Editorial principles: The Scarlatti Cantatas
Principles used in Rosalind Halton’s Scarlatti Editions
My aim in editing Scarlatti’s work has been to work from the basis of his own notational practice, by studying as many as possible of his extant autograph manuscripts, both cantatas and operas. Some of these manuscripts (Olimpia, Non sò qual più m’ingombra, Peno, Perdono) form the principal source of editions I have made. They contain a wealth of detail in bass figuring, (often neglected by copyists), and in some cases dynamic markings, tempo markings, and articulation marking. Recitatives are profusely punctuated in the composer’s hand.
Idiosyncrasies of Scarlatti’s notation.
Time-signatures: the metre 3/8 is always given with the prefix C3/8, and almost invariably barred by Scarlatti in units of 2, though he never seems to use the time-signature 6/8.
Key-signatures are generally given, as was the custom at the end of the 17th century-early 18th century, with one sharp or flat less than the modern key-signature, or in some cases, even 2 flats fewer (as in the F minor aria 1 of Peno, 1705). In widely modulating recitatives (again Peno provides an example), Scarlatti gives no key-signature, though the piece is clearly based in a key, in this case, C minor.
Use of accidentals reflects the chromatic style in which Scarlatti composed, repeating an accidental in long bars (e.g. in common time or 12/8) especially if it occurs in both halves of the bar, but not in consecutive notes. Scarlatti often gives an accidental not considered necessary in modern practice, for what we may call rhetorical effect – e.g. a semitone alteration from the previous bar, that alerts the musician to the significance of the change.
Underlay of the text is exceptionally clear in Scarlatti’s manuscripts by his detailed use of beaming. I believe it is important to preserve the beaming of the vocal part in a modern edition.
In instrumental parts, we find that even copyists close to Scarlatti, like Piero Castrucci or the Lanciani family, often change the composer’s beaming. It may be that in some cases this can indicate articulations and when an autograph source is extant, it is worthwhile to preserve the composer’s beaming of instrumental parts.
Of course, most works of this repertoire do not exist in an autograph, and some cantatas survive in 5 or more sources, which may contain notable differences. Here the editor requires a working knowledge of the copyists who formed part of Scarlatti’s circle in Rome, as they had direct access to his manuscripts. Future research may uncover a similar body of knowledge about Neapolitan copyists which will prove equally helpful as the studies of scholars such Keiichiro Watanabe (Tokyo), Hans Joachim Marx (Hamburg) and Ursula Kirkendale in the field of identifying Roman copyists.
Understanding the composer’s use of notation is the key to adopting an appropriate set of editorial principles. How ‘consistent’ can a modern edition be, given that manuscript is itself a medium not lending itself to ‘consistency’? I aim to establish the most likely readings of the text on the basis of comprehensive manuscript study. Clefs for the vocal lines are changed from soprano and alto to the modern treble clef.
Key-signatures and time-signatures are retained, and in most cases the barring (e.g. in the case of the triple time signatures noted above). The original form of dynamic and tempo markings is given, and punctuation of the text when it comes from an autograph.
Bass figuring is given in most cases as it appears in the manuscript(s). I do not supply a written out accompaniment. (My edition of Clori mia, Clori bella, 1699, for Saraband Music gives a sample written accompaniment). In the case of non-autograph cantatas, the bass figuring may be sparse; in these cases we have supplied figures in square brackets as a guide.
Dr. Steven Campbell (University of New England, 1994 – 98) and Mr. Nathan Scott (Newcastle, 1999 – ) have given invaluable advice and many hours of work in the layout and design of these editions.
Newcastle, Australia, 2000.