ISSN: 1089-747X
Copyright © 1995–2012 by the Society for Seventeenth-Century Music

Volume 9, no. 1:

Louise K. Stein*

The Origins and Character of recitado


Spanish recitado, which emerged around the middle of the seventeenth century, owes little to Italian recitative, which was introduced to Spanish musicians and courtiers and rejected, first in 1627 and again in 1652. Spanish recitado could be in triple as well as duple meter, frequently contains recurring rhythmic formulas, distinctive melodic contours, melodic sequences, expressive devices that do not depend upon declamation, and a closely spaced chordal accompaniment. It is derived from the recitational styles sometimes used in romances as far back as the first half of the sixteenth century.

1. Recitative and Calderón’s Fortunas de Andrómeda y Perseo (1653)

2. Spanish recitado

3. The Origins of recitado

4. Vincenzo Giustiniani’s Reference (1628) to Recitative alla spagnola

5. Conclusions




Notated Musical Examples

Audio Examples

1. Recitative and Calderón’s Fortunas de Andrómeda y Perseo (1653)

“… one cannot get it into the heads of these gentlemen that one can speak singing …”

1.1 So wrote the Roman stage-architect and engineer Baccio di Bianco in 1652 in a letter to his former patron in Florence.1 Baccio was referring to the creators of court musical plays in Madrid, where he had been sent to serve King Philip IV of Spain. Examined in the light of what we know about Spanish recitative or recitado, Baccio’s comments beg for contextualization. First of all, recitative was not new in 1652, for Philip IV’s court had already been introduced to the practice of recitar cantando several decades earlier, with the production of the first Spanish opera, La selva sin amor, in 1627. La selva sin amor had been an experiment in approximating Italian recitative of the Florentine variety by setting Spanish verse written in the Tuscan poetic meters. Although Lope de Vega admitted that he was “enraptured” when he heard his poetry set to music and performed in this way, the fact remains that the two private performances of La selva sin amor and the introduction of this kind of dramatic singing went without comment at the time and did not inspire other attempts to produce opera or cultivate recitative.2

1.2 In 1652, Baccio di Bianco and Giulio Rospigliosi (papal nuncio in Madrid) managed to convince the Spaniards to experiment again with a kind of recitative in which Spanish texts with italianate seven- and eleven-syllable lines were set to music in a declamatory fashion. In Act II of La fiera, el rayo y la piedra (1652) by Pedro Calderón de la Barca, a monologue for Cupido beginning “Si el orbe de la luna,” is marked “sale Cupido cantando en estilo recitativo.”3 Significantly, it is written in what was known then in Spain as the “Petrarchan” or italianate poetic meter, in seven- and eleven-syllable lines throughout (rare for Spanish song-texts). This same poetic meter is used also for the sung dialogue between Cupido and Anteros in Act III, which probably means that this section was set in recitative as well. The stage directions suggest that recitative was exploited only for the conversations of the gods in this court play of 1652, the first that Baccio di Bianco described as having included recitativo.

1.3 The recitative composed for Calderón’s La fiera, el rayo y la piedra in 1652 has not survived, but the extant vocal music for the principal court play from the next year, 1653, does. This play, Calderón’s Fortunas de Andrómeda y Perseo, is a semi-opera preserved in a presentation manuscript that was prepared as a gift to be sent to the Hapsburg court in Vienna, and this manuscript includes not only the complete text and music, but also Baccio di Bianco’s drawings of the stage designs and sets, and important descriptive rubrics of the actions and visual effects of the first performance. The music for the play was copied into the manuscript by an Italian hand. In the dialogue that opens Act I, the gods Mercurio and Palas (half-brother and half-sister to Perseo) decide to tell the mortal Perseo the circumstances of his conception, in a gentle manner, “telling it without saying it so that Perseo will know it without knowing it” (“A decirlo sin decirlo, / y a saberlo sin saberlo”). According to the manuscript rubrics alongside the text:

Palas and Mercury, in a different manner than the mortals, began their conversation in a recitative style [un estilo recitatibo], which, being a mixture of declamation and music, was neither really music nor declamation, being rather an entoned consonance, accompanied by the choir of instruments.4

1.4 An anonymous Spanish composer (probably Juan Hidalgo) composed recitative for this scene that, in musical meter, rhythm, and melodic shapes, does not resemble Italian recitative of this period (example 1). With its many repeated notes and careful declamation of the text over a static bass line, this recitative clearly was composed in light of some idealized description of early Italian recitative, but the differences between the Spanish and the Italian approaches are immediately audible.

1.5 This recitative dialogue is set in triple meter to accommodate the accentuation patterns of Spanish octosyllabic poetry; it is almost completely lacking in expressive devices, and it facilitates rapid delivery of the wordy dialogue between the gods. This recitative offers rhetorical support for Calderón’s long expository text, which governs its musical structure and melodic shapes. The melodic line is dedicated to repeated notes within highly consonant triadic motion. Changes of register, large leaps, and cadential figures are reserved for points of poetic punctuation or structural division, and for changes in the tone or topic of the dialogue. Certain words or rhetorical highpoints are emphasized with simple melodic or rhythmic devices, precisely of the sort that the singer-composers Jacopo Peri and Giulio Caccini had shunned as artificial when they wrote about early Florentine recitative.

1.6 In Palas’s first recitative, for example, the word montes (“mountains”) is set to an upward leap of an octave; later, the words muriendo (“dying”) and lamentos (“laments”) are set to suspensions in a stylized hemiola pattern. The triple meter allows for the fluent passage from repetitive, syllabic declamation into brief instances of text painting or lyrical ornamentation, although these occur mainly at the ends of sections as part of extended cadential figures. The entire scene is melodically homogeneous, without significant moments of tension or climax, as the generally arched melodic contour rises and falls usually through repeated notes of triads, with a gradual movement toward the highest notes (except for a few large leaps as melodic punctuation or word-painting). The overall impression is of a blandly balanced, quasi-lyrical declamation.

1.7 Thankfully, this sort of recitative did not replace heightened dramatic speech throughout the whole play. Its combination of declamation and musical harmony was employed only as an appropriately elevated and harmonious medium for the language or “diction” (plática) of the gods. Mercurio and Palas, in the triple meter recitatives, were thus “saying it without saying it,” but their “estilo recitatibo” was far from italianate. Although the rubrics from the presentation manuscript describe the dialogue as “recitative,” their music entirely lacks the nobile sprezzatura di canto (“aristocratic nonchalance in singing”) so important to Giulio Caccini and contemporary Italian singers. These triple-meter recitatives from 1653, the earliest extant examples of the style in Spain, demonstrate the application of the concept of recitar cantando that Baccio di Bianco had found the Spaniards so resistant to only a year before, but remind us that Spanish composers did not actually compose their passages of sung declamation in an Italian manner.

2. Spanish recitado

2.1 Slightly later examples of recitative, from other Calderón plays, from the music for other semi-operas, and from the scores of the two extant Hispanic operas, La púrpura de la rosa and Celos aun del aire matan, tell us that Spanish composers did not continue to cultivate a recitative notated in triple meter, although they did continue to use recitative selectively and to adapt the idea of recitative to Spanish practice without emulating an italianate musical style. The Spanish term recitado is both historically more correct and descriptively more appropriate to the kind of musical “speech in song” heard in the Madrid court theaters. Recitado, a mid- to late-seventeenth-century phenomenon, is declamatory, to be sure, but it is also much more lyrical, consonant, and melodically expansive than contemporary Italian recitative, whether Florentine, Roman, or Venetian. In short, Spanish recitado is more song-like than Italian recitativo.

2.2 As examples of recitado I offer two extracts from the two surviving Hispanic operas with texts by Calderón. In the first, from La púrpura de la rosa, originally set by Juan Hidalgo in Madrid in 1659 but adapted or recomposed by Tomás de Torrejón y Velasco for Lima, 1701, we hear a passage of recitado sung by Adonis, the mortal protagonist and tragic hero of the opera.5 Throughout the opera, Adonis sings with almost excessive lyricism, and it is this quality that musically identifies and distinguishes his character, although his role was actually composed for performance by a female singer-actress. In this passage of passionate declamation, he describes the tragically violent circumstances of his conception and birth. Here the musical rhythms are entirely controlled by the text (text 1), while the melody and harmony of the recitado reveal Adonis’s heightened affective state of anxiety, as he unfolds his shameful personal story to Venus in their first scene of private and confessional intimacy (example 2, audio 1).

2.3 In the second of the Calderón-Hidalgo operas, the three-act Celos aun del aire matan of 1660, the character of the chaste goddess Diana contrasts directly with La púrpura’s gentle, overtly tuneful Adonis, yet recitado is also the vehicle for Diana’s most importantly powerful moments of direct address. When Diana addresses the mortals assembled in her temple in Act II, the lyrical and powerfully expressive character of Spanish recitado highlights her long dramatic monologue. Like Adonis’s recitado from La púrpura de la rosa, this recitado, in duple time, is characterized by a sectionalization of melodic patterns and locally affective devices corresponding to the subjects treated in each section of the text. Conventionally effective but conservative musical devices project the messages of the text (text 2) with maximum attention to meaning and structural clarity (example 3, audio 2).

2.4 Out of anxiety and anger, Diana rages because another goddess has dared to oppose her dominion in her own temple, although the exact identity of this intruder eludes her. Similar to the way in which Adonis revealed his shameful past to Venus, Diana here exposes herself in a series of self-revelatory phrases in highly symbolic language. Although Calderón’s abstract poetic language is rich in symbolism, it did not offer the composer many easy opportunities for simple word-painting or text expression (and these were, even in recitative, still a mainstay of Hidalgo’s expressive musical language). We find Hidalgo’s usual rhetorical figures—small sequential fragments, the use of descending melodic leaps for text-painting, changes in the rhythmic pattern for divisions in the text, and harmonic movement to fifth- and third-related areas, according to the sense of the text. Hidalgo employed some specific and rather simple text-painting devices; he seized on the reference to hell in m. 15 (at the words “y Proserpina en el negro centro”) with an obvious gesture by plunging the melody down to its lowest notes and the voice into the dark and possibly unnatural sound of the lower register. The same device is used again in m. 30, at the bottom of a powerful descending scale figure, for the words “pues madre de horror y miedo.” Contrarily, when Diana speaks of her strength and the “power” of her intellect, she strikes out at the highest notes of the piece, with the vocal brilliance of the top register. In bursts of confidence, Diana sings expansively of her powers (mm. 27–30, 32–33, 44–46) with arching melody and sweeping scalar figures. These dissolve with a sudden loss of momentum each time her confidence is shaken by thoughts of her opponent. When the goddess questions, laments, or exclaims in frustration, the forceful scalar figures and their goal-directed harmonic progressions are replaced (mm. 34–36, 42–43) by shorter, melodically disjunct, rhythmically and harmonically static fragments of recitation. Not until her composure is restored by her resolve for revenge, does Diana fully regain her characteristic musical force, beginning at m. 54, when she curses the “crimes” of love that the other gods have already committed with impunity in the name of her rival, Venus.

2.5 Aside from these localized gestures, Hidalgo vivified Calderón’s characterization of Diana through a symbolic use of harmony. In mm. 10 through 21, the harmony moves to the chords farthest away from the C-major tonic of the piece, dwelling on harmonies of B major, B minor, and the very unusual F-sharp major. Thus, Diana’s false and fickle influence (which she emphasizes at this point in the recitative) is represented through movement to extraordinary harmonies, to the points farthest away from harmonic stability. In this section of the recitative, we find a projection of Diana’s essential nature through harmonic movement that underlines the text with its striking color and sheer daring, although the rare harmonies are reached through musically logical means. An audible illustration of Diana’s baneful powers is brought out in mm. 24–25 with the words “cuyas venenosas plantas.” The sudden introduction of the E-major chord following the G-major chord, with the juxtaposition of G–G-sharp in the melody is obvious word-painting. But the E-major chord is a sudden obstruction in the path toward stability because it abruptly interrupts the harmonic movement around the circle of fifths in the direction of stability and the tonic C-major that Hidalgo had previously established in mm. 19–24, a jarring tribute to Diana’s mutability. As a composer of recitado, Juan Hidalgo needs no apology. He was a master at setting Spanish verse. Although his characters speak in song in a Hispanic idiom, they move our affections just as effectively as do those who sing recitative monologues composed by his near contemporaries elsewhere in Europe.

2.6 Recitado was used sparingly and even avoided in most of the Spanish musical plays (the semi-operas and especially the zarzuelas), because the standard comedia—the kind of play that the acting companies were called upon to perform almost daily—relied on a traditional musical language and style of performance. While Spanish romances were characterized by affective restraint in the treatment of their texts, the Spanish comedia was known for emphatic, exaggerated histrionic expression in its spoken roles.

2.7 Recitative dialogue in Spanish plays and operas most likely posed special difficulties. The actresses from the theatrical troupes who appeared in the royal productions were of very low social standing and were largely uneducated. Most of them had grown up in theatrical families and had learned to entertain in taverns and on the streets. They came from poverty and only a rare few retired in old age to anything but a miserable existence sustained by charity. The same actresses who sang the roles of exalted mythological deities in the semi-operas spent much of their time performing secondary spoken roles in comedias in the public theaters. It is doubtful that these actresses ever heard italianate monody and recitative in performance, and it is unlikely that the practitioners of the histrionic art in Spain would trade the glory of their dramatic spoken declamations for the complication and restriction of wholly sung roles. The traditional singing styles, considerably influenced by popular music, could be brought into the palace plays because they did not require complete musical literacy of the singers. Perhaps because the singer-actresses learned the music for their roles by rote, each of the two Hispanic operas performed for the court in 1660 was delayed and required an inordinate amount of rehearsal.

3. The Origins of recitado

3.1 By focusing on the relationship between the first performers of Spanish recitado (the actress-singers) and its music, we move closer to explaining its special, salient musical characteristics and its origins within Hispanic practice. The actress-singers for whom the sung roles in Spanish semi-operas and operas were composed had grown up singing Spanish songs, in particular those ballads known as romances and the sung dances known as bailes. Both romances and bailes were incorporated into plays for the court and public theaters as the focus of musical scenes. Indeed, the conventions of theatrical music and of musical performance in the theater were linked inextricably to the history of the romance as a musical and poetic genre.

3.2 The origin of recitado is to be found in the performance practice of the romances and bailes, and not in the few brushes with italianate recitative or the infrequent experiences of Italian monody that the court composer Juan Hidalgo might have had. The reported advice of a few pushy Italians who visited Hidalgo in his studio did not take precedence over the home-grown, everyday, and widely cultivated Spanish practice of singing declamatory and decorously melodious songs to richly colored, closely spaced chordal accompaniments. Calderón and Hidalgo were working with Italian stage designers, but, as I have demonstrated elsewhere, the imprecise, descriptive verbal advice that a non-musician like Baccio di Bianco gave to Hidalgo as he leaned over his shoulder did not cause Hidalgo to compose in an Italian manner. Surely, had Hidalgo decided to imitate a foreign musical style, he would have seen his precious rehearsal time evaporate even more rapidly as the actress-singers, who could barely read music, struggled to learn their roles.

3.3 The antecedents for recitado come forth very clearly in the late sixteenth century, in the performance practices of the Spanish romances, which belonged to a large, constantly circulating, orally transmitted repertory that was at once “high” courtly music, preserved in written polyphonic songbooks, and popular music performed by everyone from the skilled improviser to the amateur at home, and from street musicians to professional actors and actresses. In the latter mode of performance, the romances were declaimed or recited with simple melodies to occasionally polyphonic but mostly chordal accompaniments played on the vihuela and, later, on the Spanish guitar.

3.4 The melodic shapes of the sixteenth-century romance are generally consistent from tune to tune, but no single description of their vocal style fits the diversity of their melodies. It is safe to say that the melodies are generally consonant and conjunct, with repeated notes and some degree of recitation. The romances generally set their texts in a declamatory, mostly syllabic fashion, in duple meter. To this extent they resemble the early stile recitativo. I have chosen four musical examples in order to make a few points about the diversity among the romances and their array of possible performance styles.

3.5 The song Mira Nero de Tarpeya , identified as a “romance viejo” by the Spanish music theorist Juan Bermudo in 1555, was later included in seventeenth-century plays by Lope de Vega and Calderón.6 Bermudo’s setting for voice and vihuela includes the somewhat old-fashioned romance text that lent an antique quality to the scenes in the plays, on the Nero story from ancient Roman history. There are three extant musical settings of this text, but they are not similar enough to suggest that they were all based on a common model. Bermudo’s setting (figure 1, example 4) presents what seems to be an ornamented version of the romance tune, with a thinly textured, pseudo-polyphonic accompaniment. The music both respects the poetic structure of the text (the division into lines) and supports the syntax, with overlapping levels of continuity. For instance, the first cadence (V–I, A minor) indeed breaks up the poetic line, since it occurs after only two words “Mira Nero.” But both the structure and the grammar of the text “Mira Nero de Tarpeya” are preserved in that the melody continues without a notated rest until the end of the poetic line. The continuation of the sentence “Mira Nero de Tarpeya / a Roma” (“Nero looks from Tarpeia to Rome”) is conveyed in spite of the rest by the continuation of the same harmony, starting the phrase “a Roma” on the same notes in both voice and accompaniment. The formulaic structure of the song is only disrupted by the repetition of the text “cómo se ardía,” and its cadence to C at m. 19. The phrase “cómo se ardía” is stated and then extended through a consequent musical phrase, and this process might indicate an interpretation or embellishment of the original model, or of a borrowed tune. To judge from the cadence points and the melody, the second phrase of “cómo se ardía” could easily be omitted; if so, the structure of the text would be preserved in the simplest manner, and the cadence on D at m. 15 would lead directly to the phrase “gritos dan” beginning on the same notes at m. 20. Bermudo’s treatment of the text, however, rhetorically highlights the most emotionally charged words in the quatrain, “how it burned,” emphasized through textual repetition and musical elaboration to mean “oh how it burned!” As do many other romances, this song has a rounded form, in that the opening phrase returns for the last line of text at m. 27 and is closed off with a five-note conjunct descent. The elements of Bermudo’s setting that I suspect to be original elaborations might give us some idea of how an older text and tune could be recomposed, but in its melodic shapes, its clear phrase structure, and the way the melody and the bass line are linked together, this setting has some fundamental qualities that also appear in other romances.

3.6 La mañana de San Juan (example 5), from Diego Pisador’s Libro de música de vihuela (Salamanca, 1552), is a particularly lyrical romance de moros, although it has a restricted range and considerable use of repeated notes in the melody. Typical of such narrative romances on historical subjects, the setting preserves the structure of the poem and guides the clear declamation of the text. In the first quatrain, which is what Diego Pisador provided in his vihuela book, there are no salient moments of affective emphasis, although the ends of lines are given some sort of cadential inflection or ornament. Pisador’s written-out accompaniment does not intabulate a pre-existing polyphonic version, but it illustrates the use of closely spaced chords. The performance I have chosen for audio 3 also benefits from the singer’s fluency with the text and the flexibility of both the singer and the continuo player in adapting the setting to accommodate the sense of successive strophes of text. The text used in this performance is longer than that provided by Pisador; it comes from a contemporary poetic collection or romancero. It seems to me that Pisador’s romance is a good example of a strophic ballad that remained tuneful and atmospheric in spite of its partly recitational character.

3.7 Río de Sevilla dates most likely from the last decade of the sixteenth century, and is a romance nuevo of a popular or pseudo-popular type, with a text in seguidillas. It exists in at least four different late sixteenth-century musical settings: in a manuscript that may be the earliest known collection of Spanish romances with alfabeto notation for guitar chords; in a second Italian poetic manuscript with alfabeto notation; in a polyphonic setting for three voices in a Spanish songbook found in Turin (example 6); and finally as published by Gabriel Bataille with notation for voice and French lute tablature in his 1609 collection of Airs de differents autheurs.7 In all of these settings, the harmonies are closely spaced, the rhythm of the accompaniment matches that of the vocal line creating homorhythmic declamation, and the effect is of a placid, recitational melody supported by an uneventful but colorful chordal accompaniment. The seguidillas text is in the short poetic lines (mainly of five, six, and seven syllables) characteristic of this baile.

3.8 Desde las torres del alma (example 7), is a romance nuevo setting by Juan Blas de Castro, the favorite court composer of songs in the early seventeenth century. It has come down to us in a polyphonic setting, which in its recorded performance by Montserrat Figueras (audio 4) has been reduced to a performing texture of voice in recitational style over an accompanimental bass.

4. Vincenzo Giustiniani’s Reference (1628) to Recitative alla spagnola

4.1 While the written-out accompaniments in the vihuela books and the alfabeto tablatures that are given with a number of Spanish songs in Italian manuscripts help us to understand the important relationship between traditional Spanish accompaniment techniques and the earliest Neapolitan and Roman styles of basso continuo practice, I hope to show as well that antecedents for the vocal style of recitado are also to be found in these traditional and popular Spanish musics. As John Walter Hill has pointed out, a connection between a specifically Spanish kind of music, a Spanish performance style, and a new, modern solo style of singing in Italy seems to have been posited by Vincenzo Giustiniani in his famous Discorso sopra la musica, which was written most likely within the year following the experiment of setting Spanish texts in Italian meters to italianate recitative in La selva sin amor. Giustiniani wrote concerning the new Italian male singers:

Having abandoned the past style, which was quite rough, and also the excessive passage work with which it was ornamented, [good musicians] now cultivate, for the most part, a recitative style embellished by grazia and by ornaments appropriate to the conceit, along with an occasional melismatic passage drawn out with judgment and articulation, and with appropriate and varied consonances.… Above all, they make the words well understood, applying to each syllable a note now soft, now loud, now slow, and now fast, demonstrating with their face and in their gestures the conceit that is sung, but with moderation and not excessively. And one voice, or at most, three voices sing to the accompaniment of one’s own instrument: a theorbo, guitar, harpsichord, or organ, according to the circumstances. Furthermore, within this style there has been introduced singing alla spagnola or all’italiana, similar to the foregoing but with more artifice and ornament, as much in Rome as in Naples and Genoa, with the invention of new arie and of new ornaments.8

4.2 It is surprising that as of 1628 Giustiniani seems to have included “singing alla spagnola” within the general category of “recitative style,” whereas the members of the royal court and household in Madrid did not recognize anything familiar in the Italian-style recitative written by the Bolognese Filippo Piccinini for La selva sin amor when they heard it in 1627. Giustiniani, however, was capable of finding a common recitational element in liturgical psalmody, singing formulas used in performances of sacre rappresentazioni, the solo madrigals of Giulio Caccini and of Giuseppino Cenci, the operatic recitative of Florence and of Rome, and the declamatory aspects of the Spanish romance.

4.3 Hill has suggested that Giustiniani was writing about declamatory performances of Spanish romances in which a solo singer was accompanied by a chordal instrument, and he has pointed out that this traditional Spanish performance style could have been well known to Giustiniani.9 Spanish romances enjoyed considerable popularity in Italy, along with the vogue for the Spanish guitar, and they circulated principally in poetic manuscripts with shorthand alfabeto notation for guitar chords, or a modified tablature (Río de Sevilla exemplifies the sort of romance that circulated in Italian sources as well as in Spanish ones). But I wonder if in the passage quoted above, Giustiniani was referring not only to solo songs, but to the performance of romances in declamatory polyphony as well, because he mentioned “one to three” voices singing at the same time. Could it be that Giustiniani wanted to say something favorable about Spanish music for political reasons? Elsewhere in his Discorso, he wrote about music at the court of the young King Philip IV, stating that:

… in our times, music has become more noble and illustrious than ever, since King Philip IV of Spain and both his brothers take great pleasure in it and are accustomed frequently to sing and play viols together, with a few other musicians to make up the necessary number, among these Filippo Piccinino, Bolognese, a most excellent performer on the lute and the pandora. And further, the king and his brothers write compositions not only for their own pleasure but also to be sung in the royal chapel and other churches while the divine offices are celebrated. This inclination and pleasure of his Majesty will be the reason why many gentlemen still delight in it, and many others apply themselves to music.10

4.4 These comments are strikingly similar to those of another Italian observer, one who actually spent time at young Philip IV’s court. In a letter of 1 July 1627, Averardo Medici, a member of the Florentine delegation that produced La selva sin amor, wrote from Madrid to Andrea Cioli in Florence that:

The king delights in music and understands so much about it that he knows how to compose counterpoint, and he plays well on the bass viol; and every evening His Majesty and the Infanti, his siblings, are entertained for an hour with a consort of viols, in which all three play, along with the maestro di cappella and an Italian chamber musician of His Majesty whose name is Filippo Piccinini.11

Given the concordance between Giustiniani’s comments about Philip IV and those in this letter, I wonder if Giustiniani knew about the reports that the Florentine diplomats sent from Madrid, and might have heard or read positive remarks about the experiment of setting Lope de Vega’s Spanish verse in Italian meters as recitative in La selva sin amor.

4.5 In light of the nature of the surviving repertory of Spanish songs from the early seventeenth century, Giustiniani’s references to a Spanish kind of recitative are highly suspicious. The sources of Spanish secular song from the early seventeenth century demonstrate a decided preference for decorous vocal polyphony, as opposed to monody, especially in the educated circles of the Spanish elite that Giustiniani wrote about. It is possible, of course, that Giustiniani was not writing about the new Spanish songs or “romances nuevos,” but about the performance of slightly older Spanish songs known as the “romances viejos.” For these, the six printed vihuela books published in Spain between 1535 and 1577 are our most important sources for extant music and for performing practices. Intended largely for an amateur musical public anxious to bring into the realm of social music-making the latest in fashionable and popular musical styles, the vihuela books preserve both reductions for voice and vihuela of polyphonic settings of romances, and settings in which the singer performs the romance over a simpler, more obviously chordal accompaniment. Because the compilers of the vihuela books were intent on capturing as much as possible about the performance practice in the notation, it can be argued that these tablatures contain examples of what was commonly done and passed on through practice. Among extant Spanish pieces for voice and accompaniment, the sixteenth-century settings of romances in the vihuela books provide the most important antecedent for Spanish recitado, and perhaps Giustiniani was thinking of these older romances and their performance practice when he wrote about the performance of Spanish recitativo.

4.6 On the other hand, the Spanish songs current in 1628 were known as “romances nuevos” and are preserved in polyphonic manuscript songbooks or cancioneros, which contain only part of a larger, now lost corpus of well-known courtly poems. These are settings of vernacular poetry (mostly amorous romances) of a high quality composed by the generation of court composers who grew up in the late sixteenth century but worked in Madrid at court in the early decades of the seventeenth century. None of these composers is credited with having composed recitative or recitado. The romances nuevos are part of a repertory not only that was bound by conventions but in which musical borrowing and recomposition were essential. Because two or more different musical settings of the same text usually share a core of common material (a common tune and characteristic rhythmic figures), it is probably safe to say that they are independent settings of a well-known tune closely identified with its text.12

4.7 If Giustiniani was writing about the contemporary romances nuevos in his Discorso of 1628, he might have mentioned them precisely because they exemplified the fusion of contemporary popular (even folkloric) elements current in everyday musical practice with the technical artifice and aesthetic decorum of polyphonic art music. In their written settings, the professional composers whose works were collected into the cancioneros were not content merely to provide the well-known tune with a correct harmonization, but worked pre-existent melodic and rhythmic material into a new and original polyphonic setting. In this way, the “art” of “composing” songs incorporated what was natively Spanish, popular, and in keeping with the artistic trends of the day. This art is related to the long-standing, continuing practice of instrumental glosas or glosses and diferencias (both improvised and composed). Moreover, it is very similar to that of contemporary poets, who based many of their poems on borrowed material (popular refrains, stylish conceits, borrowed verses), retaining a natural flavor and a popular spirit while exhibiting technical sophistication. One of the most successful of the courtier poets, Antonio Hurtado de Mendoza, praised the vocal music for the Count of Villamediana’s court play La gloria de Niquea (1622) by writing that its composer (most likely the maestro de capilla Mateo Romero) tempered “the crispness of parchment with the sweetness of the guitar,” which is to say that the composer infused the elegance of the true compositional art, contrapuntal polyphony, with the naturally sweet, popular air of the well-known tunes.13

4.8 Both the old and new romances exhibit a close relationship between textual and musical structures with short, clear-cut musical phrases, though each song has its own easy-to-remember tune and characteristic rhythmic figures. The tendency is toward homorhythm, within a correct counterpoint without extended points of imitation or contrapuntal elaboration. Phrases usually begin with a rest on the downbeat, and close with conventional cadential patterns. The tunes embedded in the more modern settings are generally confined in range and melodic movement, and do not impress us as strongly independent melodic lines, especially since the tune is often shared between the two highest voices in a three- or four-voiced arrangement. Perhaps most striking is the affective restraint that they demonstrate. Central features of the sixteenth-century romance (the strophic form, simple phrase structure, constant rhythmic continuity between bass and upper voices) became conventional and characteristic in many seventeenth-century Spanish songs as well. Early seventeenth-century settings of romances such as “Madre la mi madre,” “¡Cómo retumban los remos!,” “Por la puente Juana,” and “Mal segura zagaleja” present a homophonic, homorhythmic arrangement, in effect more chordal than contrapuntal, with short, clearly articulated musical phrases that follow the structure of the poetic text. (See the score excerpt in example 8 for “Cómo retumban los remos,” which is performed complete in audio 5) While only a small group of these songs exists today in more than one musical setting, the presence of the same tune and rhythm in two or more settings of the same text points to a tradition of musical borrowing and leads to the conclusion that most of the simple settings with well-known texts are arrangements based on well-known, pre-existent tunes.

5. Conclusions

5.1 Because I find most romances to be more song-like than recitational, it surprises me to read that Giustiniani included singing alla spagnola and all’ italiana in the same sentence. But perhaps the importance of the Spanish romance to his discussion of early recitative had to do not only with the declamatory character of the romance but with Spanish performance practice. Romances could be performed as solo songs with strummed chordal accompaniment or by groups of two or three singers, just as described by Giustiniani in his sentence: “And one voice, or at most, three voices sing to the accompaniment of one’s own instrument: a theorbo, guitar, harpsichord, or organ, according to the circumstances.” As I have noted elsewhere, the pervasive Spanish tradition of accompanied solo song, which indeed has a considerable presence in Italian, French, German, and even English manuscripts with tablature and/or alfabeto notation from the late sixteenth and early seventeenth centuries, provides a clear antecedent for Spanish continuo practice.14 Musical, iconographical, and literary sources attest to the unbroken continuity of this tradition, in which romances new and old were performed to the accompaniment of a polyphonic instrument, preferably the strummed guitar or the harp. While accompaniment on a polyphonic instrument was not peculiarly Spanish, strummed rasgueado accompaniments were thought of as Spanish in style. The Spanish taste in accompaniment on a polyphonic instrument, whether strummed or plucked, called for two principal traits: that the vocal line not be doubled in the accompaniment, and that the accompaniment be improvised largely with closely spaced chords, without a need for either root-position chords or proper voice-leading. I think it can be shown that the practice of such a rhythmically dependent accompaniment of closely spaced sonorities was particularly associated with Spanish harpists and guitarists, rather than with players of the lute, theorbo, and harpsichord. Most of the written-out accompaniments preserved in Hispanic sources share these characteristics, and the late seventeenth-century advice of the first Spanish continuo treatises and musical theorists also emphasizes them.15

5.2 This style of accompaniment points back to the accompaniments for solo songs in the sixteenth-century vihuela books. In many of the Spanish songs in the vihuela books that have newly-composed accompaniments, the virtuosity of the instrument is entirely absent. The vihuela parts provide closely-spaced chords to support the voice, and virtuoso figurations are restricted to sound only when the voice sustains tied notes in the melody or rests between vocal statements. This separation of instrumental virtuosity from functional accompaniment was important enough that Luis de Milán instructed performers: “Here the romances begin, and to perform them one must play the chords slowly, but play very quickly the [instrumental] flourishes at the ends of phrases when the voice finishes.… The singer has to hacer garganta [sing ornaments or passagework] when the vihuela is not making redobles [ornaments, diminutions or figuration].”16 The continuo player’s aim was to provide an artfully servile, full accompaniment of closely-spaced sonorities, without doubling the voice, and adding idiomatic instrumental figurations only beneath the voice’s tied notes and in the pauses between vocal phrases.

5.3 With this description of the accompaniment to a romance in mind, the musical practices of the romance can easily be heard as the origins of Spanish recitative. But beyond merely reaffirming John Hill’s points, I wish to suggest that Spanish declamatory song did not become italianate recitative (in the seventeenth century), but instead developed on its own terms as recitado. In adopting and adapting the Italian concept of theatrical and operatic recitative to the practice of Spanish singer-actresses, Juan Hidalgo invented recitado, which depended on the tradition of the romance not only for its accompanimental style, but also for its melodic practices. Spanish recitado is far too melodious, too song-like, and rhythmically and harmonically consonant to have been modeled on the musical characteristics of any kind of imported recitative. When he invented recitado, Juan Hidalgo did what came naturally to a composer steeped in the traditions of the romance, in which restrained affective singing and vocal declamation were compatible with each other and with rhythmically animated, song-like stretches of largely conjunct melody. Affectively charged declamation did call for a special kind of accompaniment and more forcefully expressive melodies, but it did not mandate the cultivation of the “tedium of the recitative.”


* Louise K. Stein ( is Professor of Musicology at the University of Michigan. Her first book, Songs of Mortals, Dialogues of the Gods: Music and Theatre in Seventeenth-Century Spain (Oxford University Press, 1993), was awarded a publication subvention from the American Musicological Society, as well as the1995 First Book Prize from the Society for Spanish and Portuguese Historical Studies. Her second book is an expanded second edition of Howard Mayer Brown, Music in the Renaissance (Prentice-Hall, 1998). Her performing edition of the first opera performed in the Americas, La púrpura de la rosa (Lima, 1701), by Tomás de Torrejón y Velasco and Juan Hidalgo, was published in Madrid in 1999 and used for the BMG Classics recording directed by Andrew Lawrence-King, for which she also served as artistic advisor and dramaturg. In 1996 the American Musicological Society recognized her with the Noah Greenberg Award for “distinguished contribution to the study and performance of early music.”

1 “… non può entrar nel capo a questi signori che si possa parlar cantando…,” extracted from letter quoted in Lorenzo Bianconi and Thomas Walker, “Dalla ‘Finta pazza’ alla ‘Veremonda’: storie di Febiarmonici,” Rivista italiana di musicologia 10 (1975): 452–3; and Louise K. Stein, Songs of Mortals, Dialogues of the Gods: Music and Theatre in Seventeenth-Century Spain (Oxford: Oxford University Press, 1993), 136–7.

2 Stein, Songs of Mortals, 185–205; see also Shirley B. Whitaker, “Florentine Opera comes to Spain: Lope de Vega’s La selva sin amor,” Journal of Hispanic Philology 9 (1984): 43–66; and Lope de Vega, La selva sin amor, ed. Maria Grazia Profeti (Florence: Alinea, 1999), 7–11.

3 Stein, Songs of Mortals, 134–8.

4 Quoted and translated in Stein, Songs of Mortals, 150; the original source is US-CAh MS Typ 258H, fol. 22v.

5 The history of this opera is explained in the Introducción to Tomás de Torrejón y Velasco and Juan Hidalgo, La púrpura de la rosa, ed. Louise K. Stein (Madrid: SGAE, 1999).

6 Juan Bermudo, Declaración de instrumentos musicales (Osuna, 1555), fol. ci, transcribed in Stein, Songs of Mortals, 111–112; the plays are Lope de Vega, Roma abrasada, Act III; and Pedro Calderón de la Barca, Séneca y Nerón, Act III. The musical sources for the three different settings of this romance text are: Bermudo, fol. 101r; Mateo Flecha, “Ensalada del Fuego,” Las ensaladas de Flecha (Prague, 1581), ed. Higinio Anglés (Barcelona: Diputación Provincial de Barcelona, 1954), 71–72; Luis Venegas de Henestrosa, Libro de cifra nueva para tecla, harpa, y vihuela (Alcalá, 1557), Monumentos de la música española, vol. 2, ed. Higinio Anglés (Barcelona: Instituto español de musicología, 1944), 151–52; the latter setting is ascribed to Palero.

7 These four sources (I-Rvat Chigi Cod. L.VI.200, p. 28, “Libro de cartas y romançes españoles”; I-MOe MS EST. 2=P.6.22, song 11; I-Tn MS. R.1–14, fol. 27r; and Airs de differents autheurs) are considered in Stein, Songs of Mortals, 354–60, 397. The Italian manuscripts are described also in Giovanni Maria Bertini and Cesare Acutis, La romanza spagnola in Italia (Turin: G. Giappichelli, 1970); and in John H. Baron, “Secular Spanish Solo Song in Non-Spanish Sources, 1599–1640,” Journal of the American Musicological Society 30 (1977): 20–42. The history and significance of the Chigi manuscript that belonged to the Duchess of Traetta have been explained in elegant detail by John Walter Hill in his Roman Monody, Cantata, and Opera from the Circles around Cardinal Montalto, 2 vols. (Oxford: Clarendon Press, 1997), 1:70–75. See also James Tyler’s article in this volume, “The Role of the Guitar in the Rise of Monody: The Earliest Manuscripts,” par. 2.4.

8 Translation from Hill, Montalto, 1:110.

9 Hill, Montalto, 1:69–75; see also Baron, “Secular Spanish Solo Song,” and other contributions to the literature cited therein.

10 Vincenzo Giustiniani, “Discorso sopra la musica de’ suoi tempi,” (Rome, 1628), ed. Angelo Solerti, Le origini del melodramma (Turin: Bocca, 1903; reprint, Hildesheim: Georg Olms, 1969), 111–112; trans. Carol MacClintock, Il disiderio; or, Concerning the Playing Together of Various Musical Instruments, Musicological Studies and Documents 9, (Rome: American Institute of Musicology, 1962), 71–72.

11 “… il Re si diletta della musica et n’intende tanto che sa comporre di contrappunto et suona franco il basso del violone; et ogni sera si trattengono Sua Maestà et gli Infanti suoi fratelli un’hora con un concerto di viole, che tuttiatre suonano, et con loro il maestro di cappella et un Italiano musico di camera della Maestà Sua che si chiama Filippo Piccinini.” Document transcribed and quoted in Whitaker, “Florentine Opera comes to Spain,” 63.

12 Concerning the romances, their musical settings, and their place in the theater in early seventeenth-century Madrid, see Stein, Songs of Mortals, 14–48 and passim; Ramón Adolfo Pelinski, Die weltliche Vokalmusik Spaniens am Anfang des 17. Jahrhunderts: Der Cancionero Claudio de la Sablonara (Tutzing: Hans Schneider, 1971); Luis Robledo, Juan Blas de Castro: Vida y obra musical (Zaragoza: Institución Fernando el Católico, 1989); and The Cancionero de la Sablonara, A Critical Edition, ed. Judith Etzion (London: Tamesis Books, 1996).

13 “… que a su maestro le debe la música haber juntado en los tonos la destreza, y el buen ayre de cantar, ajustando lo crespo del facistol, a lo dulce de la guitarra” (Antonio Hurtado de Mendoza, Fiesta que se hizo en Aranjuez a los años del Rey Nuestro Señor D. Felipe IIII [Madrid, 1623], fol. 15v).

14 Louise K. Stein, “Accompaniment and Continuo in Spanish Baroque Music,” in Actas del Congreso internacional sobre España y la música de occidente, Salamanca 1985, 2 vols. (Madrid: Ministerio de Cultura, 1987), 1:357–70.

15 Stein, “Accompaniment and Continuo,” 361–70.

16 “Aquí empieçan los romances y hanse de tañer lo que fuere consonancia a espacio, y los redobles que hay a las finales cuando acaba la voz muy apriesa … El cantor ha de hazer garganta cuando la vihuela no haze redobles” (Luis de Milán, Libro de música de vihuela de mano: Intitulado El Maestro [Valencia: F. Diaz, 1536]; ed. Leo Schrade [Leipzig: Breitkopf & Härtel, 1927; reprint, 1967], 358, 364; cited and translated in Stein, “Accompaniment and Continuo,” 364).


Text 1: From Pedro Calderón de la Barca, La púrpura de la rosa

Text 2: From Calderón de la Barca, Celos aun del aire matan


Figure 1: “Mira Nero de Tarpeya,” romance setting in Juan Bermudo, Declaración de instrumentos musicales (1555), fol. 101r

Notated Musical Examples

Example 1: “Estilo recitatibo” opening of dialogue for Palas and Mercurio, Act I of Pedro Calderón de la Barca, Fortunas de Andrómeda y Perseo (1653) 

Example 2: Adonis, section of recitado from Tomás de Torrejón y Velasco (1701) and Juan Hidalgo (1659) La púrpura de la rosa 

Example 3: Diana, recitado from Act II of Juan Hidalgo, Celos aun del aire matan (1660), libretto by Calderón de la Barca

Example 4: “Mira Nero de Tarpeya,” romance setting from Juan Bermudo, Declaración de instrumentos musicales (1555)

Example 5: “La mañana de San Juan,” romance de moros from Diego Pisador, Libro de música de vihuela (Salamanca, 1552)

Example 6: Anonymous musical setting of the romance “Río de Sevilla,” from I-Tn MS. R 1-14, fol. 27
a) Transcription of the romance as presented a 3 in the Turin manuscript
b) The same romance as a series of guitar chords
c) The same romance in an arrangement with the top voice isolated as a vocal melody

Example 7: Opening copla of Juan Blas de Castro, “Desde las torres del alma,” romance a 4

Example 8: The opening of “¡Cómo retumban los remos!” anonymous romance a 3

Audio Examples

Audio 1: Adonis, recitado from Tomás de Torrejón y Velasco (1701) and Juan Hidalgo, La púrpura de la rosa, libretto by Calderón de la Barca

Audio 2: Diana, recitado from Act II of Juan Hidalgo, Celos aun del aire matan (1660), libretto by Calderón de la Barca

Audio 3: “La mañana de San Juan,” romance de moros from Diego Pisador, Libro de música de vihuela (Salamanca, 1552)

Audio 4: Opening copla of Juan Blas de Castro, “Desde las torres del alma,” romance a 4

Audio 5: The opening of “¡Cómo retumban los remos!” anonymous romance a 3

Copyright Statement

Copyright © 2003 by the Board of Trustees of the University of Illinois. All rights reserved.

[1] Items appearing in JSCM may be saved and stored in electronic or paper form, and may be shared among individuals for purposes of scholarly research or discussion, but may not be republished in any form, electronic or print, without prior, written permission from the University of Illinois Press, and advance notification of the editors of JSCM.

[2] Any redistributed form of items published in JSCM must include the following information in a form appropriate to the medium in which the items are to appear:

This item appeared in The Journal of Seventeenth Century Music in [VOLUME #, ISSUE #] in [MONTH/YEAR], and it is republished here with the written permission of the University of Illinois Press.

[3] Libraries may archive issues of JSCM in electronic or paper form for public access so long as each issue is stored in its entirety, and no access fee is charged. Exceptions to these requirements must be approved in writing by the editors of JSCM and the University of Illinois Press.

[4] Citations to articles from JSCM should include the URL as found at the beginning of the article and the paragraph number; for example:

Noel O’Regan, “Asprilio Pacelli, Ludovico da Viadana and the Origins of the Roman Concerto Ecclesiastico,” Journal of Seventeenth-Century Music 6 (2000) <>, par. 4.3.

This document and all portions thereof are protected by U.S. and international copyright laws. Material contained herein may be copied and/or distributed for research purposes only.