Lai

(Fr.).

An extended song form cultivated particularly in the 13th and 14th centuries. The stanzas – if the poem can be divided in that way – are each in a different form and therefore have different music. Though the number of surviving examples is small compared with the total extent of medieval song these works occupy a special position for several reasons: the very irregularity of the poetic form led to large metrical and rhyming patterns that have caused the lai and its German equivalent the Leich to be described as the major showpieces of medieval lyric poetry; and there is much truth in Spanke’s useful distinction (1938) between songs that are primarily metrical in their formal concept (i.e. nearly all medieval strophic song) and those that are primarily musical (the lai and the sequence), a distinction that almost inevitably brings with it the suggestion that the lai and related forms represent by far the earliest surviving attempts at continuous extended musical composition outside the liturgy. In general it is true to say that in the 13th century the form could be extremely free, with highly irregular rhyme schemes and lines of uneven length, but that in the 14th century lais became enormously longer, with the French tradition developing a standard pattern with each stanza following a double-versicle scheme (often refined to an apparent quadruple-versicle) and a 12-stanza form in which the first and last could be related musically or even have the same music at different pitches.

1. Terminology and origins.

2. Poetic form.

3. The lai before 1300.

4. The lai after 1300.

5. Notes on the checklist of lai music.

CHECKLIST OF LAI MUSIC

BIBLIOGRAPHY

DAVID FALLOWS

Lai

1. Terminology and origins.

Many different extended forms in medieval (and later) poetry are encompassed by the word ‘lai’ and related words in other European languages. The form described above is more strictly called the lyric lai (or lai lyrique) to distinguish it from the narrative lai (or lai breton), a long poem normally in octosyllabic rhyming couplets and often associated with stories of the Arthurian cycle.

The narrative lai was most elegantly described by Chaucer’s Franklin:

Thise olde gentil Bretouns in hir dayes
Of diverse aventures maden layes,
Rymeyed in hir firste Briton tonge;
Which layes with hir instrumentz they songe,
Or elles redden hem for hir plesaunce.

In French this genre saw its first success with Marie de France, who apparently worked in England and whose 12 narrative lais date from the years after 1160, though she claimed they were adapted from Breton originals that are now lost; but it continued well into the 15th century: after 1415 Pierre de Nesson called his poem in that form lamenting the French defeat at Agincourt the Lay de la guerre. Though the narrative lai, like much other narrative verse, was evidently sung, certainly in its earlier history, no music for it survives: only the presence of empty staves above a manuscript of the Lai de Graelent (F-Pn fr.2168) and a lost manuscript of the lais of Marie de France (see Maillard, 1963, p.66) witness that this music may have been written down. (See also Chanson de geste.)

Even within the terminology ‘lyric lai’ there are poems that cannot be described as lais according to the definition adopted for this article. Several poems carrying the title ‘lai’ are found inserted in longer narratives: particularly famous examples of this category appear in the Roman de Perceforest and the Roman de Tristan en prose. For the latter there is even music, surviving notably in A-Wn 2542 (listed below); but these are mostly simple strophic songs, quatrains with the musical scheme aabc (the three melodies in F-Pn fr.776 have no such clear design), though with the peculiarity that their music is fully written out for all stanzas of the poem. Tannhäuser’s Lude Leich (D-Mbs Cgm 4997, ff.72–73v) is also isostrophic. More difficult to classify are the two ‘lais’ attributed by Beck to Charles d’Anjou and described by Stäblein (1975) as merely a series of single-stanza songs.

The word itself is hard to pin down and has been variously explained. One possibility is that it derives from the Latin laicus, implying a secular equivalent of the sequence. Intriguingly the narrative Palamede says that the lai has its name because it leaves behind (laissier) all other lyric forms. Another theory traces it back to the Low Latin leudus, found as early as Venantius Fortunatus (c550) and meaning a vernacular song in Latin metre: this is probably a latinization of a germanic word (perhaps *leuthaz, though the German word Leich is thought to derive from *laik; asterisks in this context denote hypothetical roots) and is glossed in one manuscript with the Old High German Uuinileodos (singer); the later Latin laudes and the Irish loîd (or laîd) are evidently related. The Irish word (meaning blackbird’s song) can be documented from the mid-9th century and may go back further; since the 16th century it has been used to mean any poem, but particularly songs related to the epic of Finn. There is much evidence in favour of Irish origin for the form, a theory supported by Aarburg and by Maillard, who showed how the rhythmic interest of early Irish poetry may have influenced the lai (‘Lai, Leich’, 1973, pp.326ff); but no music survives to document this and nothing in Irish poetic structure can be linked directly with the musical form that appeared in France about 1200.

The earliest recorded French uses of the word ‘lai’, those of Wace in his Roman de Brut (c1155: ‘lais et notes/lais de vieles, lais de rotes/lais de harpes et de frestels’), describe instrumental melody, but this is by no means the rule for other early references: Wolf (1841, pp.4ff) showed that the word ‘lai’ in English or Old French often meant no more than ‘song’.

Just as the word ‘lai’ itself had and still has several meanings beyond and around the specific musico-poetic form that is the subject of this article, so also many other words cover ranges of meaning that include that of the lyric lai.

(i) Descort.

Derived from the Latin discordia, descort was the standard Provençal word for lai and was carried into Old French as well as into Italian, where the word discordio appears describing a poem of Jacopo da Lentino. There has been some disagreement as to the differences between descort and lai, with Wolf (1841) and Jeanroy (1901) pioneering the opinion that there was none and Stäblein (1975) suggesting that descort was merely a later and more sophisticated name preferred by the trouvères. (An excellent and full survey of the dispute appears in Baum, 1971; see also P. Bec, La lyrique française au Moyen Age (XIIe-XIIIe siecles), i, Paris, 1977, pp.199ff.) For musical purposes the two probably were identical since the confusion appears to lie mainly in the range of other materials encompassed by the word descort: though primarily designating the same form (or forms) as the lyric lai, it is also used for a poem whose stanzas disagree in some other way (e.g. Raimbaut de Vaqeiras’s descort Ara quan vei verdejar has five isometric stanzas but their discordia lies in the fact that each is in a different language) and for a poem whose subject matter is discordance, disagreement or most characteristically severe disappointment in love (this last explaining the definition of the descort in the Provençal Doctrina de comprendre dictatz). In this context it might be worth noting that Konrad von Würzburg’s second Leich (for which no music survives) repeatedly describes itself as a Streit (‘contest’).

It must be added, however, that some scholars see substantial differences between the lai and the descort. Maillard (1963), starting from the nomenclature found in the sources, divided the extant lyric lais into lais and descorts, but his efforts (pp.128ff) to define the difference are not entirely convincing. His suggestion (pp.143–4) that the essence of the descort was in its poetical form occasionally departing from its musical form seems questionable in view of the relative frequency of troubadour strophic songs whose musical form is at variance with their poetic form, and the extremely small number of Provençal lais (or descorts) with surviving music against which the theory can be tested (see also Maillard, 1971). Yet another opinion as to the difference is offered by Gennrich (1932, pp.138ff). Although the matter is still a subject of considerable dispute, this article has been prepared in the belief that these distinctions are artificial.

(ii) Leich or leih.

The Middle High German word for lai. Kuhn (1952) argued that the earlier form was *laik, a dance-song (but see also Carol), and there is much evidence that the word ‘leih’ was originally used to denote a melody: so Browne (1956) gave examples of a sancleich (sung Leich), a herafleih (harp Leich) and a keraleih (sorrowful Leich). By the late 12th century it seems that ‘Leich’ could mean any sacred poem, and the Monk of Salzburg (c1400) used the word to describe his sequence contrafacta; moreover, the word appears in manuscripts of the 12th and 13th centuries as a designation of psalms (D-Mbs Cgm 17 and A-Gu 204), perhaps because King David was reputed to have sung them to the harp (harafleich). But otherwise the word is perhaps the most specific of all those related to the form discussed in this article: it was not used for narrative poems and seems to have had very little currency for other purposes. Since Spanke many German philologists have preferred to form the plural artificially as Leichs (since Leiche, the more correct plural, also means ‘corpse’).

(iii) Note, nota, notula.

One or other of these words is used by Johannes de Grocheio and in several French lais apparently describing lai form. In the title of the ‘Note Martinet’ (the lai J’ai trouvé), ‘note’ may be merely the designation of the melody or perhaps more particularly of the musical and poetic scheme (see also Ton (i)). Maillard cited several instances of its apparently referring to the melody of a lai; but sometimes it meant simply the melody of any song. The word is discussed more fully by Gennrich (1932, pp.167ff).

(iv) Estampie, estampida, ostampida, stantipes, stampita.

The trouvère manuscript GB-Ob Douce 308 contains a group of 19 poems in lai form having the title ‘estampie’, but none of them has music. Among the Provençal repertory, the poem Kalenda maya by Raimbaut de Vaqeiras describes itself as an estampida but is isostrophic with each stanza following the pattern aabb¹cc¹. Instrumental pieces (perhaps dances) of the 13th and 14th centuries have the title and are in double-versicle sections with refrains at the end of each stanza and with ouvert and clos endings. The form is mentioned by Johannes de Grocheio, by the author of Las leys d’amors (1328) and as late as Michael Praetorius; see Estampie.

(v) Ductia.

Another form mentioned by Johannes de Grocheio and closely related to the estampie: it has double-versicle sections with refrains but differs from most such forms in having isometric stanzas; see Ductia.

(vi) Caribo, caribetto (Provençal: garip).

An Italian word used primarily for instrumental versions of the form though also implying dance music. It was mentioned together with the nota and the stampita by the theorist Francesco da Barberino (1310). As a name for a kind of dance-song it appears in Dante’s Purgatorio (xxxi.132); and it describes a poem in lai form by Giacobo Pugliese. There is some disagreement as to whether the word derives from the Greek charis or from the name of an Arabic instrument. (See Appel, 1887, p.224.)

(vii) Consonium.

Mentioned by Francesco da Barberino as the text for a caribus, a nota or a stampita (see Gennrich, 1932, p.163).

(viii) Ensalada, ensaladilla.

Rengifo’s Arte poetica española (1592) described the form in terms strikingly similar to earlier descriptions of the lai and descort. In surviving Portuguese and Spanish music the Ensalada is not known before the mid-16th century; on earlier traces of the poetic form see d’Heur (1968) and G. Tavani: Repertorio metrico della lirica galego-portoghese, Officina romanica, vii (Rome, 1967).

(ix) Sequence, prose, conductus, planctus, versus.

These are the Latin forms that include material that could be described in Spanke’s terms as ‘musikalisch primär’. Ferdinand Wolf’s theories (1841) on the substantial identity of sequence and lyric lai were strongly opposed by Jeanroy and Aubry (1901), who asserted a little too absolutely that all sequences were built on an unvarying double-versicle pattern – a pattern that can rarely be seen in the 13th-century lai. Subsequent studies of the sequence (see Sequence (i)) may not restore Wolf’s original position, but they do suggest that both forms were initially less strict than they later became and that although the surviving history of the sequence’s development belongs to a time nearly four centuries earlier than that of the lai the two have much in common and even share some music. Gautier de Coincy’s Hui enfantés is a straight contrafactum of the Christmas sequence Letabundus which also served as source for the Anglo-Norman drinking-song Or hi parra; and it may be an evasion of the real issue to omit these works from considerations of the lai, dismissing them as mere contrafacta. Godefroy de St Victor’s Planctus ante nescia provided the melody and the form for the French lai Eyns ne soy ke pleynte fu, the English ‘translation’, Ar ne kuthe ich sorghe non (both in Corporation of London Records Office, Liber de antiquis legibus, ff.160v–161v; fig.2) and the Hungarian poem Volék sirolm tudotlan. There are several other such examples in which the Latin original is obviously far earlier and served as a model: the fullest study is that of Spanke (1936); Handschin (1954) gave good reason for thinking that both sequence and lai have their roots in the Celtic tradition; and Stäblein (1954, 1962) showed how a close examination of the Latin precursors can throw important light on the history of the lai.

Questions of definition are made even more difficult by the inserted ‘Amen’ and ‘Evovae’ phrases in lais by Ernoul le vielle de Gastinois and Leichs by Frauenlob, perhaps just as indications of modality but possibly suggesting some liturgical function (see Maillard, 1963, pp.310ff; März, 1988; Shields, 1988). Perhaps the most interesting contrafacta, however, merit inclusion on their own account: Heinrich Laufenberg’s two Salve regina parodies have no connection with the metrical scheme of the original because they set the melismatic melody syllabically.

These works all belong in a kind of no-man’s-land between the lai and the sequence, but they should not allow any confusion as to the fundamental separateness of the two forms. The classic sequence has a compactness and clarity of design that are entirely different from either the rambling motivic dialectic of the 13th-century lai or the closely defined stanza and repetition patterns of the form in the 14th century. Certain aspects of the essence of both are more easily understood if sequence and lai are seen as the same form, but many others make sense only if the two are considered separately.

Lai

2. Poetic form.

Maillard (1963) provided an extensive listing of extant poems in the lai form: far fewer than half of them survive with music. Since musical form was at the time almost invariably determined by poetic form and since there are some discussions of it in the poetical theory, it is in literary terms that the essential features of the form’s history are most easily summarized.

Jeanroy (1901) mentioned the difficulty of being certain in identifying the stanzaic divisions of the 13th-century lyric lai: the poems are often too irregular to permit unambiguous analysis and the illuminated initials that normally begin a stanza in the manuscripts are sometimes clearly misplaced in lais, as though the early scribes were as confused as anybody. Put another way, the early lai is often a mere series of poetic lines, mostly brief and all rhyming with some other line, sometimes easily divisible into larger sections but not always giving any clear clue as to formal shape. An extreme example is the beginning of Comencerai by Thibaut IV (for the music see ex.1):

Comencerai / a fere un lai / de la meillor /
forment m’esmai / que trop parai / fet de dolor, / dont mi chant
torront a plor.
Mere virge savoree / se vos faitez demoree /
de proier le haut segnor / bien doi avoir grant paor /
du deable, du felon / que en la noire prison /
nos velt mener / dont nus ne peut eschaper;

Jeanroy, surveying the 30 13th-century lais known to him (a few more have since been discovered), listed a range of six to 19 stanzas, of two to 56 lines comprising two to 11 syllables but tending to favour the shorter lines that are in general a special characteristic of the lai throughout its history.

Few lais are quite as irregular as Thibaut’s Comencerai, and one of the most prominent features of the lai is normally the principle of responsion, or repeated material. ‘Lesser responsion’ is the immediate or almost immediate repetition of a metrical or poetic scheme in the manner of the ‘classical’ double-versicle sequence: the repeated metre and rhyme scheme normally bring with them a musical repeat. But in the lai, particularly in the early stages of its history, the scheme of a single line or a couplet will often be repeated several times before new material is introduced, and this brings in its train the multifold repetition of short melodic fragments. In the 14th-century lai lesser responsion became customary throughout, normally in the form of regular fourfold responsion which was given music with alternating ouvert and clos cadences. Thus the 11th stanza of Machaut’s Le lay de l’ymage runs:

Riens ne desir / tant com li servir / a plaisir. / Mi desir
sont la jour et nuit,
pour desservir, / en lieu de merir, / li veir, / li oir:
a ce Amours me duit.
Mieus vueil languir / pour li, sans mentir, / et morir / que joir
d’autre; c’est le fruit
dont soustenir / me vueil et norir. / La metir / la querir
vueil tout mon deduit.

‘Greater responsion’ entails repeating the scheme of a larger section later within the lai. Maillard (‘Lai, Leich’, 1973, p.335) gave the following formal analysis (others are possible) of the long ‘Lai de l’ancien et du nouveau testament’: ABCDEFG EHIJ EHIJ EIJ EIJ E AB. Such greater responsion is also found in the sequence (see Double cursus). In lais of the 14th century it had reduced itself to a single case: the last stanza normally repeated the verse scheme of the first, but here the music was often transposed to another pitch, thereby both breaking the convention of the same verse scheme having the same music and also anticipating the sonata form recapitulation idea in many of its essential features by some 400 years.

Spanke (1938), in making what is still the most comprehensive attempt to analyse and categorize the 13th-century lais, was particularly reluctant to postulate any ‘development’ of form within the lai, asserting that the strictest poetic form in Machaut was already known in the 12th century. There was merely an extraordinarily wide range of structures, from the almost shapeless, such as Thibaut’s Comencerai, through the heavily repetitive but rambling, such as the two lais of Ernoul de Gastinois, to the most painstakingly balanced lai with strictly paired lesser responsions, such as the ‘Lai du chevrefeuille’ or Puis qu’en chantant. On the other hand, it is also true to say that of the 46 Provençal poems and 109 Old French poems inventoried by Maillard (not all are strictly lais within the terms of this article and many are lost), remarkably few have 12 stanzas, the number favoured by Machaut and increasingly adhered to through the 14th century. So also, the earliest theorist to discuss the form at any length, Guilhem Molinier in the first version of Las leys d’amors (1328), was suitably vague in his definition, saying that it ‘can have as many stanzas as a vers, that is to say five or ten, these stanzas being singulars, and distinct in rhyme, in music and in text; and they can each have the same metre or different metre’.

Only one of the Roman de Fauvel lais has 12 stanzas, but nearly all of Machaut’s are in that form, later described by Eustache Deschamps (Art de dictier, 1392), who himself wrote 12 such poems. He described the lai as long and difficult to write (‘c’est une chose longue et malaisee a faire et trouver’), having 12 double stanzas of 16, 18, 20 or 24 lines, each stanza with a different rhyme except the last, which should repeat that of the first without however repeating any of the actual rhyme words. He also said that lais were fairly common: in view of his own output and the continued cultivation of this intricate and elaborate form by Froissart (who gave a similar definition in his Prison amoureuse, ll.3483–514), Oton de Granson, Christine de Pizan, Alain Chartier and Georges Chastellain, this is no surprise.

An even further increase in the rigidity of the form is apparent from the Regles de la seconde rhetorique (c1411–32) which gives a brief history of the form, according special honours to Philippe de Vitry and to Machaut (see §4 below). It describes the lai as having ‘12 stanzas of which the first and the last are similar in form and rhyme while the other ten are each individual in these respects; but each stanza must have four quarters’. Baudet Herenc (1432, ed. in Langlois, 1902, pp.166ff) was even more restrictive as to the length of each line and the permissible rhyme schemes.

Poetic theory on the subject in the later 15th century and after represents a falling away at the edges. Deschamps had described and exemplified a double lai with 38 stanzas, but for Molinet a century later the form had become so rigid that double lai was merely a lai with 16 lines in the stanza rather than the 12 lines that had by then become almost mandatory: the 12-line form most favoured was that with the rhyme scheme aab aab bba bba, often called the petit lai de contradiction but also sometimes called virelai – a name that is particularly confusing in musical contexts because the form has nothing at all to do with the standard Virelai of polyphonic song. At the same time a distinction evolved between the petit lai or commun lai with only one stanza and the full-size grand lai: a late example of a ‘commun lay’ in the Jardin de plaisance (Paris, c1501) comprises four stanzas with the same metre and rhyme scheme (that described above as ‘virelai’) and commends Chartier as a writer of lais. Thomas Sebillet’s Art poëtique françois (Lyons, 2/1556) mentioned the lai and the virelai as an afterthought, declaring both to be obsolete. One particularly interesting late example of a lai (discussed in Giacchetti, 1973) is in fact a connected ‘cycle’ of seven poems in the following sequence of forms: rondeau, fatras, virelai, fatras, ballade, fatras, rondeau; but it describes itself as being a lai, and is a lai in having each stanza in a different form from the last; moreover, Willaert (1992) notes a similar poem-sequence elsewhere in the same romance, Ysaÿe le triste, and remarks that comparable sequences can be found among the Dutch poems in the Hague Chansonnier (NL-DHk 128 D 2) of about 1400.

Lai

3. The lai before 1300.

The earliest lai repertories with music are most simply described in terms of their manuscript sources. Peripheral to the main subject of this article are the inserts in the Roman de Tristan en prose (c1225–30, found particularly in A-Wn 2542; for a version from a different manuscript see fig.3). These are a series of simple stanzaic songs, very much in the musical and poetic style of the trouvère tradition at the time. Several of them are entitled ‘lai’ and they are the only surviving music with that title actually associated with Arthurian or narrative material: for these reasons they are included in the checklist below and must be considered in any study of the subject so long as the question of definition remains unsolved. But nothing in their poetry or their music suggests any connections with the genre under consideration here.

The main collection of true 13th-century lais is in the Chansonnier de Noailles (F-Pn fr.12615) which contains 17, ten of them together (fig.4). Smaller quantities of music, less clearly organized, appear in the Chansonnier du Roi (F-Pn fr.844), in F-Pn fr.845 and in the Wiener Leichhandschrift (A-Wn 2701). In two of the French sources the collection of chansons is followed by a group of two-voice motets after which the lais are to be found; and in the third collection a group of motets is nearby. All four sources are probably from the late 13th century but their repertories, partly anonymous, have such stylistic variety as to cause considerable disagreement on the early history of the lai form.

The Provençal vida of Garin d’Apchier states that he ‘made the first descort that ever was made’ (‘fetz lo premier descort que anc fos faitz’), which would place the birth of the form around 1200; and while it is surely pertinent to ask how much the anonymous 14th-century biographer really understood about the subject, just as his comment raises again the question as to whether any clear distinction between the descort and the lyric lai was intended, it seems possible that the form appeared in the troubadour and trouvère traditions only shortly before 1200. This judgment may seem dangerous in view of the far greater antiquity of the sequence tradition, of the various discussions as to the roots of the sequence, and of the Celtic roots Handschin and others impute to the lai; nevertheless it makes sense in terms of the lyric tradition of the troubadours and trouvères as it survives, a tradition that grew up gently in the 12th century, gradually trying new formal ideas and expanding its boundaries as the century progressed.

Surveying the lais of the trouvère repertory (published almost complete in Jeanroy, Brandin and Aubry, 1901), Aubry called attention to a clear stylistic division falling approximately between those that were anonymous and those whose authors were known. The former (together with the two lais of Ernoul de Gastinois, who has even been suggested as a possible author for the others because they appear immediately after his lais in the Chansonnier de Noailles) include some extremely irregular patterns and occasional apparently incomplete musical notation on the basis of which a complete edition can require reconstruction and even guesswork. Aubry concluded that the incompletely notated lais were more improvisational and that they left hints as to possible Celtic origins for the genre; and several other scholars have pointed out that the tendency to repeat relatively brief musical phrases eight or nine times before moving on is perhaps symptomatic of some derivation from the style of the chanson de geste. Stäblein (1975) went further and proposed a chronology based on the assumptions that these less formalized pieces were earlier (an opinion it is difficult to endorse wholeheartedly in view of the full formalization of the substantially similar sequence form by 1100), and that the ‘Lai des pucelles’ (Coraigeus) was already written in the early 12th century when Abelard wrote his ‘Planctus virginum Israel’ to the same melody (for a contrary opinion see Weinrich, 1969): Stäblein suggested that all these anonymous lais may date from about 1100. He also proposed a chronology for the ascribed lais beginning with the three rather stolid pieces of Gautier de Dargies early in the 13th century, then the slightly freer works of the brothers Le Vinier; the second half of the 13th century was then represented by Colin Muset, Adam de Givenchi, Thomas Herier, Jacques de Cysoing and Ernoul le vielle de Gastinois.

Rietsch (DTÖ, xli, Jg.xx, 1913/R) observed many of the same characteristics in the Leichs of the Wiener Leichhandschrift as he found in some of the lais in the Jeanroy collection, especially those of Ernoul de Gastinois: he mentioned the short rhyming lines with no obvious scheme, rarity of clear stanza patterns and the use of greater responsion. Similarly, Stäblein (1975, pp.172–3) made an extremely interesting melodic analysis of a Leich by Tannhäuser, Ich lobe ein Wîp, and pointed out (p.97) that it is closely related in style to the French lais of the time. An adequate survey of the genres in the 13th century will need to view the French and German traditions together, if only because it is likely that none of the music in the French sources dates from later than about 1250 and the chronological (and stylistic) gap before the lais of the Roman de Fauvel is slightly closed by a consideration of the Leich of Herman Damen, a long rambling piece with constantly developing material, as well as those of Frauenlob with their consistently precise repetition in double versicles.

The developing melodic material just mentioned is perhaps the most absorbing characteristic of the lai as a musical form: figures and motifs carried from one section and expanded in the next are to some extent inevitable in a sectional monodic form the size of the lai, particularly if the composer has any feeling for the need to supply some shape to his work. In the opening section of Comencerai each unit builds on the preceding one, but the fifth unit grows anew out of the melodic material that had evolved at the end of the third and fourth units (ex.1; the texts, which have one syllable for each note or neume, are represented by only the rhyming syllable at the end of each line; for the text see §2 above).

Something similar happens in Herman Damen’s Ir kristenen. The section of it in ex.2 contains short fragments repeated several times and representing a considerable change of pace after the more casual aab form that had characterized all that went before in the piece. This section is therefore transitional, setting up its own signposts, perceptibly changing character and, in the event, leading to a new section with a higher tessitura. The transitional passage treats a simple melodic figure in several different ways, finally transposing it up a tone before launching into the new section – at which point the frequency of repetition relaxes again and the melodic material for the first time in the piece has no obvious relation to that of the preceding sections. (For a similar attempt to describe part of a lai by Machaut, see Fallows, 1977.)

Lai

4. The lai after 1300.

Schrade (1958) put forward tentative but persuasive arguments for suspecting that the four lais in the Roman de Fauvel could be the work of Philippe de Vitry, who otherwise contributed some of the most distinguished and modern music to that collection. The Regles de la seconde rhetorique (see Langlois, 1902, p.12) said: ‘Aprez vint Philippe de Vitry qui trouva la maniere des motés, et des balades, et des lais et des simples rondeaux’, and continued by praising Machaut ‘le grant retthorique de nouvelle fourme, qui commencha toutes tailles nouvelles et les parfais lays d’amours’. In any case these four Fauvel lais are the only immediate precursors of the magnificent lais of Machaut. Their musical rhythm is clear, their sections are carefully balanced, and the repetitions have reached a regularity of form that suggests an effort to build on the basis of the 13th-century lai and crystallize it into a more balanced and logical shape in line with so many other innovations of the early 14th century. In these lais there is an even clearer sense of melodic progress: each section builds on the material with which the last ended, so the result is a constantly developing musical organism along the lines seen in the 13th-century lai but more continuous, more conscious of a thematic evolution.

Machaut’s lais often include this feature, and in formal terms they follow the same scheme as those of the Roman de Fauvel. But the development of the mensural system opened new avenues: he was able to vary the pace from one section to another thereby giving this still-growing form an opportunity to become even larger, with (effective) changes of tempo giving the whole piece a sense of variety and springiness that was not found earlier. In a way this is the major importance of Machaut’s lais in musical history, for none of his other works is so long as to require such full and conscious exploitation of the mensural system, using musical techniques more commonly found in the larger mass cycles of the 15th century.

Four of Machaut’s lais contain polyphony written in ‘successive notation’. This is clearly indicated only in Je ne cesse de prier where the even-numbered stanzas have the annotation ‘chace’ and make three-voice polyphony if performed in unison canon with the entries three perfections apart. S’onques douleureusement can be performed in three-voice unison canon throughout. The polyphony of Pour ce que plus proprement is written after the manner of certain St Martial sources: the first versicle in each double-versicle stanza has different music from the second, but the two fit in excellent polyphony. Finally the music for stanzas 1, 2 and 3 in En demantant combines in three-voice polyphony as does that for stanzas 4–6, 7–9 and 10–12. In all these cases there is a serious question as to how the music might best be performed: if each voice was sung and texted the cumulative effect of the poem would presumably be lost, so perhaps the canonic possibilities are merely aids towards the construction of an instrumental accompaniment more carefully controlled by the composer than it was in the monophonic repertories of the preceding generation. The polyphony of Machaut’s lais has been observed but not explained.

If Machaut’s lais must be regarded as the highpoint of the form’s history, there is little evidence that any of his successors followed this lead. That 15th-century composers seem not to have given the form their attention may perhaps be because it was still an essentially monodic form, and monophonic writing had none of the prestige it enjoyed in earlier centuries. (Securely ascribed monophonic song in the 15th century survives only from Germany.) Another reason was possibly that the composer interested in developing extended forms concentrated more on the cyclic mass which was now beginning to take its definitive form.

But two examples of later lai composition have come to light and the circumstances surrounding each suggest that they were not entirely isolated but rather examples of a larger tradition that happens to have been lost. The lai De cuer je soupire (c1400; ed. in Wright, 1974) has exact double versicles and ouvert–clos cadences: the nature of its melodic line and its changes of pace both clearly separate it from the sequence tradition. And if the brevity of its sections contrasts strongly with the inflated length of the unset lais of Chartier and Granson this may be seen as the only way of keeping the form within the scope of other secular monophony in the 15th century. A similar situation exists in Hans Folz’s ‘Kettenton’ (c1500), which Petzsch (1970) ingeniously identified as being in the lai form: again the sections are brief, and the continuous development of material passing from one section to the next is fully within the received tradition. If these are symptoms characteristic of a larger tradition, the size of the form in the 13th century was considerably expanded and reduced to formulae by composers in France and Germany after 1300, whereas the composers after Machaut reduced size and scope with the lai as they did with so much of their cultural heritage.

Lai

5. Notes on the checklist of lai music.

While it is clear that a study of the lai must take account of the poems and their form, it seems also that much can be learnt from a study of the music alone, and it is to this end that the appended list has been compiled. The fullest listing of texts is still that in Maillard (1963), where they are tabulated separately under lais (pp.71ff), descorts (pp.119ff) and Leichs (pp.153ff) and much additional material is included. In the list here four categories of material have been deliberately omitted: (i) the many poems in lai form with no surviving music: it is not suggested that these are unimportant, but that the reader of this dictionary might be grateful for a listing that confines itself to what music is actually there; (ii) other text sources for lais whose music does survive: they can quickly be ascertained from Maillard, from Tischler (1997 edn) and from the standard inventories of medieval literary genres; (iii) Latin pieces that might be described as in lai form: these are listed and discussed in Spanke (1936), and many are edited in Tischler (1997); (iv) strophic songs in which the individual stanza includes several musical repetitions, many of which are edited in Tischler (1997); (v) pieces that happen to be called lais either in manuscript titles or within the poem, but show no apparent trace of lai form (examples would be Gautier de Coincy’s Entendez tuit ensemble and Harder’s ‘Kôrwîse’), pieces that are in the form but are straight translations of sequences (such as Gautier’s Hui enfantés) and pieces showing only elements of lai form (see Spanke, 1936, pp.91ff).

Unresolved questions as to the strict definition of the lai and as to its pre-history have made some of the decisions in this last category extremely difficult. One category has been included even though it strictly has nothing to do with the lai as defined here: the lyric insertions in the Roman de Tristan en prose are listed because many are entitled ‘lai’ and they represent a complete independent category of material; moreover they are included as such in the major studies of Wolf (1841) and Maillard (1963), and their omission could therefore cause unnecessary confusion.

Editions listed are not necessarily the best or the most recent but merely the most convenient for obtaining an overview of the repertory. Descriptions of form are added only when unambiguous and relatively simple.

Lai

CHECKLIST OF LAI MUSIC

Editions: Die Sangesweisen der Colmarer Handschrift und die Liederhandschrift Donaueschingen, ed. P. Runge (Leipzig, 1896/R) [Ru]Lais et descorts français du XIIIe siècle, ed. A. Jeanroy, L. Brandin and P. Aubry (Paris, 1901/R) [J]Gesänge von Frauenlob, Reinmar v. Zweter und Alexander, ed. H. Rietsch, DTÖ, xli, Jg.xx (1913/R) [Ri]Guillaume de Machaut: Musikalische Werke, iv: Messe und Lais, ed. F. Ludwig and H. Besseler (Leipzig, 1954) [L]The Works of Guillaume de Machaut: First Part, ed. L. Schrade, PMFC, ii (1956) [S]J. Maillard: Evolution et esthétique du lai lyrique (Paris, 1963) [M]R.J. Taylor: The Art of the Minnesinger (Cardiff, 1968) [Ta]Les lais du roman de Tristan en prose d’après le manuscrit de Vienne 2542, ed. T. Fotitch and R. Steiner (Munich, 1974) [FS]Trouvère Lyrics with Melodies: Complete Comparative Edition, ed. H. Tischler, CMM, cvii/14–15 (1997) [T]

provençal lais

lyric lais in the roman de tristan en prose (A-Wn 2542)

all isostrophic

french lais before 1300

lais in the roman de fauvel (F-Pn fr.146)

lais by guillaume de machaut

in Vg [Wildenstein Gallery, New York], A [F-Pn fr.1584], B [Pn fr.1585], C [Pn fr.1586], E [Pn fr.9221] and G [Pn fr.22546] unless otherwise stated

leichs before 1300

leichs around 1300

later leichs

latest examples

Lai

BIBLIOGRAPHY

MGG1 (‘Lai, Leich’, U. Aarburg; see also ‘Descort’, J. Maillard)

MGG2 (‘Lai, Leich’, C. März)

F. Wolf: Über die Lais, Sequenzen und Leiche (Heidelberg, 1841/R)

C. Appel: ‘Vom Descort’, Zeitschrift für romanische Philologie, xi (1887), 212–30

H.R. Lang: ‘The Descort in Old Portuguese and Spanish Poetry’, Beiträge zur romanischen Philologie: Festgabe für Gustav Gröber (Halle, 1899), 484–506

A. Jeanroy, L. Brandin and P. Aubry: Lais et descorts français du XIIIe siècle: texte et musique (Paris, 1901/R)

E. Langlois, ed.: Recueil d’arts de seconde rhétorique (Paris, 1902)

O. Gottschalk: Der deutsche Minneleich und sein Verhältnis zu Lai und Descort (diss., U. of Marburg, 1908)

P. Runge: ‘Der Marienleich Heinrich Laufenbergs “Wilkom lobes werde”’, Festschrift … Rochus Freiherrn von Liliencron (Leipzig, 1910/R), 228–40

G. Hase: Der Minneleich Meister Alexanders und seine Stellung in der mittelalterlichen Musik (Halle, 1921)

F. Gennrich: ‘Internationale mittelalterliche Melodien’, ZMw, xi (1928–9), 259–96, 321–48

J. Handschin: ‘Über Estampie und Sequenz’, ZMw, xii (1929–30), 1–20; xiii (1930–31), 113–32

H. Spanke: ‘Eine neue Leich-Melodie’, ZMw, xiv (1931–2), 385–97 [Tannhäuser’s Ich lobe ein Wîp]

H. Spanke: ‘Über das Fortleben der Sequenzenform in den romanischen Sprachen’, Zeitschrift für romanische Philologie, li (1931), 309–34

F. Gennrich: Grundriss einer Formenlehre des mittelalterlichen Liedes (Halle, 1932/R)

H. Spanke: Beziehungen zwischen romanischer und mittellateinischer Lyrik mit besonderer Berücksichtigung der Metrik und Musik (Berlin, 1936/R)

H. Spanke: ‘Sequenz und Lai’, Studi medievali, new ser., xi (1938), 12–68

F. Gennrich: ‘Zwei altfranzösische Lais’, Studi medievali, new ser., xv (1942), 1–68

H. Spanke: Deutsche und französische Dichtung des Mittelalters (Stuttgart, 1943)

G. Vecchi: ‘Sequenza e lai: a proposito di un ritmo di Abelardo’, Studi medievali, new ser., xvi (1943–50), 86–101

H. Kuhn: ‘Ulrich von Winterstetten und der deutsche Leich’, Minnesangs Wende (Tübingen, 1952, 2/1967), 91–142

J. Handschin: ‘Trope, Sequence, and Conductus’, NOHM, ii (1954), 128–74

B. Stäblein: ‘Von der Sequenz zum Strophenlied: eine neue Sequenzenmelodie “archaischen” Stiles’, Mf, vii (1954), 257–68

A. Machabey: Guillaume de Machault 130?–1377 (Paris, 1955), i, 98–130

G. Reaney: ‘The Lais of Guillaume de Machaut and their Background’, PRMA, lxxxii (1955–6), 15–32 [incl. information on earlier lais]

R.J. Browne: A Stylistic and Formal History of the Middle High German Leich, 1190–1290 (diss., Yale U., 1956)

R.H. Hoppin: ‘An Unrecognized Polyphonic Lai of Machaut’, MD, xii (1958), 93–104 [Pour ce que pens proprement]

J. Maillard: ‘Problèmes musicaux et littéraires du “lai”’, Quadrivium, ii (1958), 32–44

G. Reaney: ‘Concerning the Origins of the Medieval Lai’, ML, xxix (1958), 343–6

L. Schrade: ‘Guillaume de Machaut and the Roman de Fauvel’, Miscelánea en homenaje a Monseñor Higinio Anglés (Barcelona, 1958–61), 843–50

J. Maillard: ‘Le “lai” et la “note” du Chèvrefeuille’, MD, xiii (1959), 3–13

B. Stäblein: ‘Die Schwanenklage: zum Problem Lai–Planctus–Sequenz’, Festschrift Karl Gustav Fellerer zum sechzigsten Geburtstag, ed. H. Hüschen (Regensburg, 1962), 491–502

J. Maillard: Evolution et esthétique du lai lyrique des origines à la fin du XIVe siècle (Paris, 1963)

K.H. Bertau: Sangverslyrik: über Gestalt und Geschichtlichkeit mittelhochdeutscher Lyrik am Beispiel des Leichs, Palaestra, ccxl (Göttingen, 1964)

J. Maillard: Lais et chansons d’Ernoul de Gastinois, MSD, xv (1964)

D. Poirion: Le poète et le prince: l’évolution du lyrisme courtois de Guillaume de Machaut à Charles d’Orléans (Paris, 1965/R), 397ff

H. Baader: Die Lais: zur Geschichte einer Gattung der altfranzösischen Kurzerzählungen (Frankfurt, 1966)

J. Maillard: Roi-trouvère du XIIIème siècle: Charles d’Anjou, MSD, xviii (1967)

J.-M. d’Heur: ‘Des descorts occitans et des descordos galiciens-portugais’, Zeitschrift für romanische Philologie, lxxxiv (1968), 323–39

R. Baum: ‘Les troubadours et les lais’, Zeitschrift für romanische Philologie, lxxxv (1969), 1–44 [analysis of 58 uses of the word ‘lai’ in Provençal]

I. Glier: ‘Der Minneleich im späten 13. Jahrhundert’, Werk–Typ–Situation: Studien zu poetologischen Bedingungen in der älteren deutschen Literatur: Hugo Kuhn zum 60. Geburtstag, ed. I. Glier and others (Stuttgart, 1969), 161–83

J. Maillard: ‘Lais avec notation dans le Tristan en prose’, Mélanges offerts à Rita Lejeune (Gembloux, 1969), 1347–65 [rectifies omissions in Maillard, 1963]

L. Weinrich: ‘Peter Abaelard as Musician’, MQ, lv (1969), 295–312, 464–86

M. Hasselman and T. Walker: ‘More Hidden Polyphony in a Machaut Manuscript’, MD, xxiv (1970), 7–11 [En demantant]

C. Petzsch: ‘Ein spätes Zeugnis der Lai-Technik’, ZDADL, xcix (1970), 310–23 [Folz’s ‘Kettenton’]

R. Baum: ‘Le descort ou l’anti-chanson’, Mélanges de philologie romane … Jean Boutière, ed I. Cluzel and F. Pirot (Liège, 1971), 75–98 [with extensive bibliography]

J. Maillard: ‘Coblas dezacordablas et poésie d’oïl’, ibid., 361–75

G. Objartel: ‘Zwei wenig beachtete Fragmente Reinmars von Zweter und ein lateinisches Gegenstück seines Leichs’, Zeitschrift für deutsche Philologie, xc (1971), suppl., 217–31

C. Bullock-Davies: ‘The Form of the Breton Lay’, Medium aevum, xlii (1973), 18–31

A. Giacchetti: ‘Une nouvelle forme du lai apparue à la fin du XIVe siècle’, Etudes de langue et de littérature du Moyen Age offertes à Félix Lecoy (Paris, 1973), 147–55

J. Maillard: ‘Lai, Leich’, Gattungen der Musik in Einzeldarstellungen: Gedenkschrift Leo Schrade, ed. W. Arlt and others (Berne and Munich, 1973), 323–45

J. Maillard: ‘Structures mélodiques complexes au Moyen Age’, Mélanges de langue et de littérature médiévales offerts à Pierre Le Gentil (Paris, 1973), 523–39 [Ses alegratge]

T. Fotitch and R. Steiner: Les lais du roman de Tristan en prose d’après le manuscrit de Vienne 2542 (Munich, 1974) [incl. complete edn]

C. Wright: ‘A Fragmentary Manuscript of Early 15th-Century Music in Dijon’, JAMS, xxvii (1974), 306–15 [De cuer je soupire]

C. Petzsch: ‘Folgen nachträglich eingeführter Text-Form-Korrespondenz für Text und Lai-Technik’, Mf, xxviii (1975), 284–7 [Lesch’s ‘Tageweise’]

B. Stäblein: Schriftbild der einstimmigen Musik, Musikgeschichte in Bildern, iii/4 (Leipzig, 1975), 95ff, 172ff

E. Köhler: ‘Deliberations on a Theory of the Genre of the Old Provençal Descort’, Italian Literature: Roots and Branches: Essays in Honor of Thomas Goddard Bergin, ed. G. Rimanelli and K.J. Atchity (New Haven, 1976), 1–13

D. Fallows: ‘Guillaume de Machaut and the Lai: a New Source’, EMc, v (1977), 477–83

R. Baum: ‘Eine neue Etymologie von frz. lai und apr. lais; zugleich: ein Plädoyer für die Zusammenarbeit von Sprach- und Literaturwissenschaft’, Beiträge zum romanischen Mittelalter, ed. K. Baldinger (Tübingen, 1977), 17–66

E. Mulder: ‘Einige Bemerkungen zu Machauts “Lay de L’ymage”’, Mf, xxxii (1979), 58–62

J.H. Marshall: ‘The Isostrophic Descort in the Poetry of the Troubadours’, Romance Philology, xxxv (1981), 130–57

H. Tischler: ‘Die Lais im Roman de Fauvel’, Mf, xxxiv (1981), 161–79

J. Maillard: ‘Descourt, que me veux-tu? … ’, Cahiers de civilization médiévale, xxv (1982), 219–23

D. Billy: ‘Le descort occitan: réexamen critique du corpus’, Revue des langues romanes, lxxxvii (1983), 1–28

J. Maillard: ‘Un diptyque marial chez Guillaume de Machaut: les lais XV et XVI’, Mélanges de langue et de littérature médiévales offerts à Alice Planche, ed. M. Accarie and A. Queffélec (Nice, 1984), 328–37

J. Stevens: Words and Music in the Middle Ages (Cambridge, 1986)

C. März: Frauenlobs Marienleich: Untersuchungen zur spätmittelalterlichen Monodie (Erlangen, 1987)

R. Deschaux: ‘Le lai et la complainte’, La littérature française aux XIVe et XVe siècles, ed. D. Poirion, i: Grundriss der romanischen Literatur des Mittelalters (Heidelberg, 1988), 70–85

C. März: ‘Zum musikalischen Stil in Frauenlobs Kreuzleich’, Wolfram-Studien (1988), 125–34

M. Shields: ‘Zum melodischen Aufbau des Marienleichs’, ibid., 117–24

A.I. Buckley: A Study of Old French Lyric Lais and Descorts and Related Latin Song to c.1300 (diss., U. of Cambridge, 1990)

J.H. Marshall: ‘The Transmission of the Lyric Lais in Old French Chansonnier T’, The Editor and the Text: in Honour of Professor Anthony J. Holden, ed. P.E. Bennett and G.A. Runnalls (Edinburgh, 1990), 20–32

H. Apfelböck: Tradition und Gattungsbewusstein im deutschen Leich (Tübingen, 1991) [see also reviews by C. März: Beiträge zur Geschichte der deutschen Sprache und Literatur, cxvi (1994), 130–37, and by M. Shields: Medium aevum, lxii (1993), 355–7]

V. Newes: ‘Turning Fortune's Wheel: Musical and Textual Design in Machaut's Canonic Lais’, MD, xlv (1991), 95–121

S.N. Rosenberg and H. Tischler: The Monophonic Songs in the Roman de Fauvel (Lincoln, NE, 1991)

J. Stevens: ‘Samson dux fortissime: an International Latin Song’, PMM, i (1992), 1–40

F. Willaert: ‘Courtly Love Songs in Ysaÿe le triste’, Crossings: a liber amicorum for Denis Conlon, ed. G. Kums (Antwerp, 1992), 152–63

A. Buckley: Introduction to Lyric Lais (Newton Abbot, 1992–4) [incl. edn]

D. Billy: Deux lais en langue mixte: le lai Markiol et le lai Nompar (Tübingen, 1995)

M. Staehelin: ‘Kleinüberlieferung mehrstimmiger Musik vor 1550 in deutschem Sprachgebiet, I: die Notre-Dame-Fragmente aus dem Besitz von Johannes Wolf’, Nachrichten der Akademie der Wissenschaften zu Göttingen, I: philologisch-historische Klasse (1999), no.6, 281–313