(It.: ‘song’).
A type of instrumental music of the 16th and 17th centuries that developed from the Netherlandish chanson.
JOHN CALDWELL
The spelling ‘canzona’ was fairly frequently used in Italy after 1600 and has become standard in England and not infrequent in Germany; in older Italian sources, however, ‘canzone’ and ‘canzon’ (with the plural ‘canzoni’) are practically universal, and ‘canzone’ has subsequently remained the standard Italian form. It should be noted that ‘canzone’ as the plural of ‘canzona’ is rare. (For a discussion of the Italian poetic form, see Canzone.) The word ‘canzone’ or ‘canzona’ in its instrumental connotation originally denoted an arrangement of a polyphonic song, usually a French chanson, since although arrangements of Italian works were quite common these were usually called ‘frottola’ or ‘madrigale’. Although it was used at least until the end of the 16th century to mean a straightforward arrangement, there are quite early instances of new compositions based on existing chanson material, and the term eventually came to be applied to original compositions using idioms familiar through arrangements and reworkings. Since chansons of the type favoured for these purposes (i.e. the Parisian chanson as represented in the books of Attaingnant starting in 1528) frequently began with fugal imitation, the canzona came to be considered a fugal genre. It is described by Praetorius (Syntagma musicum, iii, p.17) as a series of short fugues for ensemble of four, five, six, eight or more parts, with a repetition of the first one at the end (although this feature is in fact rare; see Fugue, §4). Expressions such as ‘canzon francese’ and ‘canzone alla francese’ appear to indicate nothing beyond the form just described (they are used both of arrangements and of original works), while ‘canzon da sonar’, a phrase that played a part in the genesis of the term ‘sonata’ (see Sonata, §I), specified only that it was an instrumental (usually ensemble) form.
Arrangements of polyphonic vocal works for keyboard are much older than the 16th century and embrace sacred as well as secular forms. The transcriptions of Netherlandish chansons in Spinacino's two lutebooks of 1507 do not differ essentially from the general type of ornamented keyboard or lute arrangement prevalent in the early 16th century and found also in the Capirola Lutebook (c1517). More significant are the arrangements of two frottolas in Dalza's Intabulatura (1508) and those for keyboard published by Antico in 1518, not because the technique of arrangement was in itself modified, but because the chordal and harmonically conceived style of the frottola corresponds more closely than that of the newer type of chanson (in spite of the latter's penchant for imitative openings) to the transcriptions to which the term ‘canzone’ was originally applied. The earliest examples of such arrangements are found in M.A. Cavazzoni's Recerchari, motetti, canzoni … libro primo (1523). It contains four pieces with French titles, one of which has been identified as an intabulation of Josquin's Plusieurs regretz (Picker, 1972). Two celebrated chansons were included in Girolamo Cavazzoni's Intavolatura cioe recercari canzoni himni magnificati (1543), Josquin's five-part Faulte d'argent and Passereau's Il est bel et bon. Cavazzoni's arrangement of Faulte d'argent is much more than a mere transcription of the original: it is a complete reworking, omitting Josquin's canon between the contratenor and quinta parts, but retaining the ternary structure of the piece and adding much that is new and interesting. Scarcely a single bar is identical with the original. It is also a masterpiece of compression, with Josquin's 72 bars reduced to 56. Passereau's chanson is treated in a similar way.
The earliest published lute transcriptions to which the term ‘canzone’ is given, either to individual pieces or by implication in the titles of the publications in which they occur, appear to be those of Melchiore de Barberiis (Intabulatura de lauto libro quinto), Domenico Bianchini, Francesco da Milano (two books, both reprints of lost originals), G.M. da Crema and Antonio Rotta (all of 1546). The publications of Barberiis include two pieces entitled simply ‘Canzun [or ‘canzon’] francese’ without any further qualification, and these may be the earliest freely composed canzonas of any kind. There are further examples in his later book (1549), although the identification of one of them as a chanson by Sermisy tends to cast doubt on the ‘original’ character of the others. These transcriptions of the newer type of French chanson must be seen in the context of the many published at about this time without the distinguishing title, both in Italy and elsewhere. The lute canzona, however, never became very common either as an arrangement of a vocal model or as a purely instrumental form. Those in G.A. Terzi's Intavolatura (1593) are merely arrangements of Florentio Maschera's ensemble canzonas (see §3 below).
A good many 16th-century keyboard canzonas are nothing more than elaborated transcriptions of chansons, for example those of Andrea Gabrieli and Sperindio Bertoldo. Often the ornamentation is very profuse, perhaps slowing down the natural speed of the plain original. One of the inheritors of this method was Peter Philips, who applied it to both Italian and French originals. Andrea Gabrieli used the terms ‘ricercare’ or ‘capriccio’ to denote a free reworking of material from a chanson or madrigal.
One of the first composers to write canzonas independently of vocal models was Claudio Merulo. Five of his 23 canzonas (published in three books in 1592, 1606 and 1611 respectively) are transcriptions of chansons. The remainder bear such titles as La Bovia, La Zambeccara and La Gratiosa. Many of these may originally have been ensemble works: four reappear in this form in Alessandro Raverii's Canzoni per sonare (1608), ten are in Johann Woltz's Nova musices organicae tabulatura (1617), in which the canzonas are mostly transcriptions of ensemble pieces, and others survive in ensemble form in manuscript. Nevertheless, their appearance in keyboard form from 1592 onwards must inevitably have paved the way for the independently conceived keyboard canzona. The first composer of such works was apparently Vicenzo Pellegrini, whose Canzoni de intavolatura d'organo fatte alla francese appeared in 1599. Like those of Merulo, they bear titles such as La Berenice and La Gentile. Pellegrini's canzonas were the first keyboard works in several sections clearly defined by contrasting speeds and metres; but these sections were not yet variants of the same material. From this point the canzona became very largely an independent form of keyboard music.
Numerous transcriptions for keyboard of ensemble canzonas, by composers such as Antonio Mortaro, Francesco Rovigo, Cesare Borgo and Ottavio Bariolla, together with original keyboard canzonas, appear in vols.x–xi (Foà, 1, 3), of the Turin tablatures (see Sources of keyboard music to 1660, §2(iii)). They include, in vol.xi, seven by Giovanni Gabrieli, of which the second is an intabulation of an ensemble canzona published by Raverii in 1608 and the fifth is a keyboard ricercare published in 1595. The other five are apparently true keyboard canzonas and probably authentic, although one of them appears in the Woltz tablature of 1617 ascribed to Adam Steigleder and the last two are ascribed to Erbach elsewhere. They show the influence of the ensemble canzona in their alternation of pairs of voices to create an antiphonal effect and in their use of repeated chords. The second, in a ritornello form with five statements of the main section, is particularly interesting. The others are each in a single section.
Other early 17th-century composers of keyboard canzonas include Banchieri, G.P. Cima, Ercole Pasquini, Macque, Mayone and Trabaci. Trabaci's canzonas (seven from his first book, 1603) are in numerous sections, and in the sixth the sections are almost for the first time related thematically to form what has come to be known as a ‘variation canzona’. The same tendency is to be found in the canzonas of Mayone, whose first book was published in the same year. But the great early master of the canzona was Frescobaldi.
Frescobaldi's first published canzonas were the five in his Ricercari et canzoni franzese (1615). They are multisectional, showing a tendency towards variation which is more pronounced in some than in others: the second is based on an ascending 4th that appears in each of its sections, and the third on a descending 4th. The capriccios (1624) are largely based on the same principle, although their treatment is more imaginative and sprightly. Six canzonas are included in the second book of toccatas (1627). All include toccata-like elements, and indeed the fifth opens with a purely chordal section and has very little fugal writing. The sixth is entirely in triple time. The variation principle is present, but in a highly subtle form in which direct thematic quotations are largely replaced by vague resemblances, and in which complete contrast is not excluded. Only in the third, where the descending chromatic four-note figure on which it is based can be recognized throughout the work, is the treatment strictly that of the variation canzona. A more subtle use of thematic transformation can be seen in the second canzona (ex.1). These splendid though little-known works represent a highpoint in the history of the form.
The canzonas in Fiori musicali (1635), intended for use after the epistle and after the communion, are similar to Frescobaldi's other canzonas. Their multisectional structure allows the performer to cut the piece short if liturgical considerations prevent a complete performance. Indeed, the second half of the final canzona in the first mass is labelled ‘alio modo, si placet’, for use either as a continuation of, or as an alternative to, the first half, although the whole piece was obviously conceived as a single entity. All embody the variation principle, the last being based on two subjects from the outset. The toccata element so noticeable in the canzonas of 1627 is less evident here. There is also a posthumous publication of Frescobaldi keyboard canzonas, the Canzoni alla francese of 1645; a few more survive in manuscript, and there is a single one (though with basso continuo) at the end of his Primo libro delle canzoni (1628), an ensemble collection that appeared twice, in parts and in score. The 1645 works have many of the characteristics of ensemble canzonas, such as special titles (La Rovetta, La Sabatina) and the sturdy independence of the parts. At the same time, there are too many features of Frescobaldi's highly individual keyboard style for them to be dismissed as mere transcriptions.
The keyboard canzona was cultivated in Italy after Frescobaldi by such composers as Tarquinio Merula, Giovanni Salvatore, G.B. Fasolo and Bernardo Storace; but with the decline in instrumental music of the strict contrapuntal textures of the Renaissance period the form became less common in Italy, and even Bernardo Pasquini wrote only a few examples. Many of its features were incorporated into the sonata and fugue. Among earlier German contributors to the genre were Christian Erbach and H.L. Hassler, both of whom are represented in the Turin collection. The form received a new lease of life at the hands of Froberger, who consistently used the variation principle in his canzonas. It is later found in the works of numerous German and Austrian composers, including Kerll, Georg Muffat and Buxtehude, and even in a single work ascribed to Bach. In many of these the variation technique is used, while in others the title seems no more than a synonym for a short fugue of sprightly character; and from this point the canzona's idioms passed into the general language of fugue.
The earliest canzonas independent of vocal models, apart from the dubious examples for solo lute mentioned earlier, appear to have been for instrumental ensemble. A solitary five-part piece entitled ‘La bella: canzone da sonare’ appears at the end of Nicolò Vicentino's Madrigali a cinque voci … libro quinto (1572). Ingegneri's second book of madrigals (1579) includes two pieces entitled ‘Aria di canzon francese per sonar’; possibly these are reworkings of vocal models. Maschera's Libro primo de canzoni da sonare, the earliest collection of original compositions in this form, probably appeared in 1582 (this edition is lost, and the work is known today from its reprint of 1584); it proved to be a most popular publication and appeared in several further reprints and transcriptions. It was followed by such volumes as Canzoni di diversi per sonar (published by Giacomo Vincenti, RISM 15883) and those of Bariolla (1594), Banchieri (Concerti ecclesiastici, 1595, and Canzoni alla francese, 1596), Borgo (1599), Mortaro (1600) and Rovigo (1600). Maschera's book contains 21 pieces, some bearing titles such as La Capriola or La Martinenga, probably deriving from patrons' names and indicating the social milieu for the performance of this type of canzona. From the canzona this usage passed to the sonata, while the similar practice of naming dances was retained at least until the keyboard publications of Couperin, Rameau and their French contemporaries.
Vincenti's Canzoni di diversi per sonar (158831) contains two works by Merulo, one apparently an arrangement of a chanson by Crecquillon. It includes anonymous arrangements of other chansons by Crecquillon, Willaert, Clemens and Gombert, while the title ‘Fantasia in modo di canzon francese’ (or ‘francesa’) is attached to two works by Gioseffo Guami. Two subsequent anthologies of importance are Raverii's Canzoni per sonare (160824) and the third part of Woltz's considerably larger Nova musices organicae tabulatura (161724); although the latter is a keyboard publication many of the pieces can be shown to be transcriptions of ensemble pieces, and most if not all of them very probably are. Several of Merulo's canzonas that are not arrangements of vocal works, though initially published for keyboard, reappear in these collections and must originally have been intended for ensemble. Among the composers represented in the two publications are Costanzo Antegnati, Banchieri, Giovanni Gabrieli, Guami, Frescobaldi, Luzzaschi, Pietro Lappi, Macque and Flaminio Tresti. Raverii's book was provided with a basso continuo, although all the works are for four, five and eight parts in Renaissance polyphonic style. The basso continuo was omitted in Agostino Soderini's four- and eight-part Canzoni (160820).
While the four- and five-part works in Raverii's collection, together with the four-part works included by Soderini, generally feature lively rhythms and fugal textures, a different approach is found in the antiphonal eight-part works. Antiphonal effects, which by then were somewhat conventional, had already been brilliantly exploited by Giovanni Gabrieli at S Marco, Venice, and some of the results of his experiments with the possibilities presented by the spaced galleries there are to be found in the canzonas, for eight to fifteen parts, published in his Sacrae symphoniae (1597). These 16 works (two of them variants of a single piece) include the famous Sonata pian e forte and a Sonata octavi toni a 12, which reveal the etymological link between the canzon da sonar and the sonata. There is too little evidence to justify any conclusions as to a precise distinction between the terms at this early date. The instruments specified in the volume are the cornett, violin and trombone, and these are used variously in five of the works; no doubt the remaining 11 works should be similarly performed. In three pieces (the seventh, eighth and ninth) the ten instruments are treated as a single ‘chorus’ (Gabrieli’s term); in ten works there are two groups, and in three (the 13th, 14th and 16th) there are three. Many begin with the typical canzona rhythm (ex.2a) or a variant (ex.2b). The latter is turned to brilliant effect in the triple-choir opening of the 14th. Gabrieli's Canzoni e sonate (1615), with a basso per l'organo, also uses the terms ‘sonata’ and ‘canzona’ without clear distinction, and extends the number of parts to as many as 22 in addition to the basso per l'organo.
The Sonata per tre violini in Gabrieli's 1615 collection points the way to a further crucial development in the history of the canzona, namely the adoption of a polarization between the upper and lower parts, in which the basso continuo and its accompanying harmonies are essential. This piece, though described on the title-page as ‘a 3 … voci’, is really in four instrumental parts, the basso continuo being thought of as outside the sphere of essential parts, although in this work it is as essential as the others. The same nomenclature is found in Frescobaldi's Primo libro delle canzoni (1628), where the basso continuo functions in three ways: as an independent bass throughout; as a partly independent bass; and as a basso seguente following the lowest sounding part at any time. The first type applies to the works for one or two ‘canti’, the second to works with one ‘basso’ part, and the third to works with two ‘bassi’ (in three or four parts) as well as to the three in the ordinary four-part texture. It should be noted that even a work for violin and harpsichord and two for harpsichord ‘alone’, added at the end of the reprint in score in the same year, require a basso continuo.
The ensemble canzona bears a close relation to two other early 17th-century forms apart from the sonata – the sinfonia and the concerto. The relationship appears in the title of Banchieri’s Ecclesiastiche sinfonie dette canzoni in aria francese (1607). These 20 works (14 sinfonias and 6 concertos) may be played on the organ alone (in which case they may be embellished); or else they may be played as ensemble pieces or sung to the texts provided for all but one (in which case the organist must provide a plain accompaniment from the basso seguente). The last of the concertos is an antiphonal work in eight parts.
Thus, in the early 17th century, canzona, sonata, sinfonia and concerto were largely interchangeable terms in the sphere of instrumental music. Usually they are provided with an organ bass, which may be merely a basso seguente following the lowest sounding part, or else a structurally essential basso continuo. The concerto and sinfonia subsequently developed for the most part from the full-textured canzona in which each pitch level is represented by a separate part, while the sonata derived from the type of canzona that emphasized a polarized texture between treble and bass.
The early Baroque sonata also retained the multisectional form of some canzonas and in many cases the use of fanciful titles. Some composers, such as Tarquinio Merula, preferred the old-fashioned name, although their works are essentially sonatas. Merula's first book (1615) consists of four-part works in Renaissance style, his second survives in an edition of 1639 and consists of works in trio-sonata texture including a separate basso continuo, and the third (1637), entitled Canzoni overo sonate concertate, marks a further advance, for the parts are specified as being for two violins, violone and basso continuo. The combinations used are two violins and basso continuo (both ‘a due’), and two violins, violone and basso continuo (‘a tre’). The fourth book (1651) is similar. Maurizio Cazzati was another important transitional figure. His first book (1642) is entitled Canzoni a 3: doi violini è violone, col suo basso continuo; his second (1648), called Il secondo libro delle sonate, adds a viola in some works, but the pieces are called ‘canzona’ except for two entitled ‘simphonia’ and one ‘sonata’. His subsequent books of sonatas (1656, 1665 and 1670) dropped the word ‘canzona’ altogether.
In Merula's earlier canzonas thematic links between the sections are often found, resulting in what amount to variation canzonas. There is no pre-imposed pattern on the number and ordering of the sections. By the time of Cazzati's later works the sections are almost always independent, and the fast movements are long. These are not fundamental differences between the canzona and sonata; they are rather organic changes in idiom which happened to take place simultaneously with the change of preference in name. By the second half of the 17th century the transition was virtually complete, and the sonata had ousted the canzona as the main form of instrumental chamber music.
In Germany the earliest published ensemble canzonas were those of Hassler (in his Sacri concentus, 1601). A singular contribution was made by Schein, whose two examples, of five and six parts respectively, in the Venus Kräntzlein (1609), together with the five-part canzona entitled ‘Corollarium’ at the end of Cymbalum Sionium (1615), offer a remarkable synthesis of canzona, dance and toccata-like elements. The 1615 work in particular combines examples of dazzling instrumental figuration with unexpected harmonies, buoyant rhythms and thematic metamorphosis. The four- and five-part works with basso continuo entitled ‘canzon’ in Scheidt's Paduana, galliarda, courante, alemande, intrada, canzonetto (1621) are on the whole more conventional and unduly prolonged, although the end of the five-part canzona on ‘O Nachbar Roland’ may be cited for its use of rapid repeated semiquavers (presumably for strings) three years before the celebrated instance, so elaborately justified on theoretical grounds, in Monteverdi's Combattimento. The form was not subsequently cultivated in Germany to any great extent, though there are examples in the published collections of Andreas Hammerschmidt (1639), J.E. Kindermann (1653) and others. Mention may also be made of the use of the word ‘canzona’ by William Young (1653), Purcell (1697) and A.L. Baldacini (1699) as a synonym for a lively fugal movement of a trio sonata.
ApelG
BrownI
MeyerMS
SartoriB
L. Ronga: Gerolamo Frescobaldi, organista vaticano, 1583–1643 (Turin, 1930)
G. Cesari: Preface to IMi, ii (1932), pp.ix–lxxxvi
J.M. Knapp: The ‘Canzone Francese’ and its Vocal Models (diss., Columbia U., 1941)
J.M. Ward: ‘The Use of Borrowed Material in 16th-Century Instrumental Music’, JAMS, v (1952), 88–98
S. Bonta: ‘The Uses of the Sonata da Chiesa’, JAMS, xxii (1969), 54–84
F. Sumner: Letter in JAMS, xxii (1969), 529–30
H.R. Chase: German, Italian and Dutch Fugal Precursors of the Fugues in the ‘Well-Tempered Clavier I’, 1600–1722 (diss., Indiana U., 1970)
D. Kämper: Studien zur instrumentalen Ensemblemusik des 16. Jahrhunderts in Italien, AnMc, no.10 (1970)
M. Picker: ‘A Josquin Parody by Marco Antonio Cavazzoni’, TVNM, xxii/3 (1972), 157–9
E. Selfridge-Field: Venetian Instrumental Music from Gabrieli to Vivaldi (Oxford, 1975, 3/1994)
E. Selfridge-Field: ‘Canzona and Sonata: some Differences in Social Identity’, IRASM, ix (1978), 111–32
D. Kämper: ‘Die Kanzone in der norddeutschen Orgelmusik des 17. Jahrhunderts’, Gattung und Werk in der Musikgeschichte Norddeutschlands und Skandinaviens: Kiel 1980, 62–78
A. Silbiger: Italian Manuscript Sources of Seventeenth-Century Keyboard Music (Ann Arbor, 1980)
A. Silbiger: ‘The Roman Frescobaldi Tradition, 1640–1670’, JAMS, xxxiii (1980), 42–87
F.G. Sumner: ‘The “Stylized Canzone”’, Essays on the Music of J.S. Bach and Other Divers Subjects: a Tribute to Gerhard Herz, ed. R.L. Weaver (Louisville, KY, 1981), 165–80
W. Apel: Die italienische Violinmusik im 17. Jahrhundert (Wiesbaden, 1983)
J.M. Harper: ‘Frescobaldi's Reworked Ensemble Canzonas’, Frescobaldi Studies: Madison, WI 1983, 269–83
N.M. Jensen: ‘Le revisione delle canzoni e il suo significato per la comprensione del linguaggio frescobaldiano’, Girolamo Frescobaldi: Ferrara 1983, 315–28
J.L. Ladewig: ‘The Origins of Frescobaldi's Variation Canzonas Reappraised’, Frescobaldi Studies: Madison, WI 1983, 235–68
G. Vecchi: ‘La canzone strumentale e la canzone-motetto a Milano nella prima metà del Seicento’, La musica sacra in Lombardia nella prima metà del Seicento: Como 1985, 81–97
D. Lo Cicerco: Preface to Florentino Maschera: libro primo de canzoni da sonare a quattro voci (Florence, 1988)
R. Judd: ‘Repeat Problems in Keyboard Settings of canzoni alla francese’, EMc, xvii (1989), 198–214
D. Sabaino: ‘Contributo ad una precisazione morfologica della canzone polifonica d'insieme: considerazioni analitiche sulle composizioni dei musicisti bresciani del Cinque-Seicento’, Liuteria e musica strumentale a Brescia tra Cinque e Seicento: Salò 1990, 191–235
A. Dell'Antonio: Syntax, Form, and Genre in Sonatas and Canzonas, 1621–1635 (diss., U. of California, Berkeley, 1991)
C. Longoni: ‘Gli organisti lombardi e la canzone all francese negli ultimi decenni del secolo XVI e nei primi del secolo XVII’, RIMS, xii (1991), 63–82
E. Selfridge-Field: ‘Genre and Instrumentation in Italian Music, 1600–1670’, EMc, xix (1991), 61–7
P. Allsop: The Italian Trio Sonata from its Origins until Corelli (Oxford, 1992)