The Fourth Dimension of a Song

How do composers respond to the sounds of language when they set words to music? What promptsthem to elongate a particular vowel in a melodic line, to emphasize an especially expressive consonant,to highlight or exaggerate the purely sonic elements of a poem? Inspired by a recent article by theeminent literary critic M. H. Abrams called “The Fourth Dimension of a Poem,” which focuses onthe physical sensation of reading poems aloud, this article explores the diverse ways that song compos-ers musicalize the different speech-sounds, or phonemes, of poetry for expressive effect. Case studiesinclude Schubert’s “Nacht und Träume” and Britten’s “The Little Old Table,” fromWinter Words.

Keywords: song, poetic sound, phonemes, text-music relations, M. H. Abrams, Schubert, Britten.

I n a recent essay called “The Fourth Dimension of aPoem,” the eminent literary scholar—and centenarian—M. H. Abrams writes about the physical act of reading a

poem, the oral actions that we make in uttering its words.1

This, for Abrams, is a poem’s so-called fourth dimension, thetactile sensation of producing the sounds of a poem, which hesees as distinct from a poem’s other dimensions (the appearanceof a poem; the sounds of a poem, irrespective of how it feels tomake them; and the meaning of a poem). “We produce thosesounds,” he writes, “by varying the pressure on the lungs,vibrating or stilling the vocal cords, changing the shape of thethroat and mouth, and making wonderfully precise movementsof the tongue and lips.”2 These wonderfully precise movements,Abrams argues, are not ancillary to poetic meaning, a merebackdrop to the “real stuff” of a poem (its form and content);they are central to how we experience and understand poetry,and they deserve a more prominent place in discussions of theart form.3 Attending to the material dimension of poetry, he

claims, reveals how the activity of enunciating the words of apoem interacts with the meanings they convey.

Anyone who has ever savored a poem by speaking it aloudwill know what some of these actions feel like. Singers certainlyknow them and train for years so that they can produce soundsthat are beautiful, consistent, intelligible, and evocative. It couldeven be argued that the enunciative component of languageis especially evident in song, even more than in poetry. Many(though clearly not all) songs exist on the page; like poems,songs can have a visible dimension. But unlike poetry, thenormal mode of experiencing a song is through performance,not silent reading.

It is this very notion—that the performative aspect of lan-guage is highlighted in song, essential to its very nature—thatgot me thinking about what Abrams’s ideas might have toteach us about musical settings of poetry. The very act ofputting words to music would seem to require an awareness ofhow the mouth moves as it performs those words and a sensitiv-ity to the effects that can be produced by emphasizing certainsounds over others—which is why it is so surprising that thephysical dimension of poetry has seldom been explored inanalyses of music and text. It is not uncommon for song analyststo discuss the sonic aspects of a poem (commenting on rhymeschemes or noting instances of assonance or alliteration). Yetmore often than not the sounds of poetry have been regardedprimarily as structuring elements within the text rather than assonic elements that a composer might respond to in and ofthemselves.4 Thus far there has been no thoroughgoing study of

I would like to thank Keith Salley for his helpful comments on an earlierdraft of this article, and also the students and faculty at the University ofVictoria and the University of Colorado Boulder, who heard presentationsbased on it. Special thanks to Elissa Guralnick for pushing my thoughts innew directions.

Abrams (2012). Ibid. (2). Abrams slightly overstates his case. Some recent scholars do in fact write

about the physical act of reading poetry. The most prominent among themis the former United States Poet Laureate Robert Pinsky. Pinsky’s booksThe Sounds of Poetry: A Brief Guide (1998) and Singing School: Learning toWrite (and Read) Poetry by Studying the Masters (2013)—as well as hisFavorite Poem Project (favoritepoem.org)—celebrate the sheer pleasure ofreading poems aloud, and the rich and diverse meanings that we derivefrom experiencing poems as physical, embodied things. See also the collec-tion of essays The Sound of Poetry/The Poetry of Sound (2009), which grewout of a 2006 conference organized by Marjorie Perloff and colleagues, andthe chapter on “Sound” from X. J. Kennedy and Dana Gioia’s An Introduc-tion to Poetry (2010), which addresses not only the different vowel and con-sonant patterns of poetry, but also the different sensations involved inuttering them.

There are, to be sure, some exceptions. The most notable is an article byDon Michael Randel (2014), published just this past summer, whichexplores the material aspects shared by music and poetry and analyzes theinteraction of musical and poetic sound in four songs from Schumann’sDichterliebe. (I learned of Randel’s article as I was making final edits to myown.) See also a recent book by the German scholar Konstantin Voigt, Versund Atonalität: Verfahren der Textvertonung in den frei atonalen LiedernArnold Schönbergs und Anton Weberns (2013), which includes some lovelyobservations about Schoenberg’s and Webern’s responses to rhythm,

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

the ways that composers musicalize the speech-sounds, or pho-nemes, of poetry for expressive effect.

My article offers just such a “phonetic” analysis of song.I want to suggest that talk of fricatives and plosives and openvowels can be extended beyond the private voice studio andmingled productively with discussion of modulatory schemes andmetric dissonances and structural melodies. Indeed, I wouldargue that the expressive power of musicalized phonemes is asresponsible for our emotional response to song as the semanticcontent of the words and the musical expression of that content.For listeners and composers alike, how words sound matters noless than what they mean.

In support of this claim, I offer close readings of two songs,Schubert’s famous “Nacht und Träume” and Britten’s “TheLittle Old Table,” from the cycle Winter Words. Each songshows a composer responding sensitively to the phonemes of apoem—but to different types of phonemes. In “Nacht undTräume,” which is in many ways the subtler of the two examples,Schubert controls the timing and length of the poem’s vowels soas to highlight a progression from one predominant vowel sound

to another and back again, a progression that reinforces a relatedpoetic and tonal shift. In “The Little Old Table,” Britten evenmore dramatically treats different consonant types in differentways, setting certain words percussively and other words lyrically,thereby underlining phonetic contrasts that might not be obviouson a first reading of the poem.

“ ̈”

I chose this song in part because of its unnaturally long notevalues, which bring the poem’s vowels to the forefront of ourattention. The singer, for example, enters on a D♯ and holds itfor nearly a full measure, at a “sehr langsam” tempo, and thendoes virtually the same thing with the B in the next measure,before continuing the melody as it arpeggiates downward.5 Themelodic stasis obviously works in conjunction with a harmonicstasis—a single tonic triad pulsates for two full measures, amusical representation of the vast, still night sky that surroundsthe lyric speaker—but the particular vowel sounds in theopening line of Matthäus von Collin’s poem also contributeto this sense of spaciousness.6 The entire poem appears inExample 1; I encourage you to speak the poem aloud, concen-trating on how it feels to produce its various vowel sounds.7

The first two stressed syllables of the poem, “Heil-’ge” and“Nacht,” feature an [a] vowel, one of the most open vowels inthe German language.8 (The “ei” of “Heil’ge” is a diphthong, acombination of [a] and [e], but it is the [a] vowel that would beemphasized when singing the word.)9 The sound of the words

. Collin, “Nacht und Träume,” text and translation.

rhyme, assonance, and alliteration. For the most part, though, commentson the relationship between music and the sounds of poetry tend to beasides made in works on other subjects. Yonatan Malin (2006, 255), forexample, writes about Schubert’s handling of the consonants in Goethe’s“Wandrers Nachtlied I.” David Lewin (2006, 359–60) discusses the rela-tion between “vowel pitch” and musical pitch in a song from Schoenberg’sBuch der hängenden Gärten. In an essay about text setting and rhythm,Michael Cherlin (1991, 71) notes how Copland registrally connects fouralliterative D words in Emily Dickinson’s “The word feels dusty”; Cherlin(2007, 373; 1986, 214) also discusses the first line of Schoenberg’s Mosesund Aron, which progresses from a “high, front” vowel toward a “low,back” one, culminating in the word “Gott” and harking back to the single“O” vowel with which the opera begins. In her book on Winterreise, SusanYouens (1991, 235, 247, 296–97) comments periodically on Schubert’sexpressive handling (and altering) of certain vowels and consonants. RufusHallmark (2010b, 351) makes a brief observation (citing a nineteenth-century critic) about a short vowel that appears on a peculiarly long note inRobert and Clara Schumann’s Rückert Lieder; and in an essay on the rela-tion between music and poetic structure he writes, “It was not only themeaning of a word to which Schubert responded but also the sound of thewords as verbal utterance” (2010a, 9) (the essay, however, focuses more onpoetic rhythm than on phonemes per se). Finally, Stein and Spillman(1996, 86–87) have some suggestive things to say about how the relativeopenness and closedness of vowels relate to melodic register, though theydo not pursue these ideas in the context of a music analysis.

Since many readers will be familiar with this song, I have not included ascore here. A free score can be accessed via the IMSLP Petrucci Library(imslp.org).

Collin’s Nachlaß contains two different versions of the same poem, onecalled “Nacht und Träume” and the other “Nachtfeyer.” The latter closelyresembles the familiar text of Schubert’s song, but the former does not. Itis impossible to know whether Schubert received both versions and thenconstructed a blended version of his own, or whether Collin gave him anunpublished copy that looks just like the song’s text. See Youens (2002,83–84) on the complexities surrounding the poem’s various versions.

I derive my translation from Youens (2002, 84). In keeping with standard texts on lyric diction, I will refer to vowel and

consonant sounds with IPA (International Phonetic Alphabet) symbolsand place these symbols in square brackets.

The [a] vowels in both words are the same, even in spoken German (bothare so-called short [a] vowels).

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

conveys a sense of wonderment that accords beautifully withtheir meaning. Susan Youens describes the opening note of thevocal line as a “long-breathed, quietly ecstatic exhalation ofawe”;10 the word “exhalation” is apt, for the vowel sound of“Heil’ge,” and also “Nacht,” resembles a sigh.

Schubert’s musical rendering of these two words emphasizesthe expansiveness of their sighing sounds—sounds that are ofcourse already present in the text but that need not have beentreated in such a way. “Heil’ge,” after all, is an adjective, syntacti-cally less important than the noun that follows it. A more“normal” approach might have been to set this word as an upbeat(reading the line as “Heil’ge Nacht” rather than “Heil’ge Nacht”).Schubert, however, gives the modifier even more weight than theword it modifies, exaggerating the assonance that connects thefirst two words. Moreover, he downplays the assonant vowels inthe second half of the line (“du sink-est nie-der”): the first syllableof “nieder” may be a half note, but “sink” is a mere eighth note;the speech-sounds may be similar, but the note values associatedwith them are not.11 Youens hears something prosodicallyawkward in the stress on the word “nieder,” but to my ears it is theweight on “Heil’ge” that is especially strange, if wonderfully so.12

The opening phrase is notably three measures long, and theexpansion results from the stretching of the first word, the dwell-ing on its [a] vowel. The melody hovers on this initial “ecstaticexhalation,” and on the exhalation that follows it, before continu-ing more rapidly through the remainder of the line, like the nightand dreams that float down with a preternatural slowness.

This broadening of the “ecstatic” [a] is part of a largerprocess that unfolds over the opening section of the song,whereby the [a] vowels of the first four lines are emphasized andthen allowed to recede as another vowel takes their place. Notethat there are four stressed [a] vowels in lines 1 and 2 (“Heil-’ge,” “Nacht,” “wal-len,” and “auch”) but none in lines 3 and 4(“Wie dein Mondlicht durch die Räume, / Durch der Men-schen stille Brust”). Instead, another vowel emerges: the [ʊ] of“durch die Räume,” “Durch der Menschen,” and “stille Brust,” avowel that does not appear in a stressed position in lines 1 and 2.Youens rightly notes the peculiarity of the downbeat accent on“Durch der Menschen” in m. 12 and suggests that Schubert mayhave had the last line of the poem in mind (“Holde Träume,kehret wieder”) when he wrote this melody, since the melodyreappears verbatim toward the end of the song with thesewords.13 Regardless of the cause of the curious stress on “Durch,”the effect is to alert us to the fact that the fourth line begins andends with the same vowel sound; the downbeat-accented“Durch” prepares us for the more emphatic utterance of an [ʊ]

vowel at the end of the phrase, where the half note on “Brust”also falls on a downbeat (which it need not have done, sinceSchubert could have finished the phrase in the middle of m. 13rather than at the beginning of m. 14, had he not repeated theword “stille”). “Brust” marks the strongly articulated cadentialarrival, the long-awaited descent to 1̂, and the end of anotherthree-measure phrase, which balances the equally expanded[a]-centered phrase of mm. 5–7.

The [ʊ] vowel of “Durch” and “Brust” is what dominatesline 5, at least as Schubert reads it (“Die belauschen sie mitLust”). The poem is structured such that the rhyming words“Brust” and “Lust” occur across a syntactic break. Schubertreinforces this break with the famous shift to G major and thefull measure of rest before the singer re-enters, but he alsoallows the lines to remain fused together sonically by throwingthe musical weight on the end of line 5, and its rhyming word“Lust,” rather than on the beginning.14 Plus, he repeats theline, giving even more emphasis to the word. The dotted halfnotes on “Lust” are the longest in the passage—three times aslong as any other—and the first of them is the high point of thephrase, a note that seems to require a dynamic intensification,in part because it reaches suddenly above the stable platform ofB, where the melodic gesture begins. These notes sound espe-cially drawn-out due to the slow pace of the harmonic progres-sion (forming, as Carl Schachter has noted, the end of a stretchof six measures of just one chord per measure);15 the accompa-niment provides no harmonic energy, no real interest of its own,to buoy the long notes of the melody. For all of these reasons,the [ʊ] vowel of “Lust” becomes the focal point of this section,rather than the first stressed syllable of the line, “be-lau-schen,”a diphthong whose target sound is the very [a] vowel that domi-nated the first part of the song. By this point [a], in short, hasgiven way to [ʊ], and the phonetic shift underlines the tonalshift.

The modulation back to B major, one of the most breathtak-ing in all of Schubert, would also not be nearly as affecting wereit not for the vocal sounds associated with it. The line that coin-cides with the return to B major—“Kehre wieder, holdeNacht!”—returns to a variant of the opening words of the poem(“holde Nacht” rather than “Heil’ge Nacht”).16 Schubert capi-talizes on the poetic device by using an analogous musical

Youens (2002, 86). The vowels are not identical—the stressed vowel in “sinkest” is short and

open [ɪ] and the vowel in “nieder” is long and closed [i]—but they aresimilar enough to be considered assonances.

Youens (2002, 87) does note that Schubert makes the adjectival word“Heil’ge” into “something extraordinary,” though she comments neitheron the assonance in the opening line nor on Schubert’s musical treatmentof it.

Ibid. (89).

If Schubert downplays the [a] of “belauschen” in favor of the [ʊ] of “Lust,”he downplays the [i] of “Die” even more. It is easy enough to imagine asetting in which “Die”—the stressed syllable in the poem’s trochaic meter,a pronoun that refers to the “Menschen” who experience the wonders ofnight—is emphasized more than the second syllable of “belauschen”: if“Die” were placed on the downbeat of m. 16 and the rest of the measurewere syllabic rather than melismatic, “Lust” would still arrive on the down-beat of the next measure. Elsewhere Schubert does stress [i] and [ɪ] vowelswith downbeat placement (see mm. 7, 8, 10, 13, and 24), but to my ears [a]and [ʊ]/[u] predominate, in part because they are given so much agogicemphasis.

Schachter (1998, 216). Some editions of the score, as well as recordings of the song, have the pen-

ultimate line as “Kehre wieder, heil’ge Nacht,” but the Neue Schubert-Ausgabe has “holde Nacht,” and I have taken it as the authoritative version.

()

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

device, returning not just to the original key but also to themelody of mm. 8–9.17 Yet the full impact of this moment hasas much to do with the preceding line, “Rufen, wenn der Tagerwacht.” Consider again the sounds of the words themselves,setting aside for the moment what they mean. The vowel soundof “Rufen” resonates with that of the previous word, “Lust,” aconnection reinforced in Schubert’s setting by the fact that thesinger returns to the same E that was heard on the first state-ment of “Lust,” in m. 17.18 More important, the line charts areverse course from [u] to [a]. Its four stressed syllables outline aprogression from the closed vowel of “Rufen” through thesemi-open vowel of “wenn” to the open vowels of “Tag” and“erwacht,” setting the stage for the most expansive [a] vowelsince the opening measures, the dotted half note on “Nacht” inm. 22.19

As Abrams might put it, singing the line “Rufen, wenn derTag erwacht” and experiencing its gradual expansion of vowelsounds involves a physical component, and the very act of per-forming the line, or imagining what it would feel like to do so,enhances its meaning. The text describes the waking of day andthe poetic speaker’s plea for night to return, the verb “rufen”suggesting an outcry. Is it too fanciful to imagine the following[a] sounds (in “Tag,” “erwacht,” and “Nacht”) as somethinglike cries themselves, exhalations of awe transformed into excla-mations of pain? In any case, when taken together, lines 6 and 7enact a purely sonorous process—the reassertion of the crucialsound from the song’s opening—that conjoins with a poeticone: the longed-for return of night and dreams.

“ ”

Thomas Hardy’s poem is about a table whose noises remind itsowner of the woman who gave it to him long ago. We neverlearn who the woman is or what the nature of her relationshipwith the poetic speaker is; nor do we (or the speaker) discoverwhat the woman meant when, upon delivering the table, she“looked at me with a thought / That I did not understand.” Allwe are told is that the table continues to creak, and that in doingso it reminds the speaker of the “history” that “hangs upon” it,causing him to meditate on the memories that attach to materialobjects.

Hardy structures his poem around the pivotal word, “creak,”which occurs twice in the first line of the poem and is pairedwith its rhyme, “speak,” in line 3. The poem appears inExample 2; again, I urge you to read it aloud, this time focusingon the consonants. Unlike “Heil’ge” and “Nacht,” whose

sound, however expressive, bears no direct relation to theirliteral meaning, “creak” is plainly onomatopoetic—it soundslike what it represents. Britten emphasizes the imitative qualityof the word by setting it (and its rhyming pair) to crisp, staccatoeighth notes, reinforced by accented dyads in the left hand ofthe piano; Example 3 shows the opening strophe. BarbaraDocherty has written that the word “creak” “releases a mimeticimpulse” in Britten.20 That impulse, however, can be felt wellbeyond the opening stanza, for Britten treats other, less obvi-ously depictive words in the same percussive, almost pointillistfashion: namely, “brought,” “looked,” “thought,” and even“it”—words, like “creak” and “speak,” that end with a [t] or [k]plosive, a consonant produced by completely closing the oralpassage and then releasing a burst of air. The effect is even morenoticeable because Britten repeats the plosive words more thanany other words in the poem: “creak” appears three times in thepoem but nine times in the song; “speak” and “brought” appearonly once in the poem but three times in the song; and“looked” and “thought” appear once in the poem and twice inBritten’s setting. At the same time, he sets words with less per-cussive consonants to lyrical melodies (and does not repeatthem), in order to dramatize the contrast between the plosiveand non-plosive speech-sounds of the poem. The manner inwhich the poem’s consonants are articulated dictates themanner in which the song’s melodies are constructed.

The most fluid line of Hardy’s first stanza, for example, isthe last one—“Of one who gave you to me!”—a line character-ized by consonant types that, unlike plosives, allow for a contin-uation of airflow: fricatives (the [v] sound at the end of “Of”and “gave,” the [h] sound at the beginning of “who”), glides(the [w] and [ j] sounds at the beginning of “one” and “you”),and nasals (the [n] sound at the end of “one” and the [m] soundat the beginning of “me”). These consonants make the line

. Hardy, “The Little Old Table.”

As Youens (2002, 90) notes, Schubert also returns to the D♯ from theopening, “emblematic of Day and B major regnant once more.”

The [u] of “Rufen” is slightly more closed than the [ʊ] of “Lust,” becauseit precedes a single consonant rather than two consonants, but the vowelsare nonetheless closely related.

“Tag” technically uses a longer vowel than “erwacht” (designated [ ] inIPA), but the distinction would be less noticeable in sung German than inspoken German.

Docherty (1988, 46); Paetsch (1998, 546) also notes that Britten uses stac-cato notes to depict the creaky table. Neither, however, comments on theother “creaky” musical and verbal sounds of the song or the more lyricalelements that contrast with them.

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

sound more mellifluous than the first line of the stanza, with itshalting, stuttering quality, heightened by the commas withinthe line, which do not appear in the final line or in any otherline of the opening stanza. The words “Of one who gave you to

me” are easier to enunciate; they roll more naturally off thetongue. Britten capitalizes on this fact by setting the line to anunbroken stream of melody, culminating in a long melisma onthe word “gave” (marked “warm” in the score), which diverges

. Britten, “The Little Old Table,” first strophe.

()

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

sharply from the staccato opening. His handling of phraserhythm enhances the effect, since the music seems to abandonthe notated meter as it abandons the plosive sounds of “creak”and “speak,” with their downbeat-accented eighths. In m. 15the word “speak” arrives in the middle of a measure, whereas upto this point it has fallen on downbeats. This subtle change trig-gers a moment of metric displacement dissonance, where thestressed syllables “one” and “gave” in mm. 16 and 17 sound likedownbeats even though they fall halfway through the measure.The bar line seems to shift precisely when the last line of thestanza appears, and in the middle of the melisma it effectivelyvanishes altogether: the melody now seems to soar above the barline, freely issuing forth like the more continuous speech-sounds associated with it.

The piano accompaniment also highlights the phoneticchange at the end of the stanza. At the very end of the melisma,the oscillating thirds in the right hand give way to a wave of risingand falling thirds, mirroring the broad rise and fall of the vocalmelody as it approaches the cadence.21 Moreover, the percussivedyads in the left hand that surrounded the words “creak” and“speak” disappear in Britten’s setting of the last line. Thesedyads are part of a canon between piano and voice, involving notjust the percussive pianistic gestures but the entire melodiccontent of the left hand. (The arrows in Example 3 show howmelodic material is passed from one part to another.) What thissuggests is a dialogue between pianist and singer, between thetable and the poetic speaker—initiated, appropriately enough, bythe table, whose sounds set its owner’s thoughts in motion.

Britten’s setting of the second stanza is provided in Example 4;it is essentially a variation of his setting of the first and likewisestresses words that end with [k] and [t] plosives: specifically,“brought,” “looked,” and “thought.” Some of these [k] and [t]words fall in roughly the same place as the comparable [k] and [t]words from the opening stanza, so it is only natural that Brittenwould handle them similarly. “Brought,” for example, comes atthe end of line 1 of this stanza, just as “creak” did in the previousstanza, and Hardy also repeats it (albeit straddling two lines ratherthan bookending one); Britten accordingly places the word and itsrepetition on downbeat eighths, as he did with “creak.”

It is where new [k] and [t] words are singled out for specialtreatment, however, that we become aware of how sensitiveBritten is to the fourth dimension of the poem. Take “looked,”in the third line of the stanza, a word that ends not with oneplosive consonant but with two ([k] followed by [t]). It isplosive-heavy, we might say, something that Britten seems topick up on: the word appears in m. 32, and in m. 33 Brittenplaces a piano dyad right where the [kt] consonant combinationlands. In m. 35 Britten repeats the word and follows it with yetanother reinforcing dyad. The reason he is able to highlight the[kt] sound with the piano is that, unlike in the first strophe, the

dyads follow rather than precede the words they emphasize.Earlier it was the piano that led the dialogue, but now it is thevoice, a change that underscores the fact that the speaker is nowaddressing the table and driving the conversation.

The final stanza is the most different from the others, notleast because it contains fewer [k] and [t] phonemes, at least ini-tially. Rather than plosives, the first line of this stanza, “Whoeverowns it anon,” uses fricatives, nasals, and glides—[h], [v], [z],[r], and [n] sounds, the same types of consonants that appearedin “Of one who gave you to me.” (Notably, there are no caesurasin this line, unlike the first lines of the other stanzas; “Whoeverowns it anon” flows syntactically as well as phonetically.) As ithappens, the consonant shift is accompanied by a vowel shift.The end-rhymes of the first stanza (“creak” and “speak” in lines1 and 3, “knee” and “me” in lines 2 and 4) involve “high”vowels, vowels that seem to have a higher “pitch” (irrespective oftheir musical setting) because the tongue is in a high positionwhen we speak them. The end-rhymes in the second stanza havea slightly lower pitch (“brought” and “thought,” “hand” and“understand”), and those in the last stanza are lowest of all(“anon” and “upon,” “know” and “ago”). Over the course of thepoem the vowel pitch gradually drops as Hardy moves steadilyaway from the high [i] vowel associated with the word “creak.” Intwo senses, then (having to do with consonants and vowels), thefinal stanza leaves behind the “creaky” speech-sounds of the pre-vious stanzas.

Britten’s music responds in kind. (The final strophe isshown in Example 5.) No more the percussive eighths of theprevious strophes, no more the canon between piano and voice,since the speaker turns his thoughts from the sounds of thetable to the people who will own it in the future: instead, wehear a steady, chant-like melody that hovers on a single pitch, B(to be sung “sweetly”). This melodic idea returns several mea-sures later for what are undoubtedly the most liquid lines of thepoem—“will never know / What a history hangs,” with their[w], [n], and [h] consonants, the last two of which are evenused in alliteration (“never know” and “history hangs”).22 Thefinal strophe is not, however, uniformly lyrical. The percussivegestures come back after the exhalation of “hangs”—coincidingwith the return of the word “creak,” which Britten states nofewer than five times. The canon between voice and piano alsoreturns, only to dissipate with the most expansive melisma yet,on the word “long.” All of this underlines yet again Britten’ssensitivity to the diverse speech-sounds of the text, his uncannyawareness of the inherent musicality and expressivity of thepoem’s phonemes themselves.

Elissa Guralnick has suggested to me that the piano’s relentless eighthnotes, which run “always smoothly” throughout the song, evoke the soundof a ticking clock, marking the passage of time and underlining the factthat the speaker is quickly losing what little remains of the past.

Docherty (1988, 45) points out that the alliteration in this poem is “delib-erately held back” until this moment. The phrase “What a history hangs,”she says, suggests an “exhalation,” “a precise analogue of the unaccountableyears since the gift was made.” One cannot help but think of Youens, whouses the same word to characterize the opening of “Nacht und Träume.”Both “hangs” and “Heil’ge” begin with an aspirated [h] and an openvowel, both are unusually extended, and both suggest a sigh of wonder (inthe first case at the immensity of the night sky, and in the second at theimmensity of the history that clings to the table).

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

Yet what strikes me above all is something much smaller.Wedged between “Whoever owns it anon” and “will neverknow,” breaking up what would otherwise be a perfectly sensi-ble grammatical construction, are three words, set off by

commas: “and hears it,” a phrase that ends with a plosive [t]and features a high [i] vowel. How should these words beread? You could very well gloss over them, ignoring thecommas and pushing onward toward the verb “will.” Or you

. Britten, “The Little Old Table,” second strophe.

()

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

could do the opposite—pause at the commas and set thewords apart, speaking them intently, so they can really beheard. This is how Britten reads the line. The three words

become three eighths, a rising third from B to D♯ (vaguelyechoed by the B–C♯–D♯ in the left hand of the piano, a whiffof the canon from the previous strophes). In what amounts to

. Britten, “The Little Old Table,” third strophe.

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

a perfect marriage of sound and sense, the music evokes thevery creak that is referenced, grammatically and sonically, bythe word “it.” Thanks to Britten, like the future owners of thelittle table, we also hear it.

As Abrams reminds us, lines like “Rufen, wenn der Tagerwacht” and “Will never know what a history hangs” have dis-tinctive textures and sonorous profiles; they demand differentphysical actions, and those actions are meaningful to readers.I hope to have shown that they are also meaningful to compos-ers and listeners. Composers have the power to highlight,amplify, or wholly transform these meanings when they direct asinger to enunciate the words of a poem in a particular way. Lis-teners have the capacity to pick up on them when they watchsomeone singing, sing along with a song, or even imagine whatit would feel like to perform a vocal line.

Let me close by suggesting a few possible avenues for futureresearch that might open up additional meanings:

1. I have focused on composers’ settings of pre-existing texts.What might be learned about the relation between musicaland poetic sound by looking at collaborations between com-posers and poets/lyricists—say, Britten and W. H. Auden,or Hanns Eisler and Bertolt Brecht, or Richard Rodgers andOscar Hammerstein? What about cases where composerswrite their own words?What about popular music?23

2. Some poets are known for their inventive use of sound—Keats, for example, who happens to be one of Abrams’sfavorites, or Gerard Manley Hopkins. What opportunitiesand challenges do poets like these pose to composers?

3. In both of these pieces, the sounds of the words and themeaning of the words seem to be of equal importance to thecomposer. But in other works the balance would seem toshift, such that sound predominates over sense. JohnAdams’s Nixon in China comes readily to mind, as does thefirst movement of Harmonium, “Negative Love,” whichbegins with seemingly meaningless phonemes—“no, no,no, . . .” and then “ne, ne, ne, . . .”—that we only laterrealize are related to actual words in the poem: “no, no, no”prefigures the poem’s crucial line of negation that appearsthree quarters of the way through (“To all, which all love,I say no”), and “ne, ne, ne” merges seamlessly with the firstline of the poem (“I never stooped so low”). Works like thesemight be ideal testing-grounds for an analytical approach thatfocuses on the enunciative component of language.

4. I mentioned the shift from high to low vowels in “TheLittle Old Table.” How does vowel pitch correlate withmusical pitch in general? My sense is that the two oftenexist in an inverse relationship, with higher notes set tolower vowels (as with [a] vowels, which some singers find

easier to negotiate in high registers than other vowels).A study of melodic high points (along the lines of whatKofi Agawu has done24) that brings vowel pitch into theequation might yield fascinating insights into the commonmelodic/poetic patterns in the vocal music of certain reper-toires, or certain composers, and the reasons why somehigh points seem particularly suited to the sonorousness ofa line of text.

No matter our answers to these questions, we are wise to pursuethem with a recognition that the words of a song are utterances,in the simplest sense of the term—material things as well as sig-nificant things, collections not just of signs but of sounds, whichwe perform, and which composers invite us to perform, indiverse and meaningful ways.

Abrams, M. H. 2012. The Fourth Dimension of a Poem andOther Essays. New York: W.W. Norton.

Agawu, V. Kofi. 1984. “Structural ‘Highpoints’ in Schumann’sDichterliebe.”Music Analysis 3 (2): 159–80.

———. 2009. Music as Discourse: Semiotic Adventures inRomantic Music. New York and Oxford: Oxford UniversityPress.

Docherty, Barbara. 1988. “Sentence into Cadence: The Word-Setting of Tippett and Britten.” Tempo 166: 2–11.

Cherlin, Michael. 1986. “Schoenberg’s Representation of theDivine in Moses und Aron.” Journal of the Arnold SchoenbergInstitute 9 (2): 210–16.

———. 1991. “Thoughts on Poetry and Music, on Rhythmsin Emily Dickinson’s ‘The World Feels Dusty’ and AaronCopland’s Setting of It.” Intégral 5: 55–75.

———. 2007. Schoenberg’s Musical Imagination. Cambridge:Cambridge University Press.

Hallmark, Rufus. 2010a. “On Schubert Reading Poetry:A Primer in the Rhythm of Music and Poetry.” In Of Poetryand Song: Approaches to the Nineteenth-Century Lied. Ed.Jürgen Thym. 3–36. Rochester: University of Rochester Press.

———. 2010b. “The Rückert Lieder of Robert and ClaraSchumann.” In Of Poetry and Song: Approaches to theNineteenth-Century Lied. Ed. Jürgen Thym. 335–74. Roches-ter: University of Rochester Press.

Kennedy, X. J., and Gioia, Dana. 2010. An Introduction toPoetry. 13th edition. New York: Longman.

Lewin, David. 2006. Studies in Music with Text. New York andOxford: Oxford University Press.

Malin, Yonatan. 2006. “Metric Displacement Dissonance andRomantic Longing in the German Lied.” Music Analysis25 (3): 251–88.

Paetsch, Annabelle. 1998. “Aspects of Narrativity and Tempo-rality in Britten’s ‘Winter Words.’” Music & Letters 79 (4):538–54.

For an excellent study of alliteration in popular song (and the interaction ofstressed alliterate syllables with musical meter), see Salley (2011). Agawu (1984; 2009, 61–73).

()

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from

Perloff, Marjorie, and Craig Dworkin, eds. 2009. The Sound ofPoetry/The Poetry of Sound. Chicago: University of ChicagoPress.

Pinsky, Robert. 1998. The Sounds of Poetry: A Brief Guide.New York: Farrar, Straus and Giroux.

———. 2013. Singing School: Learning to Write (and Read)Poetry by Studying the Masters. New York and London:W.W. Norton.

Randel, Don Michael. 2014. “Congruence between Poetry andMusic in Schumann’s Dichterliebe.” 19th-Century Music: 38(1): 30–52.

Salley, Keith. 2011. “On the Interaction of Alliteration withRhythm and Metre in Popular Music.” Popular Music 30(3): 409–32.

Schachter, Carl. 1998. Unfoldings: 505 Essays in SchenkerianTheory and Analysis. Ed. Joseph N. Straus. New York andOxford: Oxford University Press.

Stein, Deborah, and Robert Spillman. 1996. Poetry into Song:Performance and Analysis of Lieder. New York and Oxford:Oxford University Press.

Voigt, Konstantin. 2013. Vers und Atonalität: Verfahren derTextvertonung in den frei atonalen Liedern Arnold Schönbergsund Anton Weberns. Tutzing: Schneider.

Youens, Susan. 1991. Retracing a Winter’s Journey: Schubert’sWinterreise. Ithaca and London: Cornell University Press.

———. 2002. Schubert’s Late Lieder: Beyond the Song-Cycles.Cambridge: Cambridge University Press.

Music Theory Spectrum, Vol. 37, Issue 1, pp. 144-53, ISSN 0195-6167,electronic ISSN 1533-8339. © The Author 2015. Published by OxfordUniversity Press on behalf of The Society for Music Theory. All rightsreserved. For permissions, please e-mail: journals.permissions@oup.com.DOI: 10.1093/mts/mtv002

at Acquisition D

ept Serials on August 9, 2015

http://mts.oxfordjournals.org/

Dow

nloaded from