The complete published works for keyboard of Michelangelo Rossi, the material for this recording, have survived in four copies: two of them, published by G.B. Caifabri and Carlo Ricarii respectively, are preserved at the Civico Museo Bibliografico Musicale of Bologna; the other two lack publication details and are preserved respectively at the Biblioteca del Conservatorio of Naples and the Musikarchiv of the Abbey of Göttweig in Austria. While the Caifabri print carries a title-page in movable type with the coat of arms of the Cross of Genoa and bears the publication details “IN ROMA a spese di Gio: Battista Caifabri in Parione all’insegna della Croce di Genova”, those of the other editions are richly decorated copper engravings.

Ricarii shows the Aldrobrandini arms surmounted by a crown adorned with stars (part of the family emblem) held up by two winged cherubs seated on trabeation sustained by two columns. A third cherub, seated centrally on a step at the base of this elaborate theatrical backdrop to the curtain carrying the title, holds a cartouche with the words “ROMA, MDCXXXXXVII”. At the cherub’s feet the following inscription appears along the base of the step: “(SI) VENDONO IN PARIONE ALLA CROCE DI GENOVA”.

The Naples and Göttweig prints are identical: two female musicians seated on a bench on either side of a shell-shaped construction surmounted by the Barberini arms (bees) held up by six winged cherubs and surmounted by the Cross of Malta and a cardinal’s hat. One of the ladies plays a rectangular spinet that rests on the bench and a wooden trestle; the other listens while holding a violin and bow: the latter is clearly a reference to Rossi himself, who as well as being a composer of keyboard music was also known as “Michel Angelo del Violino”. At the bottom centre a cherub seems engaged in a dance step while holding in his left hand a trumpet with the letter W engraved on the bell. The date and place of printing are added by hand: in the Naples copy it is “1657” and “Roma”; in that of Göttweig it is “Roma 1640”.

The titles given in the three editions differ slightly. Naples and Göttweig: TOCCATE E’ CORENTE D’INTAVOLATURA D’ORGANO E CEMBALO DI MICHELANGELO ROSSI

Ricarii: TOCCATE E CORENTI D’ INTAVOLATURA D ORGANO E CIMBALO DI MICHELANGELO ROSSI DI NOVO RISTAMPATO DA CARLO RICARII

Caifabri: TOCCATE E CORRENTE PER ORGANO, Ò CEMBALO DI MICHEL ANGELO ROSSI,

The fundamental problem, therefore, is that of establishing the chronology of the editions, for on that issue all considerations on the sources of Rossi’s aesthetic and stylistic inspiration must depend. Fortunately, a valuable study of the Toccate and Correnti has been published by Alexander Silbiger, whose article (Alexander Silbiger: “Michelangelo Rossi and his Toccate e Correnti”, reprinted for JAMS, XXXVI, No. 1, 1983) exhaustively deals with the problems of how to date the four surviving copies of the volume.

Silbiger’s reconstruction justly puts the Naples/Göttweig edition first and he conjectures the date as 1634, in consideration of the fact that Rossi’s service with the Barberini family was abruptly interrupted that year. The absence of the publisher’s name and dedication could indeed be explained precisely by Rossi’s fall from favour and Cardinal Antonio Barberini’s withdrawal of protection and consequent lack of interest in a dedication.

The Ricarii edition comes chronologically second. Silbiger notes that the phrase “Di novo ristampato da Carlo Ricarii” is evidently added later. The ink is different and only just slotted into a rather cramped space: in fact the publisher’s name is so compressed that a comma has been inserted to separate name from surname. Moreover additions have clearly been made to the last figures of the date (as from the fourth X), which awkwardly run onto the reverse side of the cherub’s cartouche. The original figures of the date are therefore “MDCXXX”. In her preface to the Spes facsimile edition of the Toccate e Correnti Laura Alvini refers to research made with “suitable apparatus” with the aim of establishing “whether, as would appear to the naked eye, after the third X a second engraver had erased one or two numbers so that the date could be revised to 1657. Since no trace of underlying numbers emerged”, the result was “that one could take into consideration the hypothesis that the date formerly engraved was M.D.XXX.” If, on the other hand, we consider the owner of the coat of arms, Princess Olimpia Aldobrandini, a possible date for this second edition (of which the Ricarii copy is a reprint) is 1637/38, for in July 1638 Olimpia married Paolo Borghese and a later date would plausibly have required the addition of her husband’s arms.

Finally, Silbiger places the Caifabri edition last. Active between 1663 and 1695, Caifabri was a publisher who habitually adopted earlier editions to which he would affix his own device: this would also explain the phrase “a spese di”. Furthermore Silbiger acutely observes that at least the first plate of the Fourth Toccata has been changed, very likely because it had been seriously worn from overuse. Judging from appearances, even the Ricarii edition has various seriously damaged plates (for example, pages 11 and 25).

As a personal contribution, I would just like to add that the Caifabri title-page could conceivably be a reutilisation of an edition that precedes all the others (though I admit the likelihood is slim). There are two reasons. First, it uses movable type and is hence oldfashioned, compared to the engraved title-pages (though dispensing with the services of an artist-engraver could have been merely a decision to economize). Second, it is strange that it has no noble coat of arms, as was customary, but only the sign of the shop, the Cross of Genoa. Again one could conjecture that Caifabri’s practical nature induced him to use and readapt the cross of the Savoia family – a hypothesis that could lead us to Rossi’s first patron, Cardinal Maurizio di Savoia, and perhaps to a date very close to 1630. These, of course, are mere suppositions and Silbiger’s intelligent reconstruction still remains by far the most convincing. As for the title itself, that used by Caifabri is certainly the most summary and imprecise. It loses the specification “D’INTAVOLATURA D’ORGANO E CIMBALO” (which indicated not so much a specified instrumentation as the form of notation, that of keyboard tablature) and instead acquires the much more simplistic and misleading indication “PER ORGANO O CEMBALO”.

As regards Rossi’s style, Silbiger acutely observes that his uncle, Lelio Rossi, was organist in the cathedral church of S. Lorenzo in Genoa, where the maestro di cappella was Simone Molinaro, who in 1613 had published the Partitura delli sei libri de’ madrigali a cinque voci dell’Illustrissimo, ed Eccellentiss.(imo) Prencipe di Venosa, D. Carlo Gesualdo. Very likely Michelangelo studied with his uncle, a Servite friar, and collaborated with him as organist. And during his period in the service of Cardinal Maurizio di Savoia he would have had ample opportunities for meeting the chief musician of the Cardinal’s chapel, Sigismondo d’India. In Rossi, who also happened to be a distinguished madrigalist and even the composer of two operas, the chromatic influences of the above composers, Gesualdo and D’India, came together and combined with a dutiful observance of Frescobaldi’s keyboard style. In the keyboard works the result was a very personal idiom, of the type invented by the Ferrarese master: one based on “the manner of playing with cantabile affects and with a diversity of passages” so that “this way of playing must not be subject to the beat; as we see used in the modern madrigals, which, however difficult they may be, are made easier by means of the beat, by making it now languid, now swift, and even sustaining it in the air, according to their affects or meaning of the words”.

In the Rossi toccata the free sections, placed above all at the beginning and end, form a framework for the intermediate expressive sections, which are full of affects and resort to strongly accented lombardic rhythms, alternating with more extended fugal elaborations that can be traced back to the canzona. This generates a mixed toccata-canzona genre of remarkable appeal and originality, in which the various affects and effects, though inspired by the new Frescobaldi genre, are often taken to their most extreme consequences, above all in the chromatic passages, in a manner comparable with the vocal paroxysms of D’India and Gesualdo. The Frescobaldi influence is revealed above all in features such as leaps, passi doppi (passages for both hands) and trills, of which Rossi became an exemplary disseminator.

Below I have drawn attention to certain performing details that are worth noticing. To identify the bars, reference is made to the Spes facsimile editions of both Rossi and Frescobaldi:

Toccata I: Beginning, adagio (Frescobaldi’s Avvertimento no. 3). The trill is indicated to be played from the upper auxiliary. At the end of the first page the music at times shows a ‘compositional notation’, at times a ‘performing notation’ (see the notes to be held in the imitative passage in the second bar of the last line).

Toccata II: Beginning, “adagio, et arpeggiando”. The trill is painstakingly notated from the upper note and closes with a fermata (Frescobaldi’s Avvertimento no. 4). At the end of the first line, the trill is written with an initial turn (equivalent of the doppelt-cadence in the table of ornaments Bach wrote for Wilhelm Friedemann’s Clavierbüchlein). In this regard, it is interesting to note that the French harpsichordists (with the exception of Roberday) drew more inspiration from the free genre of the Italian toccata than from the contrapuntal forms of the capriccio/fantasia/ricercar, yet while they applied its essential qualities in the free ‘semibreves’ of the Prélude non mésuré, they also drew up discrete and all-embracing Tables des agréments of Cartesian clarity that freed them from the fully writtenout notation of the Italians. At the end of page 5: an episode with “lombardic rhythms” (inverted dotting) typical of Frescobaldi’s Elevation Toccatas. Beginning of page 6: a repeated-note trill using Caccini’s notation, which Frescobaldi was also to employ as from the instrument Canzonas of 1628 and in the Fiori Musicali to which Gregorio Strozzi, Frescobaldi’s last true imitator, resorted in abundance. Page 6, line 2, penultimate bar: the minim chord that closes the expressive passage includes an A natural (to comply with the standard form of unequal temperament of three sharps and two flats), but must here be normalised to include an A sharp with a suitable adjustment to the tuning. This is a very frequent occurrence in the Correnti, as we shall see below.

Toccata III: Page 7, line 3, last bar: an interesting case of a trill clearly from the upper note. This shows that the widely acknowledged ‘rule’ (in the performance practice of early music) that the Italian seventeenthcentury trill always started from the main note is simply not true. Elsewhere I have demonstrated this with examples from Frescobaldi. The documents must always be read avec discrétion! On page 9 the last section resorts to lombardic accents.

Toccata IV: The final section has a virtuoso figuration in the left hand evidently inspired by Toccata IX of Frescobaldi’s Second Book. The whole passage, however, is developed in a personal way.

Toccata V: The calm opening is similar to Toccata V of Frescobaldi’s First Book. Page 13, last line, bar 1: an elaborate trill figuration. Even the episode that begins at page 15, line 3, is inspired by Toccata V of the First Book, page 15, penultimate line. It is worth observing that the notation of the unisons is not indicated by means of rests in one hand (as, on the other hand, we find so diligently applied in Frescobaldi).

Toccata VI: The opening trill has an evident fermata on the last note, as in Frescobaldi’s Avvertimento no. 4. Page 18, end of line 3 and following line: an evident trill from the upper note. The trill of page 19, last bar, is very like a doppelt-cadence und mordant (to use the Bachian terminology). In the final episode the thirds of the right hand show a very interesting application of tied shared notes: it was thought to be a performing technique codified by Marcel Dupré, whereas it is frequent long before that, precisely to avoid re-striking unisons. C.P.E. Bach refers to it in his Versuch über die wahre Art das Clavier zu spielen (facsimile Breitkopf, Leipzig, 1976), Appendix, page 11: his example almost exactly repeats the passage in the third movement of his father’s Fifth Brandenburg Concerto.

Toccata VII: Unquestionably Rossi’s best-known Toccata. The opening is inspired by Toccata VII of Frescobaldi’s Second Book, from which he also got the idea for the final chromatic passage. The opening trill is very clearly from the upper note. Line 2, bar 2: observe the ‘performing notation’ in the right hand. On page 21, bar 2, it is worth noting the modification to the last note of the left hand: it has been corrected to a quaver, thereby erasing a final (hypothetical) semiquaver D, perhaps in order to avoid a unison with the right hand. At the end of the same page, of great interest for performance practice is the distinction between the individually notated quavers and those joined by beams to form groups of four. The final chromatic episode is a passage of rare expressive power and it singles Rossi out as a composer of the highest stature: worth noting is the tremendo lupus in the meantone tuning between E flat and G sharp on page 23, penultimate line, last bar.

Toccata VIII: At the beginning: E flat against F sharp. The first minim chord at the end of the opening episode is certainly major. End of line 3: a trill with a deceleration and a fermata on the last note. Page 25, line, 3, bar 3: there is a case of ‘performing notation’ at the third beat in the right hand: in spite of appearances it is a run of four semiquavers (B A B C, with a held B). Also worth noting is the galloping rhythm (quaver plus two semiquavers) of evident Frescobaldian influence running from the end of page 25 through the whole of page 26. The episode in the first line of page 27 is typically a discurso or glosa accompanied by the bass, of which there are many examples in Frescobaldi (e.g. Toccata VII of the Second Book, end of page 21).

Toccata IX: In the same mode as Toccata IX of Frescobaldi’s First Book. Line, 1, bar 3: the first minim chord closing the first episode is certainly major. Page 29, line 3, penultimate bar: the keyboard with a short octave limits the low note to C. On a full keyboard it is preferable to play a low A. Same page, last line, bar 2: a harsh meantone clash between F sharp and B flat in the second bar.

Toccata X: A piece of particular breadth. Page 31, line 3, beginning: interesting ‘performing notation’ for the last series of descending semiquavers with held notes. At the end of the same line: trill with a fermata. The following series of impetuous and rhythmically clashing imitative passages at the end of page 31 comes to a close at page 32, line 1, penultimate bar, with an instance of ‘performing notation’ and a printing error in the duration of the note values of the last bar. Worth observing is the painstakingly notated trill with a fermata. After a fugato, there is an expressively declaimed episode with lombardic accents at the culmination of ascending scales. The following fugato episode is divided into two sections: the second of these, at page 33, line 2, bar 3, has stretto entries for the voices and concludes with a grand descending cadenza at the beginning of the last line, inspired by the similar descending solo passage at the end of the first page (pag. 30) of Toccata X of Frescobaldi’s Second Book. The toccata ends with rapid, phantasmagoric figuration.

Correnti: This genre of music belongs to the galanterie alla moderna that Frescobaldi had already used brilliantly at the close of his First Book: the second edition with the definitive Avvertimenti to the Reader engraved by Christoforus Blancus in 1616. These galanterie were then extended in 1637 (after Rossi’s edition, therefore) with his Aggiunta, consisting of Balletti, Ciaccone and Passacaglie. Rossi’s Correnti are absolute masterpieces of their type.

Worth noting is the occasionally summary notation at the cadences, which unquestionably requires completion, and (in particular) the avoidance of notes when not contemplated in the standard meantone tuning of three sharps and two flats. There is a clear example already at the end of the Corrente I in which the D sharp of the dominant chord is first left without a sharp and then omitted before the close on E (which, being a final chord, should be major, according to Trabaci). The same occurs in Corrente II, page 36, line 1, penultimate bar, where the A (otherwise sharp) is omitted. At the end of the first part, note the trill from the upper auxiliary with a long appoggiatura: it is none other than a 4/6 trill resolving onto a 3/5 chord according to a formula that was to be standard in the following century, a fact rarely grasped by modern performers. Again this close is summarily notated, like so many others where the harmonies need completion (as in continuo realisation). One could infer from the chords avoided that the works were written for an instrument of fixed meantone temperament like the organ, but I feel that this solution can be ruled out given the essentially secular nature of the Corrente form. Besides, we have already established that the phrase “d’Intavolatura d’organo e cimbalo” referred to the tablature and not to the instrument intended. In Caifabri’s edition this specification is omitted merely for reasons of concision and commercial interest. Moreover, the title-pages of the Napoli/Göttweig copies clearly show a woman playing the harpsichord: a detail interestingly anticipated over a century earlier in Andrea Antico da Montona’s Frottole intabulate da sonare organi of 1517, the title-page of which even shows the composer himself playing a harpsichord of the Italian type (with his rival Petrucci portrayed as a peeved Barbary ape listening to him) – again in spite of the title, which clearly indicates the use of tablature and not the instrumentation.

An interesting detail is the way the right and left hands distribute the middle voice at the start of the second part of Corrente III: a similar attention to performing matters is displayed by Frescobaldi at the start of his own Corrente I of the First Book, where the C sharp and, in the second line, the G sharp are again divided between the hands. Again in Corrente V the harmony of the penultimate bar of the first part needs to be completed with an E. In the fourth-last bar of Corrente VI a D natural is clearly indicated where a dominant chord on B cadences on E. In my opinion this is a clear case of pruderie on the part of the engraver, whose musical knowledge (as was always the case) was derived from experience, which meant that he was accustomed to avoiding notes that were inappropriate for the keyboard. Such practices also explain the frequent attempts to ‘sort out’ the different voices: even in Frescobaldi’s toccatas it often happens that when an extended passage is divided between various registers, rests are inserted after the voice has moved into the next register. This is the reason why Frescobaldi prefers to use tablature for his toccatas and is forced to see it constrained within the procrustian bed of score in the Fiori Musicali: telling instances of such discomfort are found in the Toccata avanti la Messa della Madonna, bar 3, where the musical line is divided between the tenor and bass lines even though it forms a single unit, and in the Toccata avanti il Ricercare of the same Mass, bar 10, where the ascending scale is divided between the bass, tenor and soprano. But to return to the ‘difficult’ notes in meantone tuning, the scores of Rossi’s operas are actually rife with such notes, but the difference is that these scores were written for continuo players and other instruments and not for solo performance.

In Corrente VII the engraver, who was naturally expected to write from right to left, inadvertently reversed the notation of the bass. The mistake is evident if compared to a similar passage in the third bar of page 41. In the latter case, however, the notes are D C B A B and not, as before, D C B G B (a difference, incidentally, that I believe should be retained). A similar error has deceived many editors and performers of Frescobaldi’s Monica of the First Book (quinta parte, page 53, bar 1), where the sharp against the F should in fact be read as a cancellation of the preceding E flat (not notated, but implied because it follows a B flat). For the same reason Frescobaldi had put a sharp (i.e. a natural) against an E in his Corrente III of the First Book (page 68, bar 3): the E followed a B flat and would have be flatted as a routine measure.

Corrente X shows the same need for harmonic completion at the end of the first part. In certain quaver passages there is some doubt over the tritones, which are always avoided when it is a matter of flatting the E, but not when it is a question of flatting the A (for fear of ‘meantone clashes’). The same thing happens at the end of line 1, where the tritone is avoided in the soprano but not in the tenor reply. The problem is very evident at the end of the penultimate line where the alto E should receive an additional flat, as in the bass two bars later. This time, however, I lacked the courage to add a flat to the A as well. As for flatting the E in the final cadence, my solution is to add a flat in the descending passage and omit it (it is not marked, in any case) on the last E. For the toccatas I used a full-length (steso) Italian harpsichord with a single keyboard and two unison stops. It is a copy of a late sixteenth-century instrument.

For the Correnti I also used a wing spinet. Rossi’s harpsichord certainly had a short octave, as is proved in at least two places: in Toccata VIII, page 27, line 2, the minim D minor chord is divided between the two hands in such a way that the left hand can negotiate a low D, its octave and an F above that; in Toccata IX, page 29, line 3, the lowest note would more plausibly be a low A, but a C is written instead because the key was lacking. In the latter case I prefer to play an A, for obvious reasons of completeness. In similar ways the extensions of Domenico Scarlatti adapt to the contingencies of the instruments on which the sonatas were composed: where possible, it is reasonable to complete the ranges. The wing spinet was certainly known in Rome in Rossi’s day: its inventor, Girolamo Zenti da Viterbo, was one of the Barberini suppliers. The tuning of the instruments is strictly meantone, with the black notes made into sharps or flats according to necessity.

Sergio Vartolo
English Translation: Hugh Ward-Perkins