(Gk.: ‘song’).
A group of more than five or six notes sung to a single syllable. The term may be applied universally, but has been most used in reference to medieval European music, particularly chant. ‘Melismatic’ indicates one end of a spectrum; the other is ‘syllabic’, or one note to each syllable. An intermediate category, with several notes to a syllable, is sometimes termed ‘neumatic’.
3. Melisma in early polyphony.
RICHARD L. CROCKER
The word ‘melisma’ existed in ancient Greek but was not much used; it meant, vaguely, ‘song’, and conveyed none of the sense given above. (Liddell and Scott’s Greek–English Lexicon cites one instance in which it meant ‘lyric poetry’.) The current technical meaning seems to have been superimposed upon the word by German scholars in an effort to create a term for what was a puzzling feature of Gregorian chant. In the latter context the term has been used in two ways that are distinct but closely related through a characteristic train of thought. The meaning that is logically first is abstract and generic, easier to grasp in German than in English: das Melisma – ‘a manner of singing’, eventually ‘an ethos of singing’. An analogous usage in English would be ‘melody’, as in ‘melody is a basic element of music’. A desire to be more empirical, positivistic, seems to have led directly to the second usage, in which ‘a melisma’ meant a specific instance; the analogy would be ‘a melody’. The process of reification involved the search for origins, archetypes, prototypes, pure forms – an Ur-melisma. Since very few of the latter are documented, even in chant repertories (to say nothing of earlier times and styles), the search led to hypotheses about things very old (ancient Hebrew) or far away (the Near East), and also to magic and other obscurities.
The first meaning – melisma as ‘a manner of singing’ – is more fruitful: evidence of melisma in the works of composers such as Leoninus (c1170), Ockeghem (c1470) and Handel (c1740), to name but three, provides ample support for the use of the term in this particular sense. Although there is almost no instance of melisma among the few surviving fragments of music from Mediterranean antiquity, it cannot necessarily be assumed that melisma was not used then; careful stock needs to be taken of the kinds of pieces that are preserved, for these may reflect the interests of the preservers rather than musical practice. For example, the literary critic or literary historian – in modern times as in antiquity – is traditionally interested in the words, and in the melody only in so far as it has words; that in itself possibly explains why virtually none of the extant fragments shows a developed melismatic style. And since music in general was not written down, there is plenty of room for assuming that at least some of what was not written down could have been melismatic.
St Augustine, in a famous passage written about 400, commented on the expressive essence of the melisma (‘jubilus’ or ‘jubilatio’ in his terminology): he felt that singing without words expressed a joy too deep for words. His references to jubilus have been studied by Moneta Caglio, who reported that in no instance did Augustine associate the jubilus with the alleluia; he ventured the hypothesis that Augustine was connecting the gradual responsory of the Mass to a secular melismatic manner of singing called ‘jubilus’. (Wiora explored the same phenomenon.) This supposed connection to the gradual responsory was disputed by McKinnon, who regarded the reference to the secular jubilus to be one purely of exegesis. For present purposes, however, it is enough to see that there was some such secular practice, and that it could well have been a purely Latin one – although other melismatic practices further east (or for that matter anywhere) are not excluded.
The earliest recorded repertory of melismatic chant is the Gregorian, whose musical documentation exists in the first fully notated chant books (c900). The dates at which this repertory was developed are at issue, now more than ever. Only some Gregorian genres are melismatic, and the order in which the genres, or the individual chants, were developed cannot be determined. It is conceivable that melisma was one phase in a multi-phase development; or, on the other hand, that the various genres approached a melismatic phase at different rates and in different degrees; or yet again, that melismatic style was in no way correlated with the development of repertory. Progressive elaboration from one pitch per syllable through several pitches per syllable to true melismas of ‘more than a few’ (five to ten, or more) per syllable is a developmental model commonly assumed; but while it can be observed to operate in certain limited cases (the motet from about 1220 to 1350 for example), it cannot be relied on over longer stretches, or applied to more than one specific genre at a time. And on the other hand, the traditional assumption (made by Peter Wagner and others) that melisma – hence specific melismas – was ‘archaic’ is untenable; melisma, as melisma, can be newer just as well as older.
Much has been made of a presumed coordination of melismatic style and liturgical function. The melismatic chants at Mass, the gradual-responsory and alleluia, are classed as responsorial chant and sung as independent items between Epistle and Gospel, while the introit and communion, classed as antiphonal chant and sung as accompaniment to liturgical movement, are not classed as melismatic. Some genres, however, are anomalous: tracts, sung after graduals, are not usually classed as responsorial, and those in mode 8 do not have many long melismas; offertories, long regarded as antiphonal, are now rather considered responsorial, but show big melismas only occasionally, thus more often resembling introits – as do responsories from the night Office, which are certainly responsorial in form and function. Even the gradual, the archetypal melismatic form, sometimes shows few larger melismas in the respond. If the simple definition of melisma (‘more than a few pitches per syllable’) is taken, the difference between responds and introits seems to involve merely the more frequent use of groups of ten rather than five pitches per syllable; but this is simply a location on a numerical continuum – in itself a poor basis for a categorical distinction. (There may, however, be a more significant difference in style between the groups in introits as opposed to those in graduals.)
The most distinctive aspect of melismatic style in graduals is the careful avoidance of clearly repetitive pattern (with one kind of exception to be discussed), or, indeed, of any melodic design that would make the melisma seem regular or predictable. The ascents and descents, number of pitches in successive groups, and – in manuscripts that indicate lengthening – the placement of longer values all bring an avoidance of pattern so persistent as to suggest intention. This avoidance of pattern has often been taken to be the product of a way of singing in which the kind of pre-arrangement associated with written records – ‘musical composition’ – is lacking; instead it is regarded as the product of something variously called ‘improvisation’ or, more carefully, ‘oral composition’. The pre-history and very nature of melisma has sometimes been characterized as without pre-arrangement, and attempts have been made to describe its creative process in greater or lesser detail. Without documentation of the musical result, however, all such characterization and description is only imaginative; and when documentation is available, about 900, it tends to show something quite different. When a standard melisma is to be used in a certain gradual, for instance, it is often not written out in full in the earliest sources, but instead only cued; however, its use is confirmed by other manuscripts, so there can be no question but that the melisma was known by heart and reproduced exactly. All the details of melismatic composition that have been pointed out by advocates of ‘oral composition’ can just as well be explained as the products of pre-arrangement, and no essential difference between composition with or without written records need be assumed.
Persistent variety and avoidance of repetition in Gregorian melisma has one curious consequence of great importance. Often the repetition of a motif (a few pitches) is initiated, but then immediately modified (aa') to such an extent that the repetition is not apparent (and sometimes detectable only by close study of the 10th-century notation). Occasionally, however, such a repetition is exact enough to be obvious; examples can easily be found in the verses of the Christmas gradual Viderunt and the Easter gradual Haec dies. In such obvious instances the variety – still present, as in all Gregorian – is provided by the different way in which the repetition continues, making the motivic plan aab rather than aa'. This became a distinct manner of melismatic composition in several subsequent genres.
The Byzantinists have made it clear that for them melismatic composition is a later phase in the development of Byzantine chant, to be placed in the 12th century and after. Even granting our ignorance of the development of Gregorian chant before 900, the history of Latin chant can in this case be read parallel to that of Greek chant. The Gregorian alleluia, whose development can be traced directly in the documents starting in 900, shows a steady progression from modest beginnings to highly melismatic style, and this progression can be assigned to the period between the 8th and the 12th centuries. Such a burgeoning of melisma can also be found in the Mozarabic repertory, in the genres of laude, sacrificium and sono in the León Antiphoner; and while the roots of that repertory have been shown to date from before 700, the extreme melisma found in León may represent 9th- and 10th-century development, and could in principle be coordinate with the Gregorian. In the same period Frankish sources provide instances of an expanded melisma after the verse, called a sequentia, which follows the verse of the alleluia at Mass. At Milan these expanded melismas, called there melodiae secundae, assumed giant proportions (with excessive use of the aab plan for motivic detail). Moneta Caglio has placed some of this development as early as the 10th century, but Hucke argued for a later date. All of this melismatic development can be imagined as analogous and roughly synchronous.
In early medieval chant melismas tended to acquire stereotyped melodic characteristics; as a corollary, they were often detachable entities and could be inserted into or removed from a chant, or transferred to another chant. The most famous instance is provided by Amalarius of Metz (c830) who referred to a melisma sung at Rome in the responsory In medio but transferred by Frankish singers to the Christmas responsory Descendit. This melisma has a melodic shape popular at that period: a dramatic rise by leap immediately repeated and followed by a longer stepwise descent often in three-note sequences (e.g. c–b–a, b–a–g etc.). The 8th-century Gregorian alleluia also shows these melodic characteristics. A special set of melismas accompanied the enēchēmata (see Ēchēma). (Huglo has shown that the corresponding Byzantine models again lacked the melismas.) These and other melismas were often called ‘(p)neumae’ in the Middle Ages.
The Descendit melisma and other similar ones were sometimes provided with texts set in syllabic style. Some scholars have concluded that such a verbalization of the melisma is a basic step in the development of European music. Examples exist, however, to show that within post-Gregorian and medieval chant melismas were cultivated even more intensely than before.
Frankish Kyries and sequences of the 9th and 10th centuries need to be considered in this same syndrome of musical activity, but carefully. The melodies of Kyrie and sequence (as they appear in Frankish sources from the mid-10th century on) are both notated in two forms: one form has the melody in syllabic notation with Latin words underneath; the other has the melody in melismatic notation with just ‘Kyrie eleison’ or, for the sequence, ‘Alleluia’ as an incipit. It is a difficult and controversial question as to whether the melismatic notation in either case was sung as a melismatic piece with only the words ‘Kyrie eleison’ or ‘Alleluia’. But in terms of melodic style, Kyrie melodies as well as sequence melodies are to be analysed in terms of the kinds of purely musical considerations that prevail in melismatic chant. From an analytic point of view the Latin words of sequence and of Kyrie verses can be understood as articulative, both for the rhythmic detail as well as for the larger phrasing. The couplet phrasing of the sequence can, if desired, be seen as the systematic use on a large scale of the aa' repetition of melodic motifs observed in the melismas of gradual and alleluia; but the scale is so much larger that the comparison is perhaps not credible, and in any case it may be better to treat the repetition in sequences as a function of abstract musical design. The sequence melodies as well as the Kyrie melodies have the effect of a new and different style on the Carolingian scene.
Newer forms of medieval chant, such as the Aquitanian versus, also contain melismas which often appear in well-defined functions at the beginning or end of sections in otherwise basically syllabic or neumatic settings. A remarkable and characteristic melisma occurs at the beginning of the famous Alma Redemptoris mater (dating apparently from the 12th century).
While the earliest documentation of polyphony is confusing because of the several diverse styles and structures appearing synchronously, it is nonetheless clear that the documents do not allow the simplistic idea that polyphony developed by logical steps in a straight line from note-against-note counterpoint over a syllabic chant to complex forms with florid counterpoint to a melismatic chant. (It is, of course, necessary to dissociate the idea of note-against-note singing from syllabic style; the latter is a function of one voice, the former a function of two voices.) Early instances of syllabic chant set in two voices note-against-note are mostly found in the examples of singing in parallels given by theorists from the Musica enchiriadis (c900) on; those examples, which are rightly taken as witness of long-existing practices, are not part of the repertory of composed polyphony. This repertory begins with the Winchester Troper, which provides a variety of chants, including melismatic genres, with note-against-note counterpoint. The Aquitanian repertory, whose early layer consists of monophonic versus, occasionally florid, goes on to develop other versus in florid two-part discant; in these are found, among other possibilities, a syllabic or melismatic lower voice combined with a florid (therefore melismatic) upper voice; but there are also passages in which both voices are melismatic (with the same words) in note-against-note counterpoint; and occasionally both voices are melismatic (with the same words) in note-against-note counterpoint. The prevailing arrangement is florid, with persistent use of melisma of modest length (five to ten pitches for one syllable). This arrangement seems in some way essential to the style of the Aquitanian repertory.
The Notre-Dame repertory brings extreme use of melisma, in two clearly differentiated styles – organum purum and discant. Melismas are most prominent in the organal style, where the organal voice can have 30, 40 or more notes against one in the tenor; but that one was often syllabic, not melismatic, so the melisma in the organal voice was a function of a contrapuntal relationship, not a relationship with the words – which was the original definition of ‘melisma’ in chant. On the other hand, in discant style the melismas in the tenor were set in note-against-note relation with the discant voice, and make an entirely different effect. A final paradox was presented when these discant melismas were underlaid in the discant voice with syllables to produce Latin-texted discant and the French motet. Meanwhile in the conductus, melismas and syllabic passages co-existed in more ordinary combinations.
So while it is not possible to trace simple parameters of construction and development in melismatic monophony and polyphony, it can nevertheless be seen that in the intense development of European musical style from the 8th century to the 13th and beyond, melisma played a consistently leading role. It is clear that in European music after the Middle Ages ‘melisma’ is simply a stylistic option located on a complex network of intersecting continua. Applying the concept in that same way to medieval music would avoid the fruitless mythologies that were once evoked by the term ‘melisma’.
P. Wagner: Einführung in die gregorianischen Melodien: ein Handbuch der Choralwissenschaft, i: Ursprung und Entwicklung der liturgischen Gesangsformen bis zum Ausgange des Mittelalters (Fribourg, 2/1901, 3/1911/R; Eng. trans., 1901/R)
K.P. Wachsmann: Untersuchungen zum vorgregorianischen Gesang (Regensburg, 1935)
E. Wellesz: A History of Byzantine Music and Hymnography (Oxford, 1949, 2/1961)
L. Brou: ‘Séquences et tropes dans la liturgie mozarabe’, Hispania sacra, iv (1951), 27–41
J. Handschin: ‘Trope, Sequence and Conductus’, NOHM, ii (1954), 128–74
H. Hucke: ‘ Die gregorianische Gradualweise des 2. Tons und ihre ambrosianischen Parallelen: ein Beitrag zur Erforschung des ambrosianischen Gesangs’, AMw, xiii (1956), 285–314
E. Werner: The Sacred Bridge: the Interdependence of Liturgy and Music in Synogogue and Church during the First Millenium (London and New York, 1959–84/R)
W. Wiora: ‘Jubilare sine verbis’, In memoriam Jacques Handschin, ed. H. Anglès and others (Strasbourg, 1962), 39–65
C. Thodberg: Der byzantinische Alleluiarionzyklus, MMB, Subsidia, viii (1966)
M. Huglo: Les tonaires: inventaire, analyse, comparaison (Paris, 1971)
E.T. Moneta Caglio: Lo jubilus e le origini della salmodia responsoriale (Venice, 1976–7)
J.W. McKinnon: ‘The Patristic Jubilus and the Alleluia of the Roman Mass’, Cantus Planus Study Session III [Tihány 1988], ed. L. Dobszay and others (Budapest, 1990), 61–70