Composition in the Vihuela Songs of Luis Milan and Alonso Mudarra

The aim of this study is to explore the accompanied solo songs included in two collections of Spanish vihuela music from the first half of the sixteenth century and their possible links with unwritten improvisatory traditions. In contrast to the arrangements of vocal polyphony for voice and vihuela in later books, the songs of Luis Milan and Alonso Mudarra are unique not only in Spanish music, but also in a broader European context. 1 Until very late in the sixteenth century, most Jute songs in European sources are intabulated arrangements of chansons, motets, madrigals, frottole and other polyphonic genres. The questions raised by the songs considered here concern the provenance of the solo lute song and the manner in which it was cultivated by early practitioners.3 Traditional readings of the development of European music see Milan and Mudarra's vihuela songs as something of an anomaly. Older and now generally outrnoded theory, rooted firmly in source studies, held that the known tradition of solo instrumental music was born in the late fifteenth and early sixteenth centuries, and that its subsequent development was a process of gradual emancipation from the dominant polyphonic tradition that spawned it. On the other hand, the accompanied solo sang has been viewed as one of the great novelties of the late sixteenth century and, as a novelty of its time, is championed as one of the key inventions that separates the musical baroque from the renaissance. These ideas are still purveyed in many general histories of music and under-

3 Many of the observati.ons of style raised here and some of thcir possible origins werc first raised by John Ward, T11e » Vlhuela de mano« and itr Music, 1536-1576(Ph.D. diss. New Yorlc University, 1953. More recently they have been discussed from a different yet complementary standpoint by Lafargue, Par un luth (cf. fu. 1), especially Chapter 5 ofher thesis: »L: cas espagnol: ,inventi.ons< et compositi.ons sur timbre«, pp. 145-92. graduatc curricula throughout thc world, cvcn if no langer rcadily acccpted by contcmporary musicologists, many of whom havc lang abandoncd such linear notions of history. Notwithstanding any rcticcncc to admit unwrittcn cxtemporised traditions to the ccntral dcvelopmcnt of art music, thc rcputc of instrumental improviscrs in thc cvolutionary phasc of thc instrumental tradition has bccn amply acknowlcdged. Although this rccognition has rcstorcd some of the status thcsc musicians cnjoycd in thcir own time, thc lack of tangiblc cvidencc of thcir music and pcrformancc practiccs has all but precludcd its assimilation into stylc-bascd music historics. A closer undcrstanding of thc musical practices of musicians outside thc writtcn tradition helps to placc instrumental music, and accompanicd sang, on a broader historical continuum as part of an ongoing process rathcr than as a novclty of thc carly sixtccnth ccntury. Milan and Mudarra's songs offer somc sharp insights into thc practiccs of sang pcrformancc in thc pcriod parallel to and immediately prcccding thc advcnt of thc writtcn tradition, possibly illuminating somc of the extempore musical practiccs inhcritcd by them from earlier generations of non-musically litcratc musicians. Thc music published by Luis Milan in El Maestro (Valencia, 1536) is in cvcry way maturc and accomplishcd, and cvidcncc of improvisatory practicc abounds at cvcry turn in his solo writing and sang accompanimcnts. 4 Bcyond his virtuosic showmanship, Milan displays mastcry of his musical medium, achicving eloqucncc and purposc through intuitive musical skills that arc tcchnically emde in comparison to thc sophisticatcd tcchniquc of contcmporary polyphonists. This so-callcd crudcncss is relative: by comparison to sophisticatcd vocal countcrpoint his voice-lcading lacks polish, but thc deficicncics arc counterbalanccd by the sophistication and rcfincmcnt of his idiomatic vihucla writing. Thcrc is no qucstion of thc validity of his music, but wc havc no strong evidcnce of its stylistic provcnancc. Milan indicatcs clcarly that his music was gcncratcd on the vihuela and then writtcn down, dcscribing his book as »made of many works: which wcrc composcd on thc vihuela and writtcn down«. 5 The rccurrcncc of many motivcs, phrascs, and passages throughout his works is strong cvidcncc for his music being improviscd from a stockpilc of idiomatic deviccs rathcr than as uniquc works, cach craftcd to achicvc individuality. John Ward's dcscription of Milan's music as »a bridge betwccn thc improvisatory 4 Luis Milan, Libro de Musica de Pihuela de mano. Intitulado Et Maestro (Valencia, 1536;reprint Geneva, 1975). Modem editions include: Libro de Musica de Vihueta de mano intitulado EI Maestro. Publikationen Älterer Musik 2, ed. Leo Schrade (Leipzig, 1927;reprint Hildesheim, 1967); Libro de Musica de Pihuela de mano intitulado Et Maestro, ed. Ruggero Chiesa (Milan, 1974); ElMaestro, ed. Charles Jacobs (Harmony Park, Pa., 1971). 5 Milan, ibid., fol. Aiii: »un libro hecho de muchas obras: que de la Pihuela tenia sacadas y escritas«. style of the Petrucci and Attaingnant lutenists and the technically more mature style of the Francesco da Milane generation« locates him historically, and on the questions of style and provenance, Ward identified him with the Italian improwisatun whose art lay as much in the manner of performance as in the matter performed. Certainly the impress of improvisation is on many pages of El Maestro . Tue fact that the music does not lack polish and substance speaks both for Milan's native gifts and for the artistic sophistication of the tradition within which he work:ed ... [ and] ... bespeaks a schooling in an essentially perfonnance art. 6 In this sense, Ward defines Milan as the endpoint of one tradition as weil as the beginning of another, but at the same time notes the exceptionality of his songs, commenting that they »speak the language of the baroque with a renaissance dialect.« 7 Much of my own research on Milan, as weil as other studies by Craig Russell and Luis Gasser, has clone little more than expand on Ward's observations, illuminating in more detail many aspects of improvisatory style and the parameters within which Milan operated. 8 Milan's presence at the Valencian court is another significant factor in relation to the questions of the provenance of his improvisatory style. Milan was not a salaried musician, but »a kind of mattre du plaisir«, as Ward described him, at the court of Germaine de Foix and the Duke of Calabria, Ferdinand of Aragon, son of the last Aragonese ruler of aples. 9 As can be gleaned from the autobiographical passages of his later book El Cortesano (1561), among other things an account his life at the court in the 1530s, Milan's role included entertaining the ladies of the court, playing games, and singing. 10 The situations in which Milan describes himself singing always appear to be spontaneous rather than formal. As Gasser has observed, when invited to sing, Milan never had a vihuela with him, an instrument had to be brought so that he could play. 11 As the court of the former Aragonese rulers of aples, Neapolitan influence would have been unavoidable and V alencian traditions are likely to have devel- oped through a combination of local and imported influences and the interaction of Spanish and Italian musicians. The Catalan improviser Benedetto Garet!\ fl Charif;eo, was noted for extemporised singing to the lute in Aragonese Naples, probably following the practice that was common in other Italian regions in the hands of musicians such as Pietrobono and Serafino dall'Aquila, both of whom also served in Naples for a time. 12 The Neapolitan connection links this study closely to the contribution of Dinko Fabris to this volume, and his exploration of the early villanella as a Neapolitan sang sung to the accompaniment of the lute or viola da mano. The common point between our work is the evidence suggesting that both the villanella, the Spanish villancico and several other Iberian sang types originated as solo sang with lute or vihuela accompaniment. Although the available evidence is circumstantial, it is not improbable, sirnilar to the Neapolitan situation, that many of the songs in polyphonic sources such as the Cancwnero de Palacio may have originated as accompanied songs, or were performed as accompanied songs, perhaps to known melodies sung to extemporised accompaniments.
Alonso Mudarra was a very different musician to Milan, probably younger by ten years or so, from a different region of Spain, with a different cultural background, and much more strongly schooled in music of the polyphonic tradi tion. According to his own testimony in his Tres libros de musica (Seville, 1546 ), he was raised in the cultivated household of the Dukes of the Infantado in Guadalajara, possibly as a page. 13 His book of vihuela music was published in Seville on the eve of taking up a canonry in that city's cathedral, a position he held until his death in 1580. The Tres libros contains intabulations of music by Josquin and Gombert, and many fantasias that clearly indicate his assirnilation of the contemporary irnitative polyphonic style. At the same time and like other vihuelists of his time, Mudarra also included music that shows native roots, works such as his variation sets on »Guardame las vacas« and »Conde Claros« that are at the heart of Spanish popular tradition and the improvised singing of romances. Many of his own songs are likely also to have links to this native tradition. Perhaps through exposure to the professional musicians at the palace of his patrons in Guadalajara, Mudarra had gained first-hand knowledge of the 12 Such singer-lutenists are discussed by Gasser, ibid., pp. 29-33, and a rnuch wider peri;pective on Italian irnproviseri; is given in Ernile  practices of vihuela improvisation in the traditional style of non-musically literate musicians. His renowned Fantasia que contrahau la harpa, an elaborate set of folia variations, shows Mudarra to have understood older styles of improvisatory performance practice, probably from the turn of the century. lt is likely that the Ludovico whose style the fantasia imitates is the same harpist who came to Spain from aples in the retinue of the exiled Ferdinand V, perhaps the same Italian from the north who served Ercole II d'Este in Ferrara in the 1470s. 14 This interpretation of Mudarra's harp fantasia thus suggests that he shou1d be considered from a different perspective to Milan. The purpose of this f antasia was to emulate a past style of a famous musician, and he does this by employing what he understood to be the style of an earlier generation, and the particular individual devices for which Ludovico was famous. The result is a work based on variations on a traditional improvisation scheme, glossed with idiomatic devices that recall Ludovico's playing, and stands in stark contrast to his other predominantly imitative fantasias. Mudarra can thus be seen as a commentator an past tradition, creating works by means of conscious reflective processes outside performance time. In the context of this discussion, he might be characterised as a composer rather than an improviser, in contrast to Milan who was in every way an extemporising singer-songwriter and practising participant in the tradition we are seeking to investigate. In developing this discussion of Milan and Mudarra, several strands need to be drawn together in order to define the context and the broader significance of the observations made below about individual works. The fundamental themes under discussion here are extemporised musical practice in the early sixteenth century, the relationship between written and unwritten traditions, musical interconnections between Italy and Spain, and the intellectual climate that gave rise to solo sang.
Spanish music of the period immediately preceding the books of the vihuelists is best known through several courtly-popular collections of polyphony, most notably the Cancionero de Palacio. The principal genres in these collections, villancicos and rumances are the same genres of secu1ar sang that predominate in the vihuela books, close to a hundred of the former and two dozen of the latter. Same of these songs may be settings of popular tunes, although romances were also sung to simpler recitation formulae. 15 lt is also likely that many sang melodies were newly composed or improvised according to wellestablished stylistic conventions. Same of the songs of other genres in the polyphonic cancioneros that predate the earliest written vihuela music may weil have originated as accompanied solo songs, either setting new or traditional melodies, or using well known melodic-harmonic schemes. This is not surprising as many singers and composers of polyphony are likely to have played polyphonic instruments, and there are musical details in these collections that support the assertion. Irregular counterpoint, such as passages of parallel fifths, in early villancicos may be due to the transfer of instrumental chord positions to a polyphonic setting. While this is speculative and little hard evidence can be invoked in its support, it also coincides with Dinko Fabris' investigation of the early Neapolitan 'Villanella, which suggests a close link with the Pillancico and performance as solo sang with lute accompaniment.
Ramance performance was fundamentally a part of oral tradition. The composed romances in the cancioneros are developed from them and some of the best known formulae for singing romances bear names that associate them with particular songs. The schemes of »Cancle Claros« and »Gu.ardame las vacas« (romanesca) were used extensively by instrumentalists as the basis for variation writing, and the improvisation of romance accompaniments on such schemes was evidently the touchstone that led to the development of the variation genre. In addition to these two, the other scheme that proliferates in villancicos and romances of the cancioneros is the early form of the folia, called pavana in Spanish instrumental sources until the 1590s. This scheme in particular has been shown to underpin some twenty polyphonic songs in the Cancionero de Palacio and is one of the bases for arguing their connection with popular tradition. 16 In addition to court performers, the common schemes mentioned were probably also sung widely in urban contexts as weil, and the little information that survives on blind, vihuela-playing oracioneros, suggests the existence of urban professional musicians who weil may have included such music in their repertories.17 The involvement of players of plucked instruments in ensemble performance is another factor that contributes to developing a profile of improvisatory practice. Polyphonie improvisation techniques of ensemble musicians have been outlined by Keith Polk, drawing together diverse documentary evidence from northern Europe, but also noting the extensive interchange between German and Italian musicians. 18 Lute duos formed an important part of minstrel music making in the fifteenth century, and the duos included in the Spinacino lutebook published by Petrucci are no doubt related to such a performance practice.19 Polk draws extensively from Tinctoris to give concrete suggestions an the nature of polyphonic improvisation and, given that Tinctoris spent a good part of his active li.fe in aples, his explanations of improvisatory techniques may have been in part modelled an his eapolitan experience.
Traditions of extemporised solo sang improvisation most extensively documented in Italy also have relevance to Spanish practice. Recent research has cast significantly new light an the Italian humanistic practices of improvised sang using both Latin and the vernacular and practised by both prof essional minstrels and influential humanists such as Marsilio Ficino. These are perhaps two parallel traditions, although there is insufficient knowledge of the music they sang to judge the extent to which they may be interconnected. On the one band, Italian court documents indicate improvised singing to the lute practised by lutenist singers with known associations to aples. lt may have been singers like the Catalan Benedetto Gareth who influenced Tinctoris to point to the Catalan practice of reciting epics to the bowed viola, 20 and it is highly likely that he and other musicians such as the harpist Ludovico transmitted Italian practices to Aragon. The extent that their practices may have influenced singer improvisers including Luis Milan, however, remains a matter for speculation.
Even less tangible are any direct connections between Spanish sang and the extemporised singing of Italian humanists. At the heart of humanist sang extemporisation was a concern for the restoration of classical poetic metre. Beyond this, we know little of their practice other that they typically accompanied themselves an the lyra, perhaps the lute, lira da braccio, or a generalised allusion to any stringed instrument. Evidence of singing humanistic texts in Spain is scarce although Tess Knighton maintains that it formed part of general education, 21 and a passage in Luis Vives' Linguae Latinae Exercitatio, a work dedi-  ( 1577) shows a concern in Spain for the same issues that lay behind humanist Latin singing. In the broadest sense, their interest was the revival of Orphic song and their desire to emulate the singing of classical antiquity. The books of the vihuelists also abound with classical references, most notably in their titles but also in their prefaces and illustrations. lt cannot be mere accident that Milan's book, the earliest source of a collection of solo songs, includes a prominent plate of Orpheus playing the vihuela. 24 Such references are indicative of a general consciousness of the vihuelists towards the music of antiquity. Milan and Mudarra's songs are part of these widespread cultural and intellectual currents, that despite the Jack of intermediary sources, manifested themselves decades later in the Jute songs that proliferated throughout Europe, the monodic style customarily attributed to the Florentine Camerata, and the ultimate advent of opera. songs are built on formulaic patterns while others are constructed from short phrases with internal melodic repetitions that do not correspond with poetic structures. Possibly as a consequence, many of the melodies are bland and unmemorable, with little melodic interest and no discernible attempt to craft lines that have any affective relationship to their texts. This characteristic may weil be interpreted as a sign of his greater concern with the manner of performance rather than musical content, a style in which delivery and the performer's art overshadows the musical substance. A

<.)
. Due to the connections with oral tradition, the romance provides a good place to begin exploring Milan's songs. The romance is a strophic narrative form with four-line stanzas that usually repeats the same music for each strophe. Rmnances are commonly of considerable length and Gasser has calculated that performances of complete texts using the music in El Maestro may have lasted more than an hour. 25 Milan's romance »Sospiraste Baldovinos« reveals elements that might possibly reflect an improvisatory conception of the music and a possible link with the unknowns of the oral tradition. At the same time the argument cannot easily be corroborated and is presented here with due caution. While there is evidence at our disposal to suggest improvisation in romance performance using formulae for recitation and accompaniment, it supposes an unrealistic leap of faith to argue that the formulae found in this one sang are typical of a more generalised practice. The unusual feature of Milan's setting of »Sospiraste Baldovinos« is that he provides music for two stanzas, thus producing a composition that divides into two separate musical units. The melody (example 1) is in mode 8, transposed a fifth. 26 The melodic style of the two stanzas is consistent in the predominantly stepwise movement, and the two are linked through the repetition of the initial motive of the sang at the beginning of lines 4, 5, and 8.

Sos -p1 -ras -tes B.1I -do -v1 -nos
Structurally, however, the two melodies are quite different and the melody of the first stanza has several hallmarks of a recitation model: it is formulaic, simple and bland. lt would adapt to any text on any topic, and is easy to manipulate during the performance of a lang strophic sang in order to obtain a broad range of rhetorical and dramatic effect. Successful performance is dependent on delivery. Each verse of the strophe follows the same pattern, rising a fourth, then cadencing by a descending semitone or tone (example 2). The first and fourth verses use the fourth ~g, while the inner verses are one tone higher e-a. The cadences of the first two verses cadence by a descending semitoneg-f/, a third above the final , and the latter two cadence by a descending tone e-d. This first strophe could easily be derived from recitation formulae that provided a framework for improvisers like Milan to use in sang extemporisation. The melody has no other characteristics that suggest it to be a carefully crafted line; the rising fourth of each verse is of little consequence. The second strophe does not follow a similar pattern, although it is no less simple. Each verse commences on e, and moves by ascending motion, and the final cadences are strategically placed ong, d, c/, and d to give overall coherence. lt could also represent an improvisation scheme, but it is not as overtly formulaic as the melody of the first strophe. ----------.JLe Example 3: Luis Milan, »Sospirastes Baldovinos«, melodic design, strophe 2 lt is clear from diverse evidence that singers normally provided their own accompaniments on the lute or vihuela, and this is reinforced by notational format of most lute sources, particularly the Spanish ones that indicate the voice to be sung in red ciphers or apostrophes within the idiomatic tablature notation.
In El Curtesano, Milan is portrayed as a singer-vihuelist in the same way. Accordingly, the accompaniment of »Sospirastes Baldovinos« (example 4) is formulated for the singer-instrumentalist: the accompaniment of the melody line is simple, predorninantly chordal and progressively incorporating !arger amounts of passing figuration. The redobles between each verse are part of combined vocal-instrumental phrase gestures that only really make sense when the singer is also the accompanist. Milan's music prolonging the final chord of the phrases of verses 1-3 and adds a short coda to mark the end of the strophe. Setting sonnets to music presents different challenges than strophic romances or refrain-based villancicos. The music of the romance ideally needs to have a character that matches the general mood of the text, yet be malleable enough to be easily adapted to each successive strophe. Villancicos, on the other hand, achieve success more than anything else through a catchy refrain melody, cemented in the listener's memory through repetition. Neither of these techniques is applicable to the sonnet which has no repetition, and is built from concise yet complex textual relationships. Sonnet setting in the sixteenth century was most typically the domain of madrigalists who mostly chose to set these texts without repetition of any kind, furnishing each line of poetry with individually crafted music to complcment textual meaning. Among vihuclists, Milan and Mudarra are the only ones to have set the sonnet to music, but in notably different ways. Mudarra composed according to the poctic form, usually setting at least both quatrains to the same music, and attempting to match music and meaning in a general sense as weil as through attention to the detail of specific words. Milan did neither of these things and it is the absence of a meditated concern for both form and meaning that is the strengest evidence to suggest that his sonnets derive from extemporised practice. Ward concluded that Milan was not successful in setting sonnets and describes his setting of Jacopo Sannazzaro's sonnet »O gelosia d'amanti« as ambitious, but monotonous and without his more customary finish and charm. Ward also alluded to possible roots in improvisation, »a music of formulas that had been used many times by the vihueli t in singing for friends and at court«. 27 Milan's version of Sannazzaro's sonnet is doubly interesting because it was also set by Mudarra and comparison of the two versions reveals the differences with great clarity. 28 Milan's setting of »O gelosia d'amanti«, in common with his other sonnet settings is set to a largely homophonic accompaniment that underpins the melody, reinforces the end of each line through decorated cadential figures, and adds some brief passage work at the end of five verses. 29 The main structural design of the music can thus be revealed through the examination of the melody alone, and suggests that Milan may weil have composed the sang extempore or adapted another melody for the purpose, perhaps even during an extemporised performance. An experienced improviser would approach the extemporisation of a sonnet using some predeterrnined idea at least to initiate the performance, and would either know the text by memory or have a written copy of the poem at hand. To commence an improvisation, it is much easier to think in terms of the melody and accompaniment together, rather than to envisage a melody without a preconception of its harmonisation. If he did in fact extemporise »O gelosia d'amanti«, Milan appears to have used the formula of »Guardame las vacas« as his point of departure, even if not consciously aware of his choice (example 5).
In elaborating the sang further, the initial line is reworked as the second. The first of the two sub-phrases of the first verse, marked (x) in exarnple 6, is a transposition one note lower, but the same cadential turn (y) is retained at the original pitch an harmonised identically. Further motivic resemblances in the entire melody are melodic commonplaces, there is no evidence of the initial material being consciously used to generate new phrases, and there is equally Infelice paura ad ehe venisti? d 14 Hor non bastaba amor con li suoi strali? f little to suggest that Milan concerned himself with crafting a distinctive or memorable melody. The most revealing feature of this melody, however, is a question of repetition; Milan states his melody twice but with no relationship between musical and poetic structure. As can be seen above, the initial melody (phrases a~) is repeated from verse 7, the rnidpoint of the second quatrain, but with the b-phrase thrice repeated, and a substitute line (f) replacing the e-phrase of the first half to bring the song to its condusion. According to this scheme, there is no formal musico-poetic relationship.
There are two plausible explanations of this repetition scheme. The first is that Milan was adapting an extant melody to a new poem, a melody origi.nally conceived for a different text genre, and probably with a different number of phrases. In support of this p05Sibility are the several points in the setting where the W1derlay of the text is awkward or the poetic and musical accentuation does not coincide. Secondly, there are few points in the song where melodic gestures impact upon or enhance the meaning of the words. Thirdly, the threefold repeti tion of phrase b, may have occurred to avoid exhausting the melodic material before reaching the end of the song. The other explanation that appears plausible is based on a reading of the poem that ignores theoretical structural norms. Milan may have chosen to repeat the melody, borrowed or extemporised, from verse 7 because of the pairing of the lines: Tra prosperi successi, adversa sorte; Tra soavi vivande aspro veneno.
From happy events, adverse fate, From delicate food, bitter poison. This is the principal point in the poem where melodic repetition coincides with the rhetorical device in the text, and works quite weil with the paired melodic phrases a-a'. Apart from this connection, there is little of the craft associated with sonnet setting characterised by its treatment in the hands of madrigalists and points to improvisation as the most likely compositional procedure.
A broader variety of song genres is represented in Alonso Mudarra's Tres libros. His fourteen Spanish songs include traditional romances, villancicos and several settings of sonnets in the newly adopted Italianate poetic style, complemented by four Italian sonnets, and six Latin texts including contemporary, biblical and classical poetry. They reveal a composer able to create remarkably varied songs that Wlfailingly combine charm with disarming simplicity, and show great sensitivity to their texts. His texts include works by famous contemporary and classical authors and anonymous tradi tional or popular verses.
The two songs chosen here for discussion are distinct from many contemporary songs notably for their more declamatory style. As this style is once again different from the habitual melodic style of contemporary polyphony, the questions of provenance again apply: is this something original to Mudarra, or could it in some way reflect on traditional forms of musical recitation? The songs to be discussed are the weil known romance »Triste estaba el Rey David« and the less known but stWlningly beautiful setting of Dido's lament from Virgil's Aeneid, »Dulces exuviae«. These songs are not the only ones in Mudarra's collection that exhibit these particular characteristics; other songs where the same techniques are immediately evident include the romance »Israel mira tus montes«, Jorge Manrique's famous verses »Recuerde el alma dormida«, the sonnets »<Que llantos son aquestos?« and »Si amar por el hombre«, and his settings of »Nisi Dominus« (Psalm 126) and »Exsurge quare« (Ps 43: 23-4) which are exceptionally formulaic although not set to the conventional psalrnodic recitation formulae. Example 7: Alonso Mudarra, » Triste estaba el Rey David« »Triste estaba el Rey David« is a Biblical romance that laments King David's sorrow an learning of the death of Absalom. The opening of the sang (see example 7), especially, is reminiscent of ecclesiastical recitation formulae and the kinds of singing found in contemporary popular Spanish laments, although historical origins of the latter cannot be assumed with any certainty. The remainder of the sang does not continue the recitation formula, although the song's closing cadence recalls the opening phrase. The simple melodic outline throughout the sang, given schematically in example 8, shows an archaic structure more easily associable with medieval melodic design than with sixteenthcentury norms, based an phrases that ascend and descend between clearly defined points. The opening phrase an e clearly defines the Phrygian character of the piece, and the development of the melody responds to mode 4, using the lower modal tetrachord almost exclusively throughout the sang. The use of c,g, and a as melodic resting points is also characteristic of this mode. The accompaniment of the sang matches the simplicity of the melodic line, wi th some brief passages of passing notes giving momentary relief from the otherwise chordal texture. If Mudarra were in fact trying to emulate a native style outside the written polyphonic tradition, archaic or otherwise, then his choice of harmonic language matches closely what little we know of Spanish style through the cancionero repertory and the romance formulae preserved in instrumental variations. This is seen most effectively if we consider the piece not tobe in mode 4, but in a less orthodox manner, as a form of A-minor, concluding an the dominant chord. In mode 4, the principal harmonies that open and close each of the four phrases of the music would be I, iv, III, and VI, but read in A-minor, these same chords are V, i, VII, and III, the harmonic vocabulary of both »Guardame las vacas«, the pavana-folia and many other pieces. In fact, the sequence formed by the opening chord of each of the four phrases of the music (E, A minor, G and C) is identical with the V-i-VII-III opening of the pavana-folia, the same formula on which Mudarra based his deliberately retrospective Fantasia que contrahaze /a harpa. The harmonic and melodic structure of the piece is given in the analytical reduction of the sang. Mudarra's inclusion of settings of poetry by Virgil, Horace, and Ovid in his Tres libros is unique in the early Jute sang repertory, but probably indicative of a wider practice in Spain and throughout Europe. lt immediately recalls the improvisatory practices of Italian humanists cited earlier and which we now know also to have been practised in Spain in some form. lt was from the 1539 Hofhaimer and Senfl anthology Harmoniae poeticae that Mudarra arranged Hofhaimer's setting of Horace's Beatus Ille for his vihuela book. Perhaps the simple metrical style of Horace's bucolic poem was emulated by Spanish improvisers, but Mudarra's setting of »Dulces exuviae« exemplifies a declamatory style that has no equal in the early repertory of songs accompanied by the Jute or vihuela and whose origins and traditions remain an enigma. The main feature of the melodic style -it can perhaps be styled a monodic recitation -are the declamatory passages of repeated notes, generally accompanied by tatic harmony. The most striking of these passages occurs at the beginning of the distinguish them from other known early sixteenth-century songwriting styles, there are still many unknown factors that impede a deeper understanding of their provenance whether it be popular song, recitation practices, the improvisation techniques of minstrels or humanist singers, or just these vihuelists' own conception of how the singers of antiquity performed.