Some Problems of Text, Attribution, and Performance in Early Italian Baroque Keyboard Music
Manuscripts of Italian keyboard music from the early seventeenth century frequently contain textual problems that make editing and interpretation difficult while raising questions about performance practice. This study closely examines the variations on the aria di Romanesca by Ercole Pasquini and a canzona attributed to Girolamo Frescobaldi, offering suggestions with regard to rhythm, articulation, and ornamentation, as well as more global matters. Also considered are partial concordances to works of Johann Jacob Froberger and possible relationships to vocal music of the seconda pratica.
3. The Source
11. The Source
1.1 In considering the keyboard music of the late Renaissance and early Baroque in Italy, most musicians have in mind a small group of publications that appeared during the late sixteenth and early seventeenth centuries. Yet most of the keyboard music played during this period was not printed; much of it was not written down at all, and the duties of most keyboard players included substantial amounts of improvisation--whether accompanying from a continuo part or short score or filling in parts of a church service. Hence our concentration on printed music, and especially the carefully prepared publications of Girolamo Frescobaldi, may give us a somewhat skewed view of what keyboard music was like during the period and of the manner in which it was played.
1.2 The keyboard sources of the period often take the form of manuscripts that, to modern eyes, give the impression of carelessly written scores preserving a baffling repertory of fragmentary and seemingly corrupt texts. Many lack clear or reliable attributions; most raise not only the usual questions about historical performing technique and interpretation but also basic questions about editorial procedure and the relationship of score to performance. It is the latter that I would like to emphasize in considering two works from the manuscript repertory of early Baroque Italian keyboard music: the variations on the aria di Romanesca by Ercole Pasquini and Canzona Ottava from a collection in the British Library (Add. Ms. 40080) attributed to Frescobaldi.
1.3 The leading scholar of this repertory has been Alexander Silbiger, who carried out the first systematic codicological study of the sources, identifying scribes and concordances, proposing provenances and dates, and making innumerable perceptive comments on the musical styles, attributions, and historical significance of their contents. Much of this work, initially presented in his dissertation, has been updated in his introductions to a series of facsimile editions through which the most important of these sources have been made readily available. (note 1) Yet these sources and, more important, the music within them remain neglected, in part because of the difficulty of interpreting them. What I offer here is meant to serve as an introduction to the use of these sources, with the aim of making them and their repertory better known.
2.1 The first of my two case studies is a set of variations by Ercole Pasquini, Frescobaldi's immediate predecessor as organist at the Cappella Giulia, the principal musical organization at St. Peter's in Rome. (Ercole is not to be confused with the much later keyboard composer Bernardo Pasquini, who was a performing partner of Corelli in the late seventeenth century.) Ercole, like Frescobaldi, came from Ferrara, where he might have influenced the younger composer; both must have been influenced by Frescobaldi's teacher, Luzzasco Luzzaschi, organist at Ferrara and author of the virtuoso madrigalspresumably composed for the concerto delle donne, the so-called three ladies of Ferrarathat were published in 1601.
3.1 Unlike Frescobaldi, Ercole Pasquini published no keyboard music. What little survives of his keyboard writing is preserved in a number of highly problematical manuscripts. Among these is a volume bearing the title Libro di fra Gioseffo da Ravenna, which evidently refers to its possession by an abbot of the monastery of Classe at Ravenna; this manuscript is the sole source of Ercole's variations on the aria di Romanesca. (note 2) The legibility of the manuscript is reduced by its apparent physical deterioration (see figure 1), but beyond this the text itself seems to be corrupt, requiring editorial emendation. Some emendations involve errors that are common in this repertory; thus the right-hand chord on the downbeat of bar 6.8 (note 3) is apparently written a third too high, producing parallel fifths with the bass and unexpected leaps in the upper voices (see example 1). Such readings would seem to be copying mistakes of a type endemic to this form of notation, with its numerous staff lines. Yet as one continues to come across similar readings elsewhere in the piece, one begins to wonder whether every apparent anomaly is really a mistake or is rather an intended stylistic idiosyncracy of the composer. More global questions are raised by other oddities, such as the notation of certain passages in double or half the arithmetically correct note values. Is each odd reading the product of some error that occurred during composition, copying, or transcription, or was it an intended reading in the composer's lost autograph, constituting a clue to the proper performance of the music?
3.2 Before attempting to settle such practical questions of editorial method or performance practice, before deciding how to play, we must first know what exactly we are playing. This music evidently belongs to a tradition that is ambivalent at best about the binding character of the notation on the performer. We might approach the surviving sources as working documents whose readings are to be revised, supplemented, perhaps even ignored when this is where our understanding of the music and its context leads us. In practical terms this means that the performer's interpretation of the music becomes difficult to distinguish from that of making an edition; to make sense of the sources one must be actively engaged in creating a coherent text for the music. Some of these manuscripts, despite their careless appearance to a modern reader, in fact give texts containing few apparent mistakes. (note 4) But, to judge from the numerous small errors and inconsistencies found in most manuscript sources of this repertory, it would seem that many copyists and perhaps some composers as well took a more casual attitude toward notation than classically trained musicians today are accustomed to doing. Copyistswho often were probably also usersof these manuscripts frequently must have assumed that the player would automatically fill in missing notes or correct mistakes. Such an attitude would have been in keeping with the view of keyboard music expressed in Banchieri's L'organo suonarino, which gives the impression that much, perhaps most, of the music actually played by professional keyboard players during the period was largely or wholly improvised. (note 5) Such players would not have been troubled by the notation of the manuscripts, which might have served them only as a rough guide to what was actually played. Although the carefully prepared published texts of Frescobaldi were frequently re-issued and copied into manuscripts, the unpublished pieces may not even have been thought of as constituting a repertory of compositions in the now-familiar sense, forming instead a set of examples for emulation, a source of material for further improvisation or composition.
4.1 Although Ercole's works have appeared in a modern edition, (note 6) it often amounts to little more than a diplomatic facsimile of the sources, reproducing errors from the original or normalizing aspects of the original notation that are potentially significant to performance or interpretation. Yet, if an edition of the present music is to be comprehensible to non-specialists it must, given the nature of the sources, incorporate considerably more editorial interpretation or intervention than is now customary. Ideally, an edition would present a complete facsimile of each source alongside diplomatic transcriptions of each as well as an edited text; the latter would be accompanied by a complete editorial commentary and would employ typographical means to distinguish original readings from editorial supplements. (note 7) Due to practical considerations, the notated musical examples presented below realize these ideals only partially; most editorial emendations have been carried out silently, with added accidentals in parentheses.
4.2 The two works that form the focus of the present discussion are unica. But in this repertory, concordances, where present, often present distinct versions; hence, many works extant in multiple sources would require separate transcriptions and editions of all versions (each incorporating further editorial emendations). Yet it remains an open question whether it is really necessary or desirable to extend such treatment to the entire repertorythat is, whether a complete critical edition in the usual sense would be appropriategiven the availability of facsimiles of the most important sources. (note 8)
5.1 Examples of Ercole's handwriting apparently do not survive. But his autographs may well have resembled the extant manuscript copies in important respects. Some support for this hypothesis comes from recent studies of manuscript sources for the music of Frescobaldi, who may well have heard and played this music. Frescobaldi's published keyboard works might have had a similar appearance prior to their publication, to judge from the notation of a number of unpublished works in what have been identified as Frescobaldi autographs. For example, the last two pages of a manuscript canzona (in Vatican, Biblioteca Apostolica, MS Chigi Q.IV.29, ff. 15-18') contain passages that have been rendered nearly indecipherable by notes that have been blotted out and written over. The alterations include the apparent insertion of entries of the fugal subject, changes so fundamental that they are more likely to represent compositional revisions than corrections of copying mistakes. (note 9) A copyist working from such an exemplar might well have produced some of the types of apparent misreading found in the present manuscript. Nevertheless, the owners of these manuscripts evidently found some use for them, despite their problems of legibility and textual accuracy. Such musicians, in turning to printed music, are likely to have continued to follow habits acquired during improvisation and performance from imprecise manuscript copies; we cannot assume that they aspired to note-perfect reproductions of printed scores. Hence anything about historical performance practice that we can glean from the Pasquini sources may also apply to those of Frescobaldi.
5.2 Pasquini's variations on the aria di Romanesca constitute his largest and most ambitious surviving work. It may not be coincidental that his successor's partite on the same ostinato are also Frescobaldi's largest set of variationsone, moreover, that underwent significant revision between its first two editions. (note 10) The two settings share an overall shape: a general acceleration of surface motion over the course of the first few settings that is interrupted by relatively sedate variations near the center, then reaches a climax and concludes with two surprisingly simple, understated settings. (note 11) There are also occasional similarities in figuration and in harmony. On the other hand, Frescobaldi, but not Pasquini, following a convention peculiar to certain Romanesca settings, notates most of his variations in duple time even though they are actually in triple meter. Pasquini, at the outset of his first parte, seems to refer to the galliard or corrente, whereas Frescobaldi avoids clear dance references until his quinta parte, which is a corrente (absent from the first edition). In general, Frescobaldi tends toward stricter, more genuinely polyphonic textures, whereas Pasquini often seems content merely to give the impression of counterpointa distinction also evident in the canzonas of the two composers.
6.1 Not surprisingly in a piece never prepared for publication, the Pasquini text contains a greater number of apparent errors. Yet, as one grows accustomed to Pasquini's music and begins to recognize some of the conventions apparently at work in it, one becomes less confident that certain anomalies of notated pitch or rhythm are errors, as opposed to purposeful violations of the rules of counterpoint for expressive purposes or intended notational irregularities meant to evoke certain freedoms in performance. The contrapuntal irregularities tend to be limited to occasional odd leaps within individual parts, forbidden parallels, and unusual passing tones. Prominent among these are the leaps to unexpected registers mentioned earlier, which sometimes lead to parallel fifths. This is especially common at the ends of sections, but it also occurs at internal cadences. (note 12) One can also find numerous instances of parallel octaves and fifths elsewhere (particularly within the opening parte), and there are a few examples of a type of sharply dissonant passing note now associated with Monteverdi. (note 13) In general, however, the irregularities arise through surface embellishment of the underlying ostinato progression, suggesting a casual approach to counterpoint and voice-leading such as might arise in an improvisatory tradition.
6.2 Such a tradition would have favored the cultivation of the more purposeful sort of rule-breaking that Monteverdi termed the seconda pratica. Indeed, the liberties found here are somewhat reminiscent of those arising in what Caccini described as "a certain noble freedom in singing" (una certa nobile sprezzatura di canto), evidently referring to the departures from strict voice-leading and dissonance treatment characteristic of early Baroque monody. (note 14) Since, however, these keyboard pieces lack verbal texts or descriptive titles, it is difficult to show conclusively that the relatively mild contrapuntal irregularities of the present work have any specific expressive significance. The ground-bass aria was, practically by definition, a vocal genre, and vocal settings of standard arie such as the Romanesca were flourishing at the same time as the instrumental ones. Hence it is reasonable to imagine that keyboard settings represented in a general way an imitation of vocal performance, in particular the monodic type described by Caccini. But concrete parallels to vocal music, as in the form of common melodic figures, are few, apart from the use of standard embellishments such as the groppo (long upper-note trill) and diatonic scale passages.
7.1 More important than any literal imitation of the voice would have been the freedoms of tempo and rhythm prescribed by Frescobaldi in the performance of his toccatas and partitas, which he describes as resembling those appropriate to "modern madrigals" (madrigali moderni). (note 15) As Frederick Hammond has shown, that freedom must have extended to the performance of variation sets, (note 16) and Frescobaldi's instructions for adopting a "broad" (largo) tempo in variations (partite) containing running figuration (passaggi) can be easily applied here. Moreover, metrical inconsistencies in Pasquini's music suggest that even within individual variations there is not necessarily a consistent pulse. For example, in bar 6.7 (example 1a) the note values in the two hands do not add up to the same duration; the upper voice contains the equivalent of nineteen eighth notes, the lower eighteen-and-a-half. Example 1b follows an ingenious emendation by Richard Shindle that yields both correct meter and counterpoint. Yet the notation of the manuscript might instead imply free rhythm in which the dotted notes are lengthened, the sixteenths rushed, and the concluding groppo (long trill) sustained for a duration left up to the performer, who might add additional repercussions to the ornament. (note 17)
7.2 The apparent shift of the pulse at the end of bar 6.7 from the half to the whole note recurs in several other variations as well. Frescobaldi doubled the value of ostinato notes in five of his Romanesca variations, but except in a few isolated cases the shift is carried through an entire partita, implying the maintenance of a steady pulse in performance. With Pasquini the shifts tend to involve only a few beats, and, as in example 1a, they are not always carried through all voices of a passage. The florid right-hand lines of bars 10.1 and 10.3, as notated in the source (example 2a), each comprise eight quarter-note beats in place of the normal six beats per measure (that is, per underlying ostinato note). In example 2b I have emended these passages to read as "correct" bars of 3/2. But one might just as plausibly have doubled the written values of the sustained notes of the accompaniment at these points. In either case, however, the implication is that the metrical pulse in performance should not so much change as dissolve or become arbitrary, somewhat as in an eighteenth-century cadenza. Evidently, Frescobaldi's rule about maintaining a consistent tempo within a variation must be discarded in this music, which was composed by someone apparently less concerned than he with maintaining unity of affect within each section (audio 1; for score see parte 10).
8.1 Another instance in which notational irregularities may carry suggestions for performance practice occurs with the beaming of small note values, in particular the use of separate flags for individual eighth and sixteenth notes, as opposed to continuous beaming. (note 18) Thus the eighths in bar 6.6 (example 1) are each separately flagged, whereas those three measures earlier in bar 6.3 are beamed in groups of six and four (see example 3). Although this gives the initial impression of being merely a notational inconsistency, as opposed to a palpable error, it is again necessary to consider whether the notation of the source is a random anomaly or a meaningful clue to performance. The beaming in bar 6.3 seems to reflect that passage's conjunct motion in parallel tenths, as opposed to the leaps of bar 6.6, whose canonic voices move with greater rhythmic and melodic independence. Presumably this translates into a more heavily articulated (staccato) performance of the passage in bar 6.6 (audio 2; for score see parte 6; for facsimile see figure 1). A comparable situation arises when a note or chord on the beat is flagged separately from a beamed group immediately following, as on the downbeat of bar 10.4 (example 4). In the latter passage, however, the separate flagging of the note on the beat is not maintained consistently for all three entries of the sixteenth-note figure. (note 19) Such beaming could reflect the composer's, or at least the copyist's, view of how the notes are grouped into motivic figures. Yet it does not necessarily relate directly to articulation in performance; it would be wrong to assume that motivic groupings were, or ought to be, always reinforced by articulation.
8.2 Some sort of articulation, however, may indeed be implied in the passages discussed above, as well as by certain dotted figuresin particular, the use of dotting to create a separation between a figure ending with a short note on the downbeat and the figure that begins with the following note. Sometimes the dot is accompanied by a break in beaming, as in bar 7.4 (example 5). (note 20) But the copyist has failed to notate the next two occurrences of the dotted rhythm in the same manner, and, moreover, seems to have introduced a mistake in the next bar. I have emended the latter to reflect the introduction of a new imitative figure after the first note of bar 7.5. (note 21) This use of dotted rhythms (or at least dotted notation) seems to reflect a tendency for figures to move from off-beat or "bad" notes and to on-beat or "good" notes. (note 22) It suggests, moreover, that many figures beginning on offbeats might have been slurred into the following beator at least that the most noticeable articulations in such cases would have followed rather than preceded the beat. (note 23) Many modern harpsichordists and organists have been taught to avoid slurs of this type, which, although commonly marked in classical and later music, are rarely notated in earlier repertories. Yet the usual reason given today, that such slurs obscure the beat, does not effectively cover the many instances where accompanying chords or long notes in the other voices clearly define the beat.(note 24) At least in the latter contexts, a type of articulation frequently associated with "modern" playing may in fact be "early" as well (audio 3; for score see parte 10).
9.1 Outside of the particular realms of rhythm and articulation just discussed, it is difficult to relate apparent errors or inconsistencies in notation to intended idiosyncracies of performance. But it remains essential to identify and, where necessary, emend errors in the text, even though this is made problematical by our inadequate understanding of the origins of these sources, the conventions governing their use, and the musical style of the works copied in them. Questions about the text may involve only the placement of a few accidentals or ties, yet even decisions about small textual details can drastically affect the character of the music in performance. Seemingly missing or misplaced accidentals are, in fact, a recurring problem in this repertory. Accidentals are often set above or below the staff, in such a way that it is often unclear precisely which notes they affect. Particularly in figures containing small note values beamed together, it was evidently customary to write accidentals only after the notes had been entereda practice that would often have resulted in the unintended omission of accidentals. Still, some apparent omissions may reflect not so much copyist error as the composer's failure to notate accidentals whose need seemed obvious. For example, in some variations the last two measuresthe reprisa, repeating the final cadence of the ostinatoare effectively in the major mode (with B-naturals indicated by sharp signs, the usual notational practice here). But in others the accidentals needed to specify this are missing, especially in the final chord. Since, however, the change of mode is clearly indicated in the first parte, it seems reasonable to err on the side of favoring rather than avoiding the modal shift in ambiguous cases.
9.2 Other passages pose harder questions. For instance, in bar 10.4 (example 4), maintaining the F# throughout the measure (in both staves) creates melodic diminished fourths and a disjunction with the following chord of Bb. Cancelling the sharp at some point within the measurefor example, on the second note of the upper voiceproduces an equally clashing chromaticism within the measure. Although the modern convention of applying an accidental throughout the measure did not exist at the time, it is clear from written-out trills (groppi) and other figures that the copyist did not repeat sharps every time they were necessary (see example 1). Hence it cannot be assumed that the unsigned Fs in bar 10.4 were all meant to sound as F-natural. But no solution is entirely unproblematical.
9.3 Knowing of the bizarria actively cultivated by certain contemporaries of Pasquini, such as the Neapolitans Giovanni de Macque and Ascanio Mayone, modern players and listeners may be prepared to accept almost any strange-sounding progression as exotic or expressive. But, despite occasional instances of sprezzatura in the use of dissonance or chromaticism, the harmonic idiom in the surviving works of Luzzaschi and even Frescobaldi is predominantly conservative, and in general this is true of Pasquini's keyboard music as well. (note 25) In the present piece, however, the florid melodic style and certain characteristics of the underlying ostinato create recurring opportunities for mild chromaticism or irregular dissonance treatment. This occurs especially over the B-flats in bars 1 and 5 of the ground bass, where the note E in the upper voices repeatedly forms a tritone. There is no evidence that these Es should be routinely altered to E-flats, and it is possible that some harsher notated dissonances are also intended for expressive effectfor example, the augmented sixth in bar 7.3. (example 5) (note 26)
9.4 Ties, like accidentals, seem generally to have been written after the notes were already in place and thus appear to have been omitted often. It is easy to decide to add ties in a case such as bar 10.6 (example 4), where restriking every repeated note as written creates an anomalously accented dissonant chord on the fourth quarter-note of the measure. But in variation 12 (example 6) practically every vertical sonority contains a repeated note; few ties appear. As notated, the variation might be understood as possessing a dance-like character. It is transformed by the addition of ties, which produce a fairly legato polyphonic setting; that the ties should be added is implied by their occasional inclusion even here (e.g., in bar 12.3). Moreover, there is a tendency throughout this repertory to represent sustained notes in a polyphonic texture as tied notes whose duration matches that of the moving notes in other voices; given this convention, the accidental omission of ties is to be expected. The second parte from the first edition of Frescobaldi's Romanesca variations presents a somewhat similar texture. There, however, the addition of a substantial number of ties seems implausible for musical reasons: the repeated notes are motivic, constituting a point of imitation that is even treated in augmentation (example 7). (note 27) There is no such motivic justification for most of the repeated tones in example 6.
10.1 With the missing ties of parte 12 we pass beyond a local textual problem to one involving basic musical character and interpretation. An equally basic question about the character of the music arises at the very end of Ercole's Romanesche, which, in the manuscript, concludes with a bare octave (example 8). One might be inclined to fill this out as a full triad (presumably major), yet the octave might have been intentional in light of the high tessitura and light three-part texture of the preceding passage. The progression leading up to this final sonority, moreover, lacks a full dominant chord. Hence one cannot assume that this was merely a formulaic closing passage; it may have been meant to be something distinctive and unusual.
10.2 The most serious textual problem in the piece involves a passage that is clearly corrupt, requiring extensive emendation. Parte 6 opens anomalously with G replacing B-flat as the first note of the ostinato (example 3). (note 28) The following two measures contain extra beats, and there is an unidiomatic partial doubling of the bass in bar 6.3. Despite the copyist's apparently casual approach to notational details elsewhere, such palpably faulty readings are unusual within this source. They are likely to stem from the copyist's failure to understand compositional changes that had been entered with less than perfect clarity into the autograph. (note 29) The music as notated is nonsensical, and one can only conjecture as to what the composer intended here or, more to the immediate point, what a player using the present manuscript copy would have done upon encountering such a passage. It would have been impossible to overlook the problem in the text, yet, as in similar situations elsewhere, there seems to have been no effort made to correct it. (note 30) If the manuscript was used for actual performancea point open to some doubt, in light not only of the textual problems in this passage but of numerous awkward page turnsthe player would have had to substitute a more or less plausible improvisation at this point.
11.1 Many of the uncertainties about text and performance that arise in Pasquini's music recur in an unedited collection (eleven canzonas and one toccata) identified on its original title-page as Fioretti del Frescobaldi. The copyist of the manuscript (British Library Additional Ms. 40080) has been identified as Frescobaldi's engraver Nicol˜ Borbone. The contents seem to be a compilation of passages that might have been found among Frescobaldi's unpublished papers or sketches, although there are no concordances to firmly attributed pieces by the master. (note 31) Prior to the identification of the copyist, the collection had been dismissed as patently spurious. (note 32) More recently it has been proposed that the style is that of Frescobaldi late in life, (note 33) and a single, fragmentary concordance perhaps confirms this. But stylistic differences between these and his better-attributed works make participation by one or more younger composers plausible as well.
12.1 Unlike the Pasquini work and most other keyboard music of the period, these are written in an exceptionally neat hand on five-line staves. But the notation sometimes consists of nothing more than a bass line, as in Canzona Ottava (example 9; figure 2). Such passages, which are confined largely to initial bars, are presumably to be realized in a manner resembling the opening of a toccata or the transition passages in a canzona. Silbiger points to the opening of a toccata in another British Library manuscript (Add. Ms. 36661), with a doubtful attribution to "Sig. Freses Baldi," as a possible model for a realization (example 10; audio 4). (note 34) Another possible model occurs in the opening of the last piece in the present manuscript (example 11; audio 5), whose designation as a toccata rather than a canzona may merely reflect the fact that the opening passage is fully realized. The underlying bass line of this opening passage (shown as an editorial addition in example 11) is similar to the unrealized basses in the initial sections of the preceding eleven canzonas.
12.2 The first polyphonic section of Canzona Ottava (example 9, mm. 9ff.) employs a thin contrapuntal texture that is more characteristic of pieces from the anonymous or weakly attributed manuscript repertory than of Frescobaldi's published canzonas. This, together with the five-line stavesunusual for Italian keyboard notation of the periodhas invited speculation that these pieces might not be original keyboard music but rather arrangements or reductions of music for instrumental ensemble. More than one passage in the manuscript resembles ensemble music in texture or style, as in a sequence whose implied voice crossings suggest writing for two violins (example 12). The first piece in the manuscript is headed by the unusual indication Cimbalo Solo, but the similar indication spinettina sola occurs in two of the ensemble works included in Frescobaldi's Primo libro delle canzoni (Rome, 1628). (note 35) But even if the present work is a reduced score or part for an ensemble composition, it contains no gross lacunae such as would have occurred if essential lines intended for another instrumental part had been omitted. Moreover, two partial concordances spotted by Silbiger seem to strengthen the assignment to solo keyboard.
13.1 One passage from Canzona Ottava (example 13a) recurs almost verbatim in the "Freses Baldi" toccata mentioned earlier (example 13b). This, as Silbiger noted, is actually a sectional canzona similar to those found in the Fioretti collection; as in the Toccata of that collection (example 11), only the opening is in toccata style. But the "Freses Baldi" toccata is predominantly in what we would call G major, whereas the common passage (examples 13a and b) is in D minor. This suggests that the latter passage has simply been grafted into a piece to which it is not, at least by modern standards, entirely appropriate. Surely, however, there is a close connection between this work and the Fioretti pieces, for another passage from the same toccata recurs twice, in two different canzonas from the Fioretti collection. (note 36) In addition, Silbiger notes a second concordance in Canzona ottava (example 14a) to a well-known work by Froberger (example 14b). Despite the obvious differences between the two passages, there is a striking similarity of mode or key, subject, length, and general character in the two passages illustrated. The differences are too great to warrant the conclusion that Froberger was the composer of the present piece or even that he knew it. But this could be an early form of a passage that Froberger borrowed and reworked in his toccata. At the very least one may conclude that sections, or at least subjects, taken from works by Frescobaldi or his students circulated independently. (note 37) These would have had the potential of becoming parts of new compositions or improvisations, including those of Frobergerwho was sent from Vienna in 1637 to study with Frescobaldi. Froberger's re-use of this material, if that is what it is, included the addition of a chromatic countersubject and the more consistent maintenance of a true four-part texture.
14.1 For performers, the chief question raised by the Fioretti pieces is whether to play them as is or to "realize" the notated texture. Concordances involving polyphonic sections, as in examples 13a and 13b, do not show substantial variants, suggesting that elaborate "realization" was not the usual practice in such passages. On the other hand, the dense reworking by Froberger of thematic material possibly taken from the Italian repertory suggests that a considerable amount of variation and harmonic enrichment is not out of the question. Hence, in the opening passage of Canzona Ottava, it is possible to imagine both a simple realization suggested by the opening of the toccata from the same manuscript (example 15; audio 6) and a more elaborate one in a somewhat later style derived from that of Froberger (example 16; audio 7). Still, in the toccatas and canzonas of the manuscript repertoryunlike works by Bach or even the strict four-part contrapuntal works of Frescobaldi and Frobergermost of the counterpoint is more implicit than real, the basis of harmony being not strict four-part voice-leading but a lighter two- or three-part texture. It is all too easy for us to impose the norms of Froberger's music and notationor Bach'son this earlier repertory, where additions might normally have been limited to simple, straightforward flourishes over unrealized bass notes and the addition of occasional ornaments or filler notes. (note 38) Moreover, the recurrence of sections from other works, as well as partial concordances between pieces within the Fioretti manuscript, suggests that musicians did not regard compositions of this type as integral wholes, instead considering themselves free to concatenate their own versions of favorite passages into new assemblages. This view accords with remarks of Frescobaldi in his prefaces to both the First Book of Toccatas and the Fiori musicali, which allude to the possibility of ending certain pieces prior to their final cadences. In both cases, the score represents not a finished or whole composition but only a source from which the player may choose to take material for a performance. From a modern point of view this seems to reflect a cavalier attitude toward the unity or integrity of the compositions within it. Under such circumstances, there can have been little sense of composition as an autonomous art, and attributions would necessarily have been lacking or ambiguous, particularly if a piece is composed of sections taken from different sources.
14.2 I would not advocate that we therefore compile our personal selections of favorite passages from Pasquini or Frescobaldi and call them canzonas or toccatas by those composers. Yet, this is, in a sense, exactly what we are doing when we play this music as written. Unlike players of the seventeenth century, who might have skipped sections or put them into new orders or remade them into their own versions, we tend to play exactly the notes that are in the score, in the order that the score tells us to. But this is as much a choice, a selection, as any other. Is it musically the most effective one? It can be so only if the composers really had the ability to compose the individual sections comprising their works into unified wholes that are greater than the sums of their parts. We take this for granted in later music, but only experimentation and analysis can demonstrate the point in this repertory. An effort has been made to do so for a few of Frescobaldi's toccatas, but as a general principle it remains a hypothesis. (note 39) Only with further imaginative performance and analysis of this long-unheard repertory will it become possible to hazard answers about this and other questions concerning its interpretation.
* David Schulenberg (email@example.com) is Assistant Professor of Music at the University of North Carolina at Chapel Hill. He is author of The Instrumental Music of C. P. E. Bach and The Keyboard Music of J. S. Bach, as well as a performer on harpsichord and other early keyboard instruments. He is currently editing Bach Perspectives, vol. 4, and preparing a textbook and anthology in Baroque music history for publication by Oxford University Press. He is also a composer; recent works include a harpsichord sonata and Rondeaux for soprano, cello, and harpsichord on texts by Machaut.
Editor, 11/23/02: David Schulenberg (firstname.lastname@example.org
is now Professor and Chair of the Music Department at Wagner College (Staten
Island, New York). Additional sound recordings of music relevant to this
article can be downloaded from his website: http://www.wagner.edu/faculty/users/dschulen/.
1. See Alexander Silbiger, Italian Manuscript Sources
of 17th Century Keyboard Music (Ann Arbor, 1980), and the series of
which he was general editor, 17th Century Keyboard Music: Sources Central
to the Keyboard Art of the Baroque (New York: Garland, 1987-9). References
to the latter are cited below by volume title and "Garland Series" number, as the volumes are catalogued and shelved separately in many North
2. Facsimile in Ravenna, Biblioteca comunale
Classense, MS Classense 545, introduced by Alexander Silbiger,
Garland Series, 12 (New York, 1987); mod. ed. in Ercole Pasquini: Collected
Keyboard Works, ed. W. Richard Shindle, Corpus of Early Keyboard Music
[CEKM], 12 (Neuhausen-Stuttgart: American Institute of Musicology, 1966).
With the exception of the British Library manuscripts named below, I have
examined the manuscripts only in these facsimile editions.
3. Bars in Ercole's Romanesche will be cited
in the form 0.0, where the first digit is the number of the parte or
variation, the second that of the bar within it.
4. This is true, for example, of the sources identified
as Frescobaldi autographs (see below).
5. Banchieri's L'organo suonarino (Venice, 1605,
revised and reprinted through the 1630s) describes the types of music
that organists were expected to provide during services, ranging from
simple cantus-firmus settings to elaborate sonatas and other polyphonic
works. Banchieri provides examples of most of these types of music, but
even those pieces that he gives in full are often notated incompletely,
in the form of a figured bass line or of soprano and bass lines which
the performer presumably filled in at his or her discretion.
6. CEKM vol. 12, op. cit.; sources of most of the works
are also reproduced in various volumes of the Garland Series.
7. The inclusion of facsimiles might seem to render
much of the conventional scholarly apparatus redundant, but the latter
is necessary if the reader is to avoid having to repeat the editor's work
of collating sources and texts. Diplomatic transcriptions make clear the
editor's readings of ambiguous or partially illegible originals.
8. I am grateful to Professor Silbiger for his illuminating
remarks on this subject in conversation; see also his review, "Recent
Editions of Early Keyboard Music," Journal of the American Musicological
Society [JAMS], 42 (1989): 172-88. Although French sources
of seventeenth-century keyboard music are often notated with greater care,
their more numerous concordances present variant readings that raise equally
challenging editorial problems; see David Fuller, "'Sous les doits
de Chambonniere,'" Early Music 21 (1993): 191-202.
9. Facsimile in Garland Series, 15/2 ; transcription
by Harry B. Lincoln in Seventeenth-Century Keyboard Music in the Chigi
Manuscripts in the Vatican Library (CEKM vol. 32) (Neuhausen-Stuttgart:
American Institute of Musicology, 1968), 1:25-7; discussion of text and
revisions and identification of Frescobaldi as the scribe (and composer)
in Claudio Annibaldi, "La didattica del solco tracciato: Il codice
Chigiano Q.IV.29 da Klavierbüchlein d'ignoti a prima fonte frescobaldiana
autografa," Rivista italiana de musicologia 20 (1985): 44-97
(especially 66). See also Annibaldi, "Musical Autographs of Frescobaldi
and His Entourage in Roman Sources," JAMS 43 (1990): 393-425;
further discussion in the introduction to Garland Series, 15/1.
10. Frescobaldi's Partite sopra l'aria della Romanesca
appear in his first book of Toccate e partite d'intavolatura (Rome,
1615); the second edition (Rome, 1615-16) removed five partite
and added seven to those retained from the first edition. See Girolamo
Frescobaldi, Il primo libro di Toccate d'intavolatura di cembalo e
organo, ed. Etienne Darbellay, Opere complete, 2 (Milan: Edizioni
Suvini Zerboni, 1977), 50-61 and 129-133; also Darbellay, Le Toccate
e i Capricci di Girolamo Frescobaldi, Opere complete, Supplemento
ai Voll. II-IV, 47-48.
11. This description holds especially for the revised
version of Frescobaldi's work.
12. At the ends of sections: conclusions of
parti 2 and 8; at internal cadences: bars 2.1, 5.8, 7.4, 7.10,
8.5, and 9.8).
13. For instance, the aria di romanesca "Ohimé,
dov'é'l mio ben" from Monteverdi's Seventh Book of Madrigals contains
some ferocious passing dissonances on the phrase "fera cagion" (bars 2.12-13); similar voice-leading occurs in Ercole's Romanesche
at bar 7.1, last beat (between the top three voices), and elsewhere, including
the passage shown in example 5.
14. Giulio Caccini, Le nuove musiche (Florence,
1602), preface; see the discussion of the term sprezzatura in note
10 to the translation by Wiley Hitchcock, in his edition of Le nuove
musiche (Madison, WI: A-R Editions, 1970), 44-45.
15. See the preface to his first book of toccatas
and partitas (second edition); translation and commentary in Christopher
Hogwood, "Frescobaldi on Performance," in Italian Music and
the Fitzwilliam (Cambridge: Fitzwilliam Museum, 1976), 14-22.
16. Frederick Hammond, Girolamo Frescobaldi (Cambridge,
Mass.: Harvard University Press, 1983), 225-6.
17. On the other hand, the apparent rhythmic error
in bar 6.5 seems more likely to be the result of copyist oversight than
of intentionally imprecise rhythmic notation. Shindle's reading is taken
from his edition, op. cit., p. 83.
18. See Etienne Darbellay, "Peut-on découvrir
des indications d'articulation dans la graphie des tablatures de clavier
de Claudio Merulo, Girolamo Frescobaldi et Michel-Angelo Rossi?," in International Musicological Society, Report of the Eleventh Congress,
Copenhagen, 1972 (Copenhagen: Wilhelm Hansen, 1974), 342-50.
19. Likewise in bar 3.5.
20. Also in bar 8.4.
21. Two subsequent statements of the same figure lack
both dotting and the break in the beaming: bars 8.5 (alto), 8.6.
22. The nomenclature of "good" and "bad" notes, used, e.g., by Girolamo Diruta, Il transilvano (Venice,
1593-1610), derives from the fact that in sixteenth-century embellishments
(divisions) "good" notes are consonant whereas "bad" notes tend to be passing dissonances.
23. Further examples in bars 4.5, 6.4-5, and 10.3.
24. As in Pasquini's Romanesche, bars 4.5,
25. Pasquini's vocal music is also conservative; I
am grateful to Richard Shindle for furnishing me copies of his unpublished
editions of Pasquini's motets and madrigals.
26. This augmented sixth might, however, be an artifact
of a copying mistake; the sharp could have been meant for an f' on the
downbeat of bar 7.4 (where, however, only a' appears in the upper staff).
Surely intended, however, is the cross relation C/C#'' in bar 4.3b.
27. For example, the same repeated-note motive is
presented simultaneously in m. 1 by the alto and by the soprano in augmentation.
28. I am grateful to Prof. Silbiger for pointing out
to me the significance of this apparent substitution.
29. Such corrections occur in the presumed Frescobaldi
autograph described earlier. Unfortunately, there exist no manuscript
copies of this source that might demonstrate how copyists actually dealt
with hard-to-read autographs.
30. In other passages the manuscript does contain
blotted-out noteheads and other signs of correction made during the initial
entry of the text; here the only such correction appears to be the alteration
of c' in bar 6.2 from an eighth to a quarter note.
31. Complete facsimile in London, British Library,
MS Add. 40080 ("Fioretti del Frescobaldi"), introduced by
Alexander Silbiger, Garland Series, 2 (New York & London, 1987); I
have been unable to consult the edition by Andrea Marcon and Armin Gaus
(vol. 1 in the series Musica antiqua nova arte scripta, Zimmern ob Rottweil:
Edition Gaus, 1994), reviewed by Frederick Hammond in this Journal,
3 , no. 2). Silbiger originally questioned the attribution to Frescobaldi
(Italian Manuscript Sources, 156-7), but the identification of
the copyist (by Silbiger; see his introduction to Garland Series, 2, and
his comment appended to Hammond's review) has strengthened the attribution.
32. E.g., by Hammond, Girolamo Frescobaldi,
292. The title, which forms a parallel to the composer's Fiori musicali,
appears to be in the copyist's hand; it could be rendered as "favorite
pieces" or "selections" by or from the composer.
33. Silbiger, introduction to Garland Series, 2: ix-xiii.
34. Garland Series, 2: xi; edited by W. R. Shindle
in Girolamo Frescobaldi, Keyboard Compositions Preserved in Manuscripts,
CEKM vol. 30/1, 75. These passages bear no resemblance to the partimenti
by such later composers as Bernardo Pasquini (see Garland Series,
8), which generally include figures and employ a style familiar from late-Baroque
continuo parts. On the other hand, similar things do occasionally occur
in earlier seventeenth-century ensemble music; see, for instance, ex.
2 in Danilo Constantini and Nicola Sansone, "Le toccate per violino
e spinetta di Francesco Spagnoli detto Rusca: Un esempio di tastiera concertante
nel seicento italiano," in Tradizione e stile: Atti del II Convegno
Internationale di Studi sul tema La musica sacra in area Lombardo-Padova
nella seconda metˆ del '600, Como-Villa Gallia 3-5 settembre 1987,
ed. Alberto Colzani et al. (Como: AMIS, 1989), 193-208.
35. I am grateful to Prof. Silbiger for this observation.
36. See Garland Series, 2: xix-xx for details. Silbiger
discusses further partial concordances involving the Fioretti in
Italian Manuscript Sources, 66-70.
37. The resemblance between the two passages is slightly
greater when one follows the original reading of Froberger's autograph
in bar 17, bass, third note from the end: d' (as in example
14b) in place of the e' supplied in most modern editions. Facsimile
of the autograph in Vienna, …sterreichische Nationalbibliothek, Musiksammlung,
Mus. Hs. 18706 (Froberger Autographs), introduced by Robert Hill,
Garland Series, 3/1, fol. 7r.
39. See Anthony Newcomb, "Guardare ed ascoltare
le toccate," in Girolamo Frescobaldi nel IV centenario della nascitˆ:
Atti del Convegno Internazionale di Studi (Ferrara, 9-14 settembre 1983),
ed. Sergio Durante and Dinko Fabris, Quaderni della Rivista Italiana di
Musicologia a cura della Societˆ Italiana di Musicologia, 10 (Florence:
Olschki, 1986), 281-300.
David Schulenberg, harpsichord
1. Ercole Pasquini, Romanesche, Ravenna, Biblioteca Comunale Classense, MS Classense 545, parte 10.
2. E. Pasquini, Romanesche, parte 6.
3. E. Pasquini, Romanesche, parte 7.
4. "Sig. Freses Baldi": Toccata (CEKM vol. 30/1, no. 20), from London, British Library, Add. Ms. 36661 (f. 28'), bars 1-9.
5. Frescobaldi (?): Toccata, from London, British Library, Add. Ms. 40080 (f. 35'), bars 1-6.
6. Frescobaldi (?): Canzona Ottava, from London, British Library, Add. Ms. 40080 (f. 22), bars 1-8, simple editorial realization.
7. Canzona Ottava, bars 1-8, elaborate editorial realization.
(set in Finale¨ by John Sheridan)
1. Ercole Pasquini, Romanesche, Ravenna, Biblioteca Comunale Classense, MS Classense 545, bars 6.6-10 (ff. 103'-104): (a) diplomatic transcription; (b) edition.
2. E. Pasquini, Romanesche, bars 10.1-3 (f. 105'): (a) diplomatic transcription; (b) edition. On the downbeat of bar 10.3, d' appears to have been blotted out in the lower staff.
3. E. Pasquini, Romanesche, bars 6.1-4 (f. 103'): (a) diplomatic transcription; (b) edition.
4. E. Pasquini, Romanesche, bars 10.4-6 (f. 106): (a) diplomatic transcription; (b) edition.
5. E. Pasquini, Romanesche, bars 7.3-5 (f. 104): (a) diplomatic transcription; (b) edition.
6. E. Pasquini, Romanesche, bars 12.1-3 (f. 107): (a) diplomatic transcription; (b) edition.
7. Frescobaldi, Partite sopra la Romanesca (Toccate e partite d'intavolatura di cimbalo, Rome, 1615), bars 2.1-3.
8. E. Pasquini, Romanesche, bars 13.9b-10 (f. 107'): (a) diplomatic transcription; (b) edition.
9. Frescobaldi (?), Canzona Ottava, from London, British Library, Add. Ms. 40080 (f. 22), bars 1-13a.
10. "Sig. Freses Baldi": Toccata (CEKM vol. 30/1, no. 20), from London, British Library, Add. Ms. 36661 (f. 28'), bars 1-9; analytical bass line added.
11. Frescobaldi (?): Toccata, from London, British Library, Add. Ms. 40080 (f. 35'), bars 1-6; analytical bass line added.
12. Canzona Ottava, bars 48b-50 (f. 24).
13. (a) Canzona Ottava, bars 54-56 (f. 24'); (b) "Sig. Freses Baldi": Toccata, bars 30-32.
14. (a) Canzona Ottava, bars 28-30 (f. 23); (b) Froberger, Toccata II, from Vienna, …sterreichische Nationalbibliothek, Musiksammlung, Mus. Hs. 18706, bars 17-19 (f. 7).
15. Canzona Ottava, bars 1-8, simple editorial realization.
16. Canzona Ottava, bars 1-8, elaborate editorial realization.
Figure 1. Ercole Pasquini, Romanesche, beginning of parte 6; Ravenna, Biblioteca Comunale Classense, MS Classense 545, fol. 103v, reprinted from facsimile edition (see note 2) by permission of Garland Publishing, Inc.
Figure 2. Frescobaldi (?), end of Canzona Settima and beginning of Canzona Ottava; London, British Library, Add. Ms. 40080, fol. 22, reprinted from facsimile edition (see note 31) by permission of Garland Publishing, Inc.
Copyright © 1998 by the Society for Seventeenth-Century Music. All rights reserved.
 Copyrights for individual items published in The Journal of Seventeenth-Century Music (JSCM) are held by their authors. 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 author(s), and advance notification of the editors of JSCM.
 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:
 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, who will act in accordance with the decisions of the Society for Seventeenth-Century Music.
 Citations to articles from JSCM should include the URL as found at the beginning of the article and the paragraph number; for example:
Jonathan Glixon, "Far il buon concerto: Music at the Venetian Scuole Piccole in the Seventeenth Century," Journal of Seventeenth-Century Music 1 (1995) <http://www.sscm-jscm.org/v1/no1/glixon.html>, par. 2.3.
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.