Text and Tune

On the Association of Music and Lyrics in Sung Verse

by Teresa Proto (Volume editor) Paolo Canettieri (Volume editor) Gianluca Valenti (Volume editor)
©2016 Conference proceedings 371 Pages
Series: Varia Musicologica, Volume 21


This book offers an overview of issues related to the regulated, formal organization of sound and speech in verse intended for singing. Particularly, it is concerned with the structural properties and underlying mechanisms involved in the association of lyrics and music. While in spoken verse the underlying metrical scheme is grounded in the prosody of the language in which it is composed, in sung verse the structure is created by the mapping of specific prosodic units of the text (syllables, moras, tones, etc.) onto the rhythmic-melodic structure provided by the tune. Studying how this mapping procedure takes place across different musical genres and styles is valuable for what it can add to our knowledge of language and music in general, and also for what it can teach us about individual languages and poetic traditions. In terms of empirical coverage, the collection includes a wide variety of (Western) languages and metrical/musical forms, ranging from the Latin hexameter to the Norwegian stev, from the French chant courtois to the Sardinian mutetu longu. Readers interested in formal analyses of vocal music, or in metrics and linguistics, will find useful insights here.

Table Of Contents

  • Cover
  • Title
  • Copyright
  • About the author
  • About the book
  • This eBook can be cited
  • Contents
  • Introduction
  • Poetry and music in archaic and classical Greece. Some thoughts
  • Medieval liturgical drama, Carmina Burana and the Arnaut Daniel’s sestina: Music and literature
  • M’es belhs dous chans: Melody, metre and imagery in a ‘love verse’ of early troubadours
  • Poetic rhythm in musical notations of the 14th century: The amateur tradition of grand chant courtois under the patronage of the autonomous system of ars musicae
  • For the anisosyllabic whim of the Romance Middle Ages: Disciplines and non-regularity in the lyric poetry
  • Creation, appropriation and development of the “sung verse” in the medieval musico-liturgical drama Officium Stellae
  • Textsetting of multilingual poems: The example of Bruder Hans’ Ave Maria
  • Norwegian gamalstev: A millennium of sung verse
  • Verse structure and time patterns in the a mutetus extemporary sung poetry of Southern Sardinia
  • The prosody of Basque songs: A methodological proposal
  • Text-to-tune alignment and lineation in traditional French songs
  • Stress-to-beat mismatches in French rap
  • New directions in Italian song lyrics?
  • Traditional metrics in Javier Krahe’s lyrics: Accords and discords
  • The challenge of identifying vowel phonemes in singing
  • Textsetting in translation: Rhythmical (non-)equivalence in the works of three Scandinavian ‘singer-translators’
  • Three dimensions of singability. An approach to subtitled and sung translations
  • What can the cross-cultural study of children’s clapping games teach us about the universality of sung verse?
  • List of Contributors



The expression ‘sung verse’ covers a wide variety of metrical-melodic forms, which are found across time and space in all human cultures. While one culture may lack instrumental music, vocal music seems to be universal. Around this basic idea, specialists from different research areas – from metrics to anthropology, from linguistics to musicology – were brought together at a conference held in Rome in 2012. Our goal at that time was to encourage an interdisciplinary discussion on issues regarding the regulated, formal organization of sound and speech in verse intended for singing.

A few years later, a number of those papers are collected here. The heterogeneous character of the book reflects the multifarious nature of the subject matter itself, which can be approached from different angles and perspectives. Scholars focused on the historical development of (Western) vocal music and those interested in formal analyses and metrical typology may find useful insights here, for the approaches best represented are the historical-descriptive and the formal-analytical. The first part of the collection includes mainly papers of the first kind, while the second covers more methodological and theoretical issues; however, intersections and overlaps between the two are evident throughout.

Each paper is representative of (a group of) specific issues that emerge when dealing with the formal properties of sung verse, considered simultaneously as a metrical, melodic and textual object. Even in the contributions with a strong historical or empirical slant, attention is always given to the theoretical implications. Conversely, even in the most formalistic approach considerations of historical and empirical importance may not be disregarded. This reflects the twofold nature of sung verse, which is, on the one hand, a cultural product, subject to change and linked to such notions as style, genre, imitation etc.; and, on the other hand, a product of human creativity, whose form is partly ← 7 | 8 → determined by our innate, cognitive systems, which apply restrictions at certain levels of its organization and perception.

The twofold nature of sung verse can be traced not only along the axes of culture and cognition, but also, and more importantly, in its ‘double articulation’ of text and melody. While in spoken verse the underlying metrical scheme is grounded in the prosody of the language in which it is composed, in sung verse the structure is created by the mapping of specific prosodic units of the text (syllables, moras, tones, etc.) onto the rhythmic-melodic structure provided by the tune. Studying how this mapping procedure takes place across different musical genres and styles is valuable for what it can add to our knowledge of language and music in general, and also for what it can teach us about each language and poetic tradition in particular.

In terms of empirical coverage, the collection includes a wide variety of (Western) languages and metrical/musical forms, ranging from the Latin hexameter to the Norwegian stev, from the French chant courtois to the Sardinian mutetu longu. No distinction is made between ‘high’ and ‘low’ poetry, nor between folklore and literature, and the same importance is attached to children rhymes, rap and folk poetry as to opera and Troubadours’ songs. Indeed, such a distinction would not make sense in a book concerned with the structural similarities found cross-culturally in sung verse and the universal mechanisms underlying the relationship between lyrics and melody. Traditionally, folk poetic forms have been studied almost exclusively from the point of view of ‘culture’, as part of a group’s or nation’s ‘folk-lore’. However, recently there has been a shift in this area of research and scholars have started to look at the abstract properties of these ‘cultural objects’ and to point to the common features shared by similar though unrelated poetic forms. A good example is provided by Andy Arleo’s line of research, which tackles the hypothesis of a Universal Children’s Rhythm valid for all forms of childlore across the world. In the exploratory study of children’s clapping songs presented here, the author claims that in different parts of the world clapping follows the same binary patterns. From our perspective, further studies of this kind would be most welcome, in that they would not only improve our knowledge of the forms of sung verse attested across the world, but would also provide a good testing ground for predictions as to which structures are common cross-culturally, and which are rare or impossible. ← 8 | 9 →

Within a single poetic tradition, investigating the formal relationship between words and music can help to unravel the dense network of cross-references and quotations among authors or between author and sources, which would otherwise go unnoticed. Good examples are the detailed analyses proposed by Antoni Rossell and Giorgio Monari for Arnaut Daniel’s sestina and for the early troubadours respectively. By approaching the sestina as an auditorily experienced object, Rossell finds interesting connections between the poem and its sources, both at the melodic and textual level. Monari investigates the distribution of a specific melodic progression, which seems only to occur in restricted metrical contexts across the production of the early troubadours (Jaufre Rudel, Marcabru and Bernart de Ventadorn). According to Monari, the appearance of this melodic phrase in Jaufre Rudel’s songs evokes the general idea of ‘singing for love’. An argument in favour of this interpretation is provided by the fact that the lyrics set to the tune consistently make reference to the act of singing. Moreover, the occurrence of the same melodic phrase in other early troubadours’ songs contributes to shed light on the poets’ views of fin'amors and on the sophisticated interplay of allusions and cross-references that characterizes the production of the early troubadours.

Oliver Vogel approaches another important tradition within the poetic production of medieval France, i.e. the grand chant courtois. His point of departure is the observation that while dance tunes from the 13th and early 14th century (for instance, the rondeaux) were easily re-used in later centuries and became suitable material for other genres of musica mensurata (for example, motets), trouvère songs did not. This difference is convincingly explained in terms of a higher rhythmical freedom peculiar to the Old French metrical line, which could not be easily adjusted to restrictions imposed by mensural notation. In the analysis, specific manipulations of the textual and melodic material are shown to be in place in the first attempts at a mensural notation, in order to counterbalance the loss of freedom. Sometimes, specific musical features can be exploited to underline the meaning of the text. This is well illustrated in Nausica Morandi’s study of the Officium stellae, a musico-liturgical drama performed during the Liturgy of Epiphany between the 10th and the 15th centuries. She shows how existing metrical texts (in Latin and mainly hexameters) can be set to (new or traditional) ← 9 | 10 → liturgical melodies in such a way as to build up the dramatic character of a play. Special attention is devoted to the formal devices and strategies adopted in order to emphasize particular points in the drama, such as the use of neumatic vs. melismatic passages to identify specific situations or characters (much in the same spirit as the later Wagnerian Leitmotiv), or the regular occurrence of the same pitch interval or melodic formula in parallel positions in the text (e.g. on pairs of rhyming words). Melodic ascents and descents are used in a way that anticipates the “word-painting” technique found in later madrigals.

The boundary between metrical and unmetrical can be highly controversial when dealing with written poems that (in all probability) were originally sung, but have been handed down to us without music. Without the metrical template provided by the tune, these texts may still appear – to a certain extent – metrical; however, a closer analysis reveals irregular patterns whose existence is difficult to explain in terms of prosodic metrics. In his analysis of the medieval anisosyllabic ghiribizzo (irregular whim), and based on the example of some Italian and French medieval poems, Fabio Sangiovanni tries to establish a methodology capable of distinguishing those lines that are purposely created as anisosyllabic from those that are made irregular due to corruptions in the textual transmission. Understanding the relationship between ‘process’ and ‘datum’ in anisosyllabic lines would be of help not only for the purposes of textual criticism, but also for historical metrics and linguistics, as it may contribute to a more precise dating of specific graphemic and/or phonetic changes.

A large part of the poetic production of the Middle Ages is interspersed with records of texts which are without musical notation since all performers knew that they should be sung to a particular well-known tune. Lyrics without melodies are not isolated cases, as shown by Patrizia Noel Aziz and Levente Seláf in their paper examining a late 14th century Ave Maria. Here the lack of musical notation is coupled with the complexity of a multilingual tradition, as the poem is glossed in several languages (German, Latin, French, and English). A twofold problem presents itself to the metricist: on the one hand, searching for the musical template underlying the poem(s) may prove fruitless, as hardly any reference is made to the actual melody; on the other hand, studying this material as part of either versification system may be ← 10 | 11 → misleading, for spoken metrical templates may capture some of the structural features of the text, but fail to predict and explain deviations.

Sometimes a musical form or song only survives in oral tradition, with no further written attestations beyond a certain point in time. This is the claim made by Storm-Mathisen concerning the Norwegian Gamalstev, whose origins can be traced back to approximately a thousand years ago. The Gamalstev is likely to be a continuation of the Old Norse poetry preserved in the Edda manuscripts. Its structure is strikingly similar to the ljóðaháttr metre, used e.g. in the Hávamál and the Lókasenna; moreover, their affinity in terms of accentual patterns is such that Old Norse verse can easily be sung to the tunes of Gamalstev. This is another good example of the contribution that folklore studies can offer to metrics. By bridging the gap between oral and written poetry, it can play a central role in revealing the regularities that hide in heterogenous and apparently distant forms.

The object of Paolo Bravi’s research also belongs to oral poetry. The mutetu longu is an extemporary genre performed in the Campidanese dialect of Sardinia by semi-professional poets. Despite the existence of a regula poetica – a set of rules established in the 20th century for regulating the rhyme patterns – , the mutetu seems to lack any clear rules of line structure. However, through a quantitative and qualitative analysis of live recordings, which takes into account both acoustic measurements and considerations of syllable duration, prominence distribution and line length, Bravi shows that not all patterns are acceptable. Indeed, the analysis points to the existence of implicit unconscious models that experienced poets (and listeners) have internalized in the course of their life-long training. Such a finding is also valuable for the contribution it makes to metrical typology, and to debates of a more theoretical nature such as, for example, the distinction between verse instance and verse design. From the point of view of the methodology employed, this kind of research looks promising: the acoustic analysis developed here shows that tools can be borrowed from other areas of linguistics – in this case, the software PRAAT originally created for the study of acoustic phonetics – and adjusted to the needs and purposes of metrical analysis.

The use of acoustic measurements for the study of sung poetry is also encouraged in the paper by Varuṇ DeCastro-Arrazola. Here the main focus is on the methods employed in the data preparation stage ← 11 | 12 → preliminary to the study of textsetting constraints, i.e. restrictions that apply to the setting of a text to a tune. The author argues against relying exclusively on metrical grids à la Halle and Lerdahl (1993) as the basis both for empirical analysis and for building up hypotheses and models. Using the example of a traditional Basque song, where musical pitch apparently correlates with phonological pitch (at least in some positions), it is argued that traditional metrical grids would fail to capture this important property of Basque verse, simply because they lack any representation of pitch. For other types of verse the use of metrical grids makes perfect sense. This is the case for rap and hip-hop which are chanted on a regular rhythm rather than sung to a full melody. Daniela Rossi makes use of metrical grids in order to illustrate differences between the text-to-tune alignment of French rap and that of French traditional song. Her detailed analysis of stress-to-beat mismatches points to systematic violations of traditional textsetting, which can be ascribed both to syncopation and to a different application of the local maximum constraint, as stipulated by Dell and Halle (2009).

By contrasting Rossi’s paper with Dell’s account of traditional French songs, it appears clear that the templates underlying the two singing idioms are quite different. However, the idea, put forward in Dell’s account, that song lyrics do not have an inherent metrical structure, seems to hold true for French rap, where the text looks very much like prose when written down. In particular, the two pieces of information that in Dell’s theory should be incorporated into the representation of the melody, i.e. line-end location and melismas, have a different status in rap when compared with traditional song. Melismas do not seem to occur, and mismatches between metrical and textual lines are so commonplace that they should perhaps be considered as the distinctive mark of rap as a poetic genre, together with complex rhyming schemes.

The proposal made by Dell remains a challenging one, especially since it questions a unified account of spoken and sung poetry. Moreover, the relationship between the text of songs and literary verse within one individual language raises questions concerning their origin and evolution. Are the formal similarities observed between them due to chance, or to imitation, or even to direct/indirect filiation? The role played in contemporary vocal music by the literary poetic models is investigated in two contributions, those of Luca Zuliani and Clara Martínez Cantón ← 12 | 13 → respectively. While the former points to the emergence of a new distinctive “language” in Italian contemporary songs, which greatly departs from the traditional schemes, the latter focuses on the contribution of traditional metrics to the art of Javier Krahe, a Spanish songwriter from the last decades of the 20th century. These studies show two alternative approaches that are typically found in modern developments of traditional singing idioms, in which either an effort is made to maintain the metrical and prosodic patterns established within a poetic tradition, by appealing, for instance, to renowned predecessors (as Krahe does with the poets of the Siglo de Oro); or the opposite occurs, and traditional style is rejected entirely, opening the way for experimentation. One of the reasons for this may be connected to the phonology of the language. For Italian, this is illustrated by the loss of the apocopated forms traditionally used in oxytonic rhymes to supplement the rare oxytones present in the language. Although these forms derive from a compromise between the rigid structure imposed by the traditional melodies and the phonology of everyday speech, in the second part of the 20th century they started to be felt archaic and artificial and eventually became associated with trivial pop songs.


ISBN (Softcover)
Publication date
2015 (November)
lyric poetry Basque Songs French Rap Hans Bruder
Bern, Berlin, Bruxelles, Frankfurt am Main, New York, Oxford, Wien 2015. 371 pp., num. b/w ill., 25 coloured ill.

Biographical notes

Teresa Proto (Volume editor) Paolo Canettieri (Volume editor) Gianluca Valenti (Volume editor)

Teresa Proto is a postdoctoral fellow affiliated with Leiden University and the Meertens Institute, Amsterdam. Her publications reflect her interest in language, music and the interactions between the two. Paolo Canettieri is a romance philologist and researcher in cognitive science. He is full professor at the University of Rome ‘Sapienza’. He is Editor in chief of the journal «Cognitive philology». Gianluca Valenti is a postdoctoral researcher at the University of Liège. His main interests are romance philology, metrics, history of science and lexical studies. He has recently published La liturgia del trobar.


Title: Text and Tune