(Ger. Textunterlage). In the notation of vocal music, the alignment of notes and syllables; in oral traditions, the relationship of notes and syllables as performed. The term ‘underlay’ tends to imply a notational procedure of placing syllables beneath notes; the reverse procedure, that of placing notes above syllables, is occasionally referred to as ‘music overlay’, and in cases where order of copying is itself under discussion the neutral terms ‘text placement’ and ‘texting’ are sometimes preferred.
DON HARRÁN
Although it pertains to vocal music of all times and cultures, text underlay is most often discussed in connection with European art music, with special emphasis on 15th- and 16th-century polyphony. This is because the most detailed prescriptions for underlay, which form the cornerstone of later practice, occur in the writings of 16th-century theorists. The applicability of these prescriptions to earlier periods, of which the most problematic seem to be the 14th and 15th centuries, has been questioned, and text underlay has been described as ‘one of the biggest problems facing performers and editors of music before 1600’ (Meconi).
Text underlay connects with the topic of text-music relations at large by virtue of the most fundamental of its prescriptions, stated in an uninterrupted line of writings from antiquity on, namely that there should be an appropriate alliance of speech and song (see also Text-setting). Text underlay thus concerns different kinds of accommodation between music and words in matters of both syntax and semantics. Specific relationships between notes and syllables reflect the varying ways music and language settle one another’s demands.
At one extreme there is a text that shapes the music and at the other music that imposes its order on the text; in between there are a great many textually or musically induced combinations. That 16th-century theorists were aware of the relevance of the general to the specific is clear from the writings of Vicentino and Zarlino, both of whom preface their instructions on underlay with chapters on music and speech. They sometimes took extremist positions: Jean Le Munerat described heated debates between grammarians (i.e. humanists) and musicians at the College of Navarre in the 1490s on the primacy of speech over music or the contrary (see Harrán, 1989).
It might seem strange that a topic of such consequence had limited resonance in detailed writings on music before 1500 (with one exception, a folio added to Antonius de Leno’s Regulae de contrapuncto, c1440; see Harrán, 1978). However, general prescriptions on the adaptation of music to speech can in fact be culled from many sources from Plato onwards.
Prescriptions for text underlay can be extracted from the larger, more systematic discussions of several 16th-century theorists, principally those by Lanfranco (8 rules), Zarlino (10) and Stoquerus (15). Zarlino based his set on Lanfranco’s (Harrán, ‘New Light’, 1973) and Stoquerus seems to have based his on Zarlino’s. The number of rules expands considerably when one takes into account specified exceptions. For example, Lanfranco wrote that ‘in mensural music all single notes (excepting, nearly always, the semiminim) carry their own syllable’: from this the rule emerges that in mensural music the semiminim and obviously smaller values almost never carried syllables. About 110 distinct points of procedure in all can be signalled in more that 360 theoretical statements from antiquity to the 17th century (see Harrán, 1986). These include general ordinances for adapting music to speech and more specific rules for accentuation, syntax, elision, the notes best suited to carrying syllables, how syllables should be positioned within the phrase, semiminims (single or in a series) and subsequent notes, dots and subsequent notes, dissonances, syncopations, leaps, repeated notes, repeated motives or phrases, textual replication and, with respect to performance, breathing, articulation and pronunciation.
These rules fall into three broad categories: (1) a suitable relationship between speech and music, as covered by rules which, in a sense, pre-empt and subsume all others, be they for the affective handling of words or for the details of syllable placement; (2) syntax, or the division of music in accordance with larger and smaller semantic units; and (3) accentuation, or the adaptation of pitches to long/short or accented/unaccented syllables or larger portions, among them the ‘sentence accent’ discussed by Stoquerus.
Five basic assumptions lie behind the rules. One is that vocal music as composed and performed should relate in one or more ways to speech. Another is that the structure and content of the text should be audible. There are different kinds of audibility: declamatory writing facilitates the apprehension of words, but their emotive content is sometimes more finely perceived in more florid styles. The third assumption is that syllables should be treated either syllabically or melismatically; it follows that some note values are suited to carrying their own syllable (notes of medium or longer duration) and others to melismatic use (shorter values), and that different procedures of text underlay obtain for syllabic and melismatic textures. The fourth is that ‘accentuality’, or the musical differentiation between emphasized and unemphasized portions of speech, should not be construed as merely an adaption of notes to word stresses, but rather as a principle that operates on varying morphological and semantic levels, each premising its own kind of textual–musical relationships. The fifth is that these same relationships are subject to a principle of ‘integrality’ whereby, in composition and performance, separate demands are reconciled.
Some works exemplify these rules more by breaking them than by observing them. 16th-century theorists formulated the rules mostly in relation to their own music, which, on the whole, illustrates the operability of these rules; just as there is consistency between the various styles of composition from the 1520s on, so there is also consistency between instructions for underlay from the 1530s (Lanfranco) to the 1570s (Stoquerus). However, the same theorists implied that rules changed according to changing styles or practices. For example, Lanfranco spoke of rules for placing words under melodies whereas Zarlino’s rules were for placing notes under words, a distinction which signals different approaches to text-setting: an earlier one in which words often bent to musical (or otherwise scribal) considerations, and a later one in which music was more noticeably shaped by words. Lanfranco, moreover, said that his instructions pertained to masses and motets, but implied different procedures for chansons and madrigals, which lay beyond his purview. Vicentino remarked on the different accentual treatment of different languages. All the theorists intimated exceptions to the rules in specific stylistic or rhythmic situations, and Stoquerus spoke of varying degrees of observance between earlier and later generations, represented by Josquin and Willaert; he gave a set of ‘compulsory’ rules for both and a separate set of ‘optional’ ones for each. The rules, then, provide a framework for addressing problems of note–syllable relations but cannot always be rigorously enforced, since the theorists themselves suggested their adjustment where necessary to musical realities.
As a premise for operability of the rules in earlier styles, one might refer to Stoquerus’s thesis that the rules accord with natural speech habits. Given the age-old connection between music and speech and the fundamental affinity between musical and speech behaviour as cognate forms of rhetorical expression, it can be assumed that certain rules are indeed of relevance to earlier styles. Lanfranco and Zarlino, moreover, exposed their rules by comparison with those for plainchant, noting their similarity. Since plainchant was the basis for early sacred polyphony (e.g. organum) and persisted as borrowed material for later masses and motets, it would seem that a line of continuity can be drawn, in text–music relations, from earlier to later styles. The same can be said of secular traditions: the concern of the composers, in secular monophony, with textual form, syntax and often prosody carried over into its later polyphonic equivalent, if not in all voices then at least in those that carried text.
Early polyphony raises fewer problems of text underlay for the modern editor than does the music of the 15th century. As Edwards has put it (1987):
we need only go back to around the middle of the 15th century to find ourselves at the end of an extended era when composers wrote many songs in which scope for ambiguity in text underlay is minimal; enough to reveal an extraordinary long-term consistency of texting practice and to provide a more than adequate guide to what is required in those songs which might otherwise land us in difficulties. Moreover, the earlier sources, while by no means free of ambiguity and error, characteristically show a concern with the correspondence between notes and words that is quite lacking in their later counterparts.
The problems that beset the editor and performer of music from the late 14th century and the 15th are due to the often haphazard inscription of words in the sources, not to speak of textual variants. As preliminaries to their solution, the sources (including literary sources of the texts, if available) have to be ranked and collated; decisions must be made about which part or parts were sung with text (the appearance of a textual incipit in a voice may not necessarily mean that full text should be added) and about the treatment, within a phrase, of textless initial, medial or final portions: these may be performed on instruments, vocalized to a single syllable, or texted; and the text itself must be completed or corrected when marred by scribal negligence or inaccuracy.
Manuscript text placement is often complicated by the profusion of melismas. It is not that the melismas were lacking in earlier polyphony; however, they had usually occurred in initial or terminal positions. In the late 14th century and the 15th it seems that melismas invaded the phrase in all its portions; the resultant major difficulty is how to spread the syllables in relation to a phrase’s melodic and rhythmic content. In Kyrie sections of English, Burgundian or early Netherlandish masses, for example, it is often unclear how the words ‘Kyrie eleison’ and ‘Christe eleison’ should be aligned with their notes, and how many times they should be repeated, if at all.
One noticeable difference between earlier and later underlay practices is the often ‘incorrect’ accentuation of words in the former, with weak syllables receiving musical stress (by being placed at the beginning of a tactus or by rhythmic extension) and strong ones ignored. Zarlino, by contrast, prescribed as a cardinal rule the appropriate handling of syllabic stress, a rule that reflected his humanist orientation.
From the casual or even indifferent placement of words or from their faulty musical accentuation it need not be concluded that earlier works were open to as many interpretations as their different versions in the sources, that composers, scribes and singers were uncertain or unconcerned about how the words were to be delivered, or that since it did not seem to matter to them, then neither should it to modern editors. Rather, it may be possible to arrive at a more orderly text underlay by invoking certain larger principles, provided their applicability is sustained by the melodic and rhythmic material at hand (see Perkins, 1974). Because of their generality, the following principles allow ample room for alternative solutions.
(1) syntactic articulation: co-ordinating the musical phrases with larger and, as far as possible, smaller verbal units;
(2) initiality: it is usually safe to assume that the first note of a phrase will carry the first syllable of a word (in practice, however, syllables may need to be extended across phrase boundaries);
(3) finality: the last syllable is likely to fall on the last note of a phrase, though a terminal melisma may force it to occur earlier;
(4) ligature groupings: there are two kinds of ligatures, those explicitly notated and those implied by musical design; in both kinds the syllable usually falls on their first pitch only, though ligatures sometimes have to be broken;
(5) successivity: for purposes of audibility it is preferable, when feasible, to present syllables in close succession, leaving the last accented one to be delivered on the notes that remain;
(6) accentuality: long notes suggest an adaptation to stressed syllables and shorter notes to unstressed ones. But other kinds of accents should be considered, namely the rhetorical accent, which implies its own patterns of stress; and the musical accent, whereby certain words are highlighted for inherently musical reasons (see Harrán, 1986);
(7) verbal substantiality: where possible, significant notes (tonally or rhythmically accented or melodically extended) should be associated with significant syllables (belonging to nouns, verbs, adverbs, etc., but not to conjunctions or to definite or indefinite articles);
(8) linguistic consistency: different systems of accentuation apply to different languages, and the same language may be pronounced differently by speakers of different nationalities;
(9) textural demands: syllabic writing premises a tighter co-ordination of musical and verbal elements than does melismatic writing.
Other difficulties in earlier music, with no easy solution, may have to do with a piece of music’s artificial association with a certain text, as happens in contrafacta, where a new text, with its own syntax and accentuation, is applied to a melody that originally carried a different text; or may have to do with the adaptation to music of successive stanzas, whereby grammatical and syntactical inaccuracies are sometimes unavoidable (few sources transmit texts underlaid beyond their first stanza).
J. Quadflieg: ‘Über Textunterlage und Textbehandlung in kirchlichen Vokalwerken’, Kjb, xviii (1903), 95–38; xxi (1906), 197–223
E.E. Lowinsky: ‘A Treatise on Text Underlay by a German Disciple of Francisco de Salinas’, Festschrift Heinrich Besseler, ed. E. Klemm (Leipzig, 1961), 231–51
E.E. Lowinsky, ed.: The Medici Codex of 1518 (Chicago, 1968), i, 90–107
D. Harrán: ‘New Light on the Question of Text Underlay Prior to Zarlino’, AcM, xlv (1973), 24–56
D. Harrán: ‘Vicentino and his Rules of Text Underlay’, MQ, lix (1973), 620–32
L.L. Perkins: ‘Toward a Rational Approach to Text Placement in the Secular Music of Dufay’s Time’, Dufay Conference: Brooklyn, NY, 1974, 102–14
H.M. Brown: ‘Words and Music in Early 16th-Century Chansons: Text Underlay in Florence, Biblioteca del Conservatorio, MS Basevi 2442’, Formen und Probleme der Überlieferung mehrstimmiger Musik im Zeitalter Josquins Desprez: Wolfenbüttel 1976, 97–141
D. Harrán: ‘In Pursuit of Origins: the Earliest Writing on Text Underlay (c1440)’, AcM, 1 (1978), 217–40
L.L. Perkins and H.Garey, eds.: The Mellon Chansonnier (New Haven, CT, 1979), i, 137–48
M. Bent: ‘Text Setting in Sacred Music of the Early 15th Century: Evidence and Implications’, Musik und Text in der Mehrstimmigkeit des 14. und 15. Jahrhunderts: Wolfenbüttel 1980, 291–326
D. Harrán: ‘On the Question of Word-Tone Relations in Early Music’, ibid., 269–89
F. Tirro: ‘La stesura del testo nei manoscritti di Giovanni Spataro’, RIM, xv (1980), 31–70
H.M. Brown, ed.: A Florentine Chansonnier from the Time of Lorenzo the Magnificent, i (Chicago, 1983), 168–80
R. Jonsson and L. Treitler: ‘Medieval Music and Language: a Reconsideration of the Relationship’, Music and Language (New York, 1983), 1–23
D. Harrán: Word-Tone Relations in Musical Thought: from Antiquity to the Seventeenth Century, MSD, xl (1986) [incl. general bibliography]
W. Edwards: ‘Text Underlay in Marguerite of Austria’s Chanson Album Brussels 228’, Muziek aan het hof van Margaretha van Oostenrijk, ed. G. Moens-Haenen (Peer, 1987), 33–47
D. Harrán: In Search of Harmony: Hebrew and Humanist Elements in Sixteenth-Century Musical Thought, MSD, xlii (1988)
D. Harrán: In Defense of Music: the Case for Music as argued by a Singer and Scholar of the Late Fifteenth Century (Lincoln, NE, 1989)
G. Towne: ‘A Systematic Formulation of 16th-Century Text Underlay Rules’, MD, xliv (1990), 255–87; xlv (1991), 143–68
L.M. Earp: ‘Texting in 15th-Century Chansons: a Look Ahead from the 14th Century’, EMc, xix (1991), 194–212
H. Meconi: ‘Is Underlay Necessary?’, Companion to Medieval and Renaissance Music, ed. T. Knighton and D. Fallows (London, 1992), 284–91
W. Edwards: ‘ Parallel Performance Traits in Medieval and Modern Traditional Mediterranean Song’, RdMc, xvi (1993), 27–35
J. King: Texting in Early Fifteenth-Century Polyphony (diss., U. of Oxford, 1996) [incl. bibliography]
L.L. Perkins: ‘ Towards a Theory of Text-Music Relations in the Music of the Renaissance’, Binchois Studies: New York 1995 (forthcoming)
D. Harrán: ‘How to “Lay” the “Lay”: New Thoughts on Text Underlay’, MD, li (1997) (forthcoming)
G.M. Boone: Patterns in Play: a Model for Text Setting in the Early French Songs of Guillaume Dufay (Lincoln, NE, 1999)