The Dialogue between Words and Music in the Composition and Comprehension of Song

What general advice might one give to someone who forms the intention to write a song? Two general strategies suggest themselves. First, start out with a verbal structure of some kind—a theme or narrative embodied in words—and seek a musical structure to fit this verbal structure. Second, start with a melody embodied in a musical structure of some kind, and then seek a corresponding verbal structure to fit this musical structure. Are there principled reasons to prefer one of these strategies or, indeed, to favor a third strategy over both of them? A brief autobiographical reflection might bear upon this question. As a graduate student, I used to write and occasionally perform songs in that genre usually labeled “singer–songwriter.” Drawing a veil of silence over the songs themselves, what interests me here is a phenomenon I encountered in my compositional endeavors. If I began, in line with the first strategy, by trying to fashion an effective verbal structure with some chosen theme, this nearly always resulted in something that did not work when I tried to set it to music. On the other hand, the second strategy was equally unsuccessful if I began by abstracting the broadly metrical features of a musical structure—rhythm, meter, phrasing—and then sought an appropriately themed verbal structure having those features. This puzzled me, since one might think that it is on such broadly metrical features abstracted from a musical structure that “fit” with a verbal structure should depend. The strategy that did seem to work was to begin with a musical structure and then try to match sounds and emotional tones of words to the fullness of the musical structure by, for example, singing random words to accompany the music. Only then could I hope to find a verbal structure, with the right sort of thematic or narrative coherence, that had the necessary sonic and affective qualities to “fit” with the musical structure. If we assume that my experience was not completely idiosyncratic and reflected something about the way in which words and music work together in song, we might wonder why music should have such a priority over words in what might strike one as an equal collaboration between two distinct artistic media. (I return to some of the assumptions built into this way of phrasing the question later.) But, of course, my experience lends itself to a more obvious explanation—namely, my personal failings as a songwriter! For even a peremptory survey of the relevant historical and philosophical literature suggests that others—those presumably having the songwriting talent that I obviously lacked—have experienced no such difficulties constructing successful songs by setting preexisting verbal structures to music. Indeed, the assumption that this is possible is implicit in one of the most popular accounts of how songs have meaning. On what Jeanette Bicknell has termed the “propositional” model of song meaning, “the meaning of a song can be reduced to the meaning of its verbal text.”1 On this model, the task of the musical accompaniment is to furnish the words with an amplified emotional resonance or representational force. In support of the first strategy and the propositional model, we might point to successful songs that were indeed composed by providing a musical setting for a preexisting poetic text. Some of these belong to the folk–blues tradition—for example, Joan Baez's interpretation of the musical setting by R. Dyer Bennett of Byron's poem “So We'll Go No More A roving.”2 Most obviously, however, this model seems to apply to the lieder tradition. Take, for example, the many musical settings of Johann Wolfgang Goethe's Der Erlkönig by contemporary composers, including Carl Loewe and Franz Schubert. To see how musical settings of texts in accordance with the propositional model are intended to work, I shall briefly describe the latter two settings.3 The story narrated in Goethe's poem involves four characters—a narrator, the father, his son, and the Erlkönig (and the horse upon which the father and son are riding). In setting the poem to music, Schubert assigned each character a distinctive vocal range, type of accompaniment, and rhythmic signature, and assigned some characters a distinctive musical mode. The galloping of the horse is implied throughout the work by rapid triplets that mimic hoofbeats, while musical dissonances are used to suggest the child's distress. The piano sustains throughout a sense of eeriness. As the tension in the narrative rises, so does the pitch of the music and of the child's cries to his father. The song ends with a dramatic pause preceding the words announcing the child's death. Loewe's accompaniment is more representationally and musically complex in many ways. He also includes a representation of hoofbeats, using semiquaver groups of six in nine eight time as opposed to quaver triplets in common time. While he also uses distinctive modes and vocal ranges to represent the different characters, the musical setting for the voice of the Erlkönig departs in harmony, dynamics, melodic shape, and time signature from that for the other characters. Another notable feature of Loewe's setting is a striking musical contrast between the speech of the father and that of the son, something that itself contributes to the representation of the wildness of the night and the ghostly presences within it. In its simple form, the propositional model ignores the possibility of playing off the propositional content of a text against the emotional content of its musical supplement, perhaps for ironic purposes. This is one of Jerrold Levinson's criticisms of the propositional model, but his much more nuanced account of the relationship between textual meaning and musical meaning in song nonetheless remains, I think, within the propositional paradigm more broadly conceived.4 In considering how we should understand the relationship in a song between text, vocal line, and accompaniment, Levinson stresses that “much interest in a song resides in how the ‘natural music’ of words and phrases in a given language is fitted or joined to the pure vocal melody composed to bear those words and phrases.”5 He further maintains that, in assessing the success of a song, “the primary relevant standard is fittingness or mutual suitability of the textual and musical aspects, as measured by the worth of the aesthetic whole they constitute.”6 Mutual suitability is a matter of complementing one another in a rewarding way, “a matter of holistic working, not necessarily based on internal matching or mirroring.”7 Levinson claims that normally “the music of a song attempts to reflect or express the tone and content of a poetic text,” as in the Erlkönig example, although we must also allow for other ways in which music can expressively complement a prior text. A more general model of song that can embrace “art music” such as lieder, sophisticated rock, jazz, and Broadway must “think of song as the explicit combination or pairing of self constituted music with its own expressive character, and self contained intrinsically complete text (typically some preexistent lyric).”8 Levinson therefore does not consider how the construction of the poetic text might itself need to take account of the independent demands imposed by a musical structure. There is, however, a serious problem for the propositional model of song meaning, whether naive or sophisticated, if we seek to generalize it beyond the examples discussed thus far. What is striking about the lyrics of many songs widely viewed as successful is the poverty of their propositional meanings. The librettos for great operas notoriously cannot stand alone as texts, and the same applies to the lyrics of many songs we cherish. For example, those who are moved by a song such as “Hey Jude”—frequently voted one of the very best modern songs—are presumably untroubled by the semantic infelicities in its lyrics.9 Or contrast the emotional power of a “standard” like “What's New?” with the flatness of its lyrics viewed as text.10 Bicknell characterizes this phenomenon as follows: “Songs are a of and as such are to the and of texts to be a of and in To there are to these as not by the lieder but also by the of lyrics by and what we must here is how such examples are to what many of the songs we consider most effective to the point in different can the setting of verbal to music in of into I have two is what compositional strategy should one if one to a song that I possible and on their The other is how should we understand the of words and music to the in a of and how should we why many successful songs seem to primary in this is to that the by these two be is of the ways in which words and music work together in the of a song that the most successful general strategy of song the music the words the we understand why this is the we can also in the of it to the that we can equally by with a preexisting text. In these I upon a for about how artistic meaning, and the meaning of more is to be we begin by a different what of is a song? seems to that two distinct verbal and one necessary of a This suggests that we should think of song as a form, and of artistic song as a that in a way distinct from the and musical of is what Levinson a as an from the combination or of Levinson the that song is a form, that music itself have out of The latter is therefore viewed as a having words and music as its two Levinson his of however, it seems to have a to forms that in their features. for from the historical combination of distinct as what I have termed need to briefly this in to consider its for like that have language as one of their we a artistic content by of an artistic is a that or the in that In the of a for example, the be a on the of a we of different in that in of its it is a with a upon which of for be we from that might the of upon the The very same could apply to a by when we describe the artistic of a work of we normally that the we take to have in its to this of for example, of and the in of an artistic that a of is its an artistic that we the artistic to an artistic content that it is for example, on a that in of that are in the artistic or of the While this is usually in of that as the for a can be for that a A for example, can be both in and in of the way it has to a to of those who with the an of such we can about and be here to consider which as the for in both artistic and verbal example, this no to be those verbal that have such some are intended to be of and to be of should we understand such between verbal I have that we understand these of between in general in of how they are intended to content they have on their primary in general from other in the in which the of that the artistic is intended to in the of is this that a artistic In of these distinctive ways of for a distinctive of on the of the To a by for example, a different of from that to a First, the to the relevant of an artistic is necessary if we are to the content artistic to some of their many of the different of the to the of and the not a of distinct but also does so in a to an of the this to they are to be from other verbal not in of their distinctive or their distinctive or but in of in of their they are intended to as for the of The more features distinctive of are the content is they can as such given as to how the text is to be are by the and must be by the if is to the The between and on the other hand, is a in the of of the verbal to which we must in this general way in to the content those the of structure, the of which the of and so how the same of words might to content in both and and in the of of and of A is a from The of over of that one might have in a in the The the of the I I my I to a to what with the he To some He I was have and a He I I bear to no more I we this of words in we assume that it is intended to in distinctive for example, that the words have and and and that infelicities are not to be as they might be in More in of of the of the we take the of words to have chosen to its its its and its of In this the words are not words have these as But, in these the the thematic meaning of the poem reflected in its from and that in contemporary the of and and so while the words of the are not to its we with the same of that we to the of the example, the is the out of the In this has the to the that as textual for it out between with other in the and so of which as take to be relevant to of the The same in a not for such indeed the same words in a work of such as a they for a different of from these words in the can that while and and verbal a they in that and their artistic by of different artistic are with of even in the latter the can the different that what we in in and in this in return to question about the of song as a As we whether or not something is a in Levinson's sense upon whether it from two preexisting artistic media. we can a more in a given a given time of their task in of two or more distinct artistic demands must be in a If they we such an I even if it is not a form, is in the that the about its But, to understand a or form, we must be as to the of the distinct artistic that it This is more it might as we need to in the between and artistic media. Two forms a both the same of in their artistic in their artistic media. this is possible the of a given of can be used in different ways to As we the for both and is but different of language a in the of artistic content in the two something that is in the that we are to to different of their The point in the is that words and sounds are the for song, not their artistic media. To understand song as a form, we need to whether the artistic or in of which we should to the verbal in a song is the same as that for the or for uses of we need to the same question for the musical artistic or Levinson a question about music in the of dramatic are the musical qualities of music to be in settings it is intended to not but also the qualities of the music be to the of the Levinson claims that, in dramatic song, the should start with musical of with the representationally and it in the of musical the dramatic must begin with of musical and within so as to representational He thus a priority to music over the representation of dramatic in dramatic He also suggests that, in a the musical in dramatic song a different artistic from music, since it for a different of to the of the musical He of the different or that are in when to music that is to be of of or of to of and so be musically by the when they not be in the I think that we can on about song by considering an song in First, the of song, no the of dramatic song, be to with of musical and within so as to representational Second, the verbal in song to the of song meaning an artistic or distinct from those in in and and in as of and other from not as failings in the music used in dramatic song, so the that most song lyrics from in in a of and in not as failings in In both this is these are not failings given the artistic in such In the of this I some in support of these and also what is on in those examples that seem to fit very the of sophisticated propositional model What more general in forms by to which their is to be who this question was In a much he that the must as a is “much he to assumptions about and in the But, in on such have the general in which he as a This general I think, be to to the in of his more general assumption that the of in for is a matter of in a about a that in a The for forms can then be in of on successful so He a of general in and other First, he maintains that are always to more for a given the combination of different artistic is it the of Second, he that a successful must if different artistic are in a work, there must be an between these in the the in a work cannot be the of the cannot be a between and sounds for a as an of expressive For this to be the of the work must be in each must as and complete This is what he the of the different that for expressive must distinct to the expressive of the we have the aesthetic of that we take as What might this about song in general as a First, language and music must work a common expressive Second, their to this must be we should prefer pure music or pure sounds if are always the must be the of what is in is, the in which a is there is also a feature of song, it from other In the the distinctive expressive of the different artistic are by distinct and sounds in the of and speech in the of in the of song, the expressive of the text and the vocal line, to Levinson's are of by the same of the In singing a song, the voice is to expressively both of the sense and resonance of the words itself an with expressive to those of other This expressive of the voice in song applies even in those that seem to fit most to a simple propositional for example, the different to the of the father and son in Loewe's setting of Der is therefore simple in such a to see the intention to be to complement in musical the preexisting narrative content of the poetic expressively the different the experienced by the characters, the tension in the narrated the and so we might the demands on the voice by this expressive to how the words that feature as in the artistic to the of song meaning. If this suggests that words in song belong to a distinct artistic or A further to the simple propositional of what is on in a like the settings of Der Erlkönig is that the text is itself a work of As it does not work, verbal as a simple but also uses language for expressive purposes. If the music does complement the poetic text in the way that the propositional of musical meaning is there not the of or, indeed, of the expressive of the poetic wonder whether the poetic text it is used as the verbal for a song, to as a poetic text, or whether it contributes to the of content a distinct artistic having the verbal text as its In a very for such a He ways in which the of meaning in poetic texts is by of musical in the lieder tradition. He claims that in a this is we take as text a preexisting poem or something to language in the meaning of so many of the The of words is an obvious but music also the of verbal with its implied to a song to its preexistent text, it be to the song as a work of that to the same words as the he is setting of Goethe's text includes the as follows: that the first two are the third line, on the seems and the even both and it is a be to the first two by the to the and and the of their my is my on the other hand, are my is the the song ignores of these music to the way we is the that are and are the and the when the are not when there is He further that is a to a vocal setting to the way we or a text, language a as as music is I that, language as one of the for it contributes the of the artistic or meaning in a different from language in other set poetic texts to music, song meaning is in of a different artistic from that by the we to the words of a song, it is different of those words that a in the of I take the of examples by to support this a preexisting text is set to music, features of the text that are when it as a poem and in the interests of the artistic content to be by the obviously more to be about the distinctive features of the verbal artistic of For example, as the by the text of a song seem to a in the artistic interest and the of music, have to their with and it that used to the of as we the philosophical in or the autobiographical the of to of or the of to of The To this that we upon that are necessary for an account of the and of more First, of the of an bear upon its and in that this content an artistic This the of of interest in musically of or of the and of in and the of and Second, of not so much upon the content an artistic as on the of an having that as perhaps in the of and this is with the that the content by the words of the song itself upon an artistic different from the one if these words are in a Indeed, it might also be that the words used to the artistic content of a song in of one artistic are also intended to a further content of the work the song in of another artistic we of a more general about the artistic which language contributes the of song when poetic texts are set to music, the words not their but the in of which they normally to the propositional content of the in which they are seem to be in This can be if we about those texts that seem to songs when set to music, and that therefore seem to support the propositional account of song meaning. As we in the artistic of and in a way that does of those broadly sonic that words can be as with sonic both to the as in and to one as in and their metrical and those qualities that give an affective In these work together with the sense of the words in a way by the of the The that have set to music seem to be that have so as to upon these broadly sonic that words can with music. Indeed, many of these were based upon songs or song For example, Byron's poem seems to have in by the of a song as that was in in one Goethe's Der Erlkönig it intended to be as a song, which why so many of his in to Loewe and sought to set it to But, even in the of such suggests that, when they are set to music, the of the words and their are from the structures in which they are and further from the way in which language in the artistic or of This the that these poetic texts are with the of that can be into an with the of which music its to return to I think we have the of an of why texts that seem to when as can as The verbal in such the right of the the affective and metrical of the expressively with the in line with Levinson's we to itself work when with words in those that strike as in not so in songs the artistic in song from these to on a of for like my there indeed be reasons why the best way to a successful song is to start with a musical structure and seek a text that work with that structure in the distinctive artistic of song, the and affective qualities of the words are if one to start with a poetic text that is with many of the qualities that bear upon the verbal of artistic content in

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.