent dominant that indexes a subsequent tonic. An ending inharmonic terms is an expanded cadence, the complement of the beginning. If thelarger gesture of beginning is represented as IV, then the reciprocal ending gestureis VI. Te ending fullls the harmonic obligation exposed in the beginning, but

not under deterministic pressure. As with the beginning and ending of the begin-ning, or of the middle, the location of the beginning and ending functions of theending may or may not be straightforward. In some genres, endings are signaled bya clearly marked thematic or tonal return or by a great deal of fanfare. In others, wesense the ending only in retrospect; no grand activity marks the moment of death.

Similar attributions can be given for other musical dimensions. In doing so, weshould remember that, if composition is gured essentially as a mode of play, whatwe call norms and conventions are functional both in enactment and in violation.On the thematic front, for example, we might postulate the imperatives of clear state-ment or denition at the beginning, fragmentation in the middle, and a restoration

of statement at the ending, together with epigonic gestures or effects of reminiscence.In terms of phrase, we might postulate a similar plot: clarity (in the establishment ofpremises) followed by less clarity (in the creative manipulation of those premises)yields,nally, to a simulated clarity at the end. In addition to such structural proce-dures, we will need to take into account individual composerly routines in the cho-reographing of beginnings and endings. Beethovens marked trajectories, Schubertsway with extensive parentheses and deferred closure, Mendelssohns delicately bal-anced proportions, and the lyrical inection of moments announcing home-goingin Brahmsthese are attitudes that might be fruitfully explored under the aegis of abeginning-middle-ending scheme. We have space here for only one composer.

As an example of the kinds of insights that might emerge from regarding aRomantic work as a succession of beginnings, middles, and endings on different


5/11


levels, I turn to Mendelssohns Song without Wordsin D major, op. 85, no. 4 (repro-duced in its entirety as example 2.1). Te choice of Mendelssohn is not accidental,for one of the widely admired features of his music is its lucidity. In the collection

Andante sostenuto.

4

s

7

s

s

10

cresc.

13

cresc.

16

pi

19

Example 2.1. Mendelssohn, Song without Wordsin D major, op. 85, no. 4.

(continued)


6/11

56 PART I Teory

of songs without words, each individual song typically has one central idea thatis delivered with a precise, superbly modulated, and well-etched prole. Te com-

positional idea is ofen affectingly delivered. And one reason for the composersuncanny success in this area is an unparalleled understanding of the potentials ofbeginning, middle, and ending in miniatures. I suggest that the reader play throughthis song at the piano before reading the following analytical comments.

We might as well begin with the ending. Suppose we locate a sense of home-going beginning in the second half of bar 26. Why there? Because the rising minorseventh in the melody is the rst intervallic event of such magnitude in the com-position; it represents a marked, superlative moment. If we follow the course of themelody leading up to that moment, we hear a physical rise in contour (starting onF-sharp in 24) combined with an expansion of intervals as we approach the high G

in bar 26. Specically, starting from the last three eighth-notes in bar 25, we hear, insuccession, a rising fourth (AD), a rising sixth (GE), and nally a rising seventh

22

s

cresc.

25

cresc. dim.

28

cresc.

dim.

31

34

Example 2.1. continued


7/11


(AG).Ten, too, this moment is roughly two-thirds of the way through the song,is underlined by an implicative 6/5 harmony that seeks resolution, and representsthe culmination of a crescendo that has been building in the preceding 2 bars. Temoment may be gured by analogy to an exclamation, an expected exclamation

perhaps. It also marks a turning point, the most decisive turning point in the form.Its superlative quality is not known only in retrospect. From the beginning, Men-delssohn, here as in other songs without words, crafs a listener-friendly messagein the form of a series of complementary gestures. Melody leads (that is, functionsas a Hauptstimme); harmony supports, underlines, and enhances the progress ofthe melody; and the phrase structure regulates the temporal process while remain-ing faithful in alignment. Te accumulation of these dimensional behaviors pre-pares bar 26. Although full conrmation of the signicance of this moment willcome only in retrospect, the balance between the prospective and retrospective,here as elsewhere in Mendelssohn, is striking. Luminous, direct, natural, and per-

haps unproblematic (as we might say today), op. 85, no. 4 exemplies carefullycontrolled temporal proling.

Ultimately, the sense of ending that we are constructing cannot be understoodwith respect to a single moment, for that moment is itself a product of a numberof preparatory processes. Consider bar 20 as the beginning of the ending. Why bar20? Because the beautiful opening melody from bar 2 returns at this point afersome extraneous, intervening material (bars 1219). For a work of these modestdimensions, such a large-scale return readily suggests a reciprocal sense of closurewithin a tripartite formal gesture.

If we continue to move back in the piece, we can interpret the passage begin-

ning in bar 12 as contrast to, as well as intensication of, the preceding 11 bars. Notethe quasi-sequential process that begins with the upbeat to bar 12. Phrase-wise, themusic proceeds at rst in 2-bar units (114133, 134153; these and subsequent des-ignations of phrase boundaries in this paragraph all include an eighth-note prex),then continues in 1-bar units in the manner of a stretto (15 4163and 164173), andnally concludes with 2 relatively neutral barsneutral in the sense of declininga clear and repeated phrase articulationof transition back to the opening theme(174193).28Te moment of thematic return on the downbeat of bar 20 is supportednot by tonic harmony as in bar 2 but by the previously tonicized mediant, thusconferring a more uid quality on the moment and slightly disguising the sense of

return. Te entire passage of bars 1219 features rhetorically heightened activitythat ceases with the thematic return in bar 20. If, in contrast to the earlier hearing,the passage from bar 20 to the end is heard as initiating a closing section at the larg-est level of the form, then bars 1219 may be heard as a functional middle.

Finally, we can interpret the opening 11 bars as establishing the songs prem-ises, including its material and procedures. A 1-bar introduction is followed by a4-bar phrase (bars 25). Ten, as if repeating (bar 6), the phrase is modied (bar7) and led through B minor to a new tonal destination, F-sharp minor (bars 893).

28. Bars 174182begin in the manner of the previous 1-bar units but modify their end in order to lead

elsewhere.


8/11

58 PART I Teory

In what would normally be a conrmatory gesture (bars 94113), F-sharp minor isreplaced by the more conventional dominant, A major.Tis is a deliberate shifingof gears, as if to say, Tis is really where we want to be.

If we now return to our initial proposition to hear bar 263as the beginning of

the end, we must still reckon with the fact that this moment is still some way fromthe ending. How does Mendelssohn sustain the compositional dynamic betweenbars 26 and 37? By creating a web of events, all of them promoting the largeragenda of closure. Te indispensable technique is none other than repetition,which Mendelssohn uses in exact and varied forms. Let us follow the main eventsfrom bar 26 on. Afer the high point on G at bar 263, a preliminary attempt to closeis made in bars 283291. But the cadence is evaded: melodically, we hear not 321but 325 (FEA not FED), the 1 sounding in an inner voice so that the lessconclusive melodic 5 can initiate a second attempt at closure. Te local harmony at283291is not V6/45/3I (with the second and third chords in root position) but

the more mobile V6/4V4/2I6.29Part of Mendelssohns strategy here is to embedthe more obvious gestures of closure within a larger descending-bass pattern thatwill lend a sense of continuity to the closing moment. Tis line starts with bass Aon the third beat of 28, passes through G (also in 28) then falls through F-sharp,F-natural, and E before reaching a mobile D on the downbeat of 30, making roomfor an intervening A at 294. A similarly directed bass line preceded this one andserved to prepare the high point of bar 26. We can trace it from the third beat ofbar 23: DCB (bar 23), AAGF(bar 24), then, transferred up the octave,EDCB (bar 25), and nally A (downbeat of 26), the whole spanning an octaveand a half.

Unlike the attempt at closure in bars 2829, the one in bars 3132 reaches itsdestination. A conventional 321 over a VI offers what was previously denied.Many listeners will hear the downbeat of bar 32 as a dening moment, a longed-formoment, perhaps, and, in this context, the place where various narrative strandsmeet. Schenker would call this the denitive close of the composition;30it marksthe completion of the works subsurface structural activity. Syntactic closure isachieved. We might as well go home at this point.

But syntactic closure is only one aspectalbeit an important oneof the fullclosing act. Tere is also a complementary dimension that would secure the rhe-torical sense of the close, for although we have attained 1 over I, we need to savor

D for a while, to repose in it, to dissolve the many tensions accumulated in thecourse of the song. Tis other dimension of closure can be described in differentways: as rhetorical, as gestural, or even as phenomenal. In this song without words,Mendelssohn writes a codetta-like segment (bars 32end) to meet this need.Teselast 6 bars are a tonic prolongation. We sense dying embers, a sense of tranquility,the serenity of homecoming, even an aferglow. We may also hear in them a senseof reminiscence, for the sense that death is approaching can be an invitation to

29. Here and elsewhere, I follow Schenkerian practice in understanding cadential 6/4s as dominant-

functioning chords featuring a double suspension to the adjacent root-position dominant chord.

Hence the symbol V6/45/3.

30. Schenker, Free Composition, 129.


9/11


relive the past in compressed form. It is as if key moments in the form are madeto ash before our very eyes, not markedly as quotations, but gently and subtly, asif in a mist, as if from a distance. One of the prominent elements in this ending isa simple neighbor-note motive, ABA or 565, which was adumbrated in the very

rst bar of the song, where B served as the only non-chord tone within the tonicexpression. Subsequently, the notes B and A were associated in various contexts.Ten, in bars 3233, the ABA gure, now sounding almost like a wail, presses themelodic tone A into our memories. Te V6/5 harmony in the second half of bars32 and 33 may also remind us of the high point in bar 26. Ten, too, we experi-ence a touchingly direct 54321 descent across bars 3335. Tis collection ofscale degrees was introverted in bars 2133, sung in V but without 4 in bars 1011,introverted again in bars 201213, embedded in bars 2829, heard with 5 playingonly an ornamental role in bars 3132, before appearing in its most direct andpristine form in bars 324353. Even the dotted-note anacrusis at bar 324has some

precedent in bars 1112, where it energized the rst major contrasting section inthe song. And the extension of the right hand into the highest register of the piecein the penultimate bar recalls salient moments of intensication around bars 16and 17 and of the high point in bar 26 and its echo in 29.Tese registral extensionsafford us a view of another world. Overall, then, the last 6 bars of Mendelssohnssong make possible a series of narratives about the compositional dynamic, amongwhich narratives of closure are perhaps most signicant.

We began this analysis of Mendelssohns op. 85, no. 4, by locating the beginningof the ending in bar 26; we then worked our way backward from it. But what if webegin at the beginning and follow the course of events to the end? Obviously, the

two accounts will not be wholly different, but the accumulation of expectationswill receive greater emphasis. As an indication of these revised priorities and soas to ll in some of the detail excluded from the discussion so far, let us comment(again) on the rst half of the song (bars 119). Bar 1 functions as a gestural pre-lude to the beginning proper; it familiarizes us with the sound and guration ofthe tonic, while also coming to melodic rest on the pitch A as potential head tone.Te narrative proper begins in bar 2 with a 4-bar melody. We are led eventually tothe end of the beginning in bar 11, where the dominant is tonicized. Mendelssohnsprocedure here (as also frequently happens in Brahms, for example, in the songWie Melodien zieht es mir, op. 105) is to begin with a head theme or motif and

lead it to different tonal destinations. In the rst 4-bar segment (bars 25), theharmonic outline is a straightforward IV. A second 4-bar segment begins in bar6, passes through the submediant in 78, and closes in the mediant in bar 9. But,as mentioned before, the emphatic upbeat to bar 10, complete with a Vii6/5 of V(thinking in terms of A major), has the effect of correcting this wrong destina-tion. If one is looking to locate the end of the beginning, one might assign it to theemphatic cadence on the dominant in bar 11. Yet, the end of the beginning and thebeginning of the middle are ofen indistinguishable.Te exploratory potential sig-naled by A-sharp in bar 7, the rst nondiatonic pitch in the song, confers a gradualsense of middle on bars 711. Tis sense is intensied in a more conventional way

beginning with the upbeat to bar 12. From here until bar 20, the music movesin ve waves of increasing intensity that conrm the instability associated with a


10/11

60 PART I Teory

middle. Example 2.2 summarizes the ve waves. As can be seen, the melodic proleis a gradual ascent to A, reached in wave 4. Wave 3 is interrupted in almost strettofashion by wave 4. Wave 5 begins as a further intensication of waves 3 and 4 butdeclines the invitation to exceed the high point on A reached in wave 4, preferring

G-sharp (a half step lower than the previous A) as it effects a return from what, inretrospect, we understand as the point of greatest intensity. Wave 5 also adopts thecontour of waves 1 and 2, thus gaining a local reprise or symmetrical function. Itemerges that the tonicized mediant in bar 9 was premature; the mature mediantoccurs in bars 1920.

16

17

18

14

12

wave 1

wave 2

wave 3

wave 4

wave 5

Example 2.2. Five waves of action across bars 1220 in Mendelssohn, Song with-out Wordsin D major, op. 85, no. 4.

Stepping back from the detail of Mendelssohns op. 85, no. 4, we see that thebeginning-middle-ending model allows us to pass through a Romantic composi-tion by weighing its events relationally and thus apprehending its discourse. Temodel recognizes event sequences and tracks the tendency of the musical material.In this sense, it has the potential to enrich our understanding of what musiciansnormally refer to asforma complex, summary quality that reects the particularconstellation of elements within a composition. Tere is no mechanical way to

apply a beginning-middle-ending model; every interpretation is based on a read-ing of musical detail. Interpretations may shifdepending on where a beginningis located, what one takes to be a sign of ending, and so on. And while the generalfeatures of these functions have been summarized and in part exemplied in theMendelssohn analysis, the fact that they are born of convention means that someaspects of the functions may have escaped our notice. Still, attention to musicalrhetoric as conveyed in harmony, melody, phrase structure, and rhythm can proveenlightening.

Te beginning-middle-ending model may seem banal, theoretically coarse,or simply unsophisticated; it may lack the predictive power of analytical theories

that are more methodologically explicit. Yet, there is, it seems to me, some wis-dom in resisting the overdetermined prescriptions of standard forms. Tis model


11/11


substitutes a set of direct functions that can enable an individual analyst to getinside a composition and listen for closing tendencies. Musicology has for a longtime propagated standard forms (sonata, rondo, ternary, and a host of others) notbecause they have been shown to mediate our listening in any fundamental way,

but because they can be diagrammed, given a two-dimensional visual appearance,and thus easily be represented on screens and blackboards and in books, articles,and term papers. A user of the beginning-middle-ending model, by contrast,understands the a priori functions of a sonata exposition as mere designation;a proper analysis would inspect the work afresh for the complex of functionsmany of them of contradictory tendencythat dene the activity within, say, theexposition space. To say that a dialogue is invariably set up between the normativefunctions in a sonata form and the procedures on the ground, so to speak, is animprovement, but even this formulation may overvalue the conventional sense ofnormative functions. Analysis must deal with the true nature of the material and

recognize the signifying potential of a works building blocksin short, respondto the internal logic of the work, not the designated logic associated with externalconvention. Reorienting thinking and hearing in this way may make us freshlyaware of the complex dynamism of musical material and enhance our apprecia-tion of music as discourse.

High Points

A special place should be reserved for high points or climaxes as embodiments ofan aspect of syntax and rhetoric in Romantic musical discourse. A high point is asuperlative moment. It may be a moment of greatest intensity, a point of extremetension, or the site of a decisive release of tension. It usually marks a turning pointin the form (as we saw in bar 26 of example 2.1). Psychologically, a single high pointtypically dominates a single composition, but given the fact that a larger wholeis ofen constituted by smaller parts, each of which might have its own intensitycurve, the global high point may be understood as a product of successive localhigh points. Because of its marked character, the high point may last a moment,

but it may also be represented as an extended momenta plateau or region.No one performing any of the diverse Romantic repertoires can claim inno-

cence of high points. Tey abound in opera arias; as high notes, they are sites ofdisplay, channels for the foregrounding of the very act of performing. As such,they are thrilling to audiences, whose consumption of these arias may owe not alittle to the anticipated pleasure of experiencing these moments in different voices,so to speak. Te lied singer encounters them frequently, too, ofen in a more inti-mate setting in which they are negotiated with nuance. In orchestral music, highpoints ofen provide some of the most memorable experiences for listeners, serv-ing as points of focus or demarcation, places to indulge sheer visceral pleasure.

Indeed, the phenomenon is so basic, and yet so little studied by music theorists,that one is inclined to think either that it resists explanation or that it raises no

Documents

AGAWU_Começo Seção Central Final