Since Alessandro Stradella and Bernardo Pasquini, the keyboard toccata had been a once-through, more or less motoric single movement. Bipartite movements, with a double bar near the middle and repeats, derived from dances. These could be grouped together in suites. But around 1700 composers of keyboard sonatas, searching for ways to imitate the Corelli/Vivaldi ensemble sonata, began struggling to find its true form. Bipartite movements with increased formal and technical ambition, sometimes in toccata style, sometimes as expanded dances — sometimes called sonatas, sometimes toccatas — began to appear. Toccatas grew more complex. Terminology remained vague; often the same piece would be called sonata in one source, toccata in another.
Then composers like Galuppi, Pescetti, Alberti and Durante began creating multi-movement keyboard sonatas, raising the form to the level of ensemble works. They created structures with varying numbers of movements balanced in different ways. Older composers of single-movement sonatas/toccatas found themselves out of fashion, and went back to their archives to regroup their works.
At least this is what I think happened to two prominent masters: Benedetto Marcello and Domenico Scarlatti, both students of Francesco Gasparini in Venice. Marcello’s four- and five-movement sonatas all give the impression of having been assembled from earlier single movements, surely inspired by his association with the virtuoso Scarlatti.
Scarlatti‘s keyboard works all come down to us as single movements, numbered separately and called either sonata or toccata in their sources. But as Ralph Kirkpatrick first emphasized, in the 13 semi-parallel volumes of the Venice and Parma manuscripts they are mostly grouped in pairs by key. As this tremendous archiving project began, I think Scarlatti went to his shelves and moved forward chronologically from pieces which had not already been preserved in the Essercizi, printed in 1738, and in the 1742 and 1749 royal collections. He organized stacks of folios by key and looked for likely pairs. As he went along he revised, and composed some new works to make pairs with older sonatas, as well as some new pairs. The few works which remained unmarried in the great sorting show up as wallflowers in the ballroom groups of 30.
After 1754 the work of copying slowed as more and more newly-composed sonatas were created. The final volumes were mostly new works, actually composed in pairs — but still numbered as separate, single movements. That fact should never be lost sight of; nor should the fact that different pairs are offered in different sources.
The pairs as they appear in Venice and Parma exist in a shadow state which continues to cause controversy. Are they obligatory, or mere suggestions? The preceding has been a prelude to an heretical view: many of the pairs are in fact incomplete, intended as fixed companions to ephemera.
The most satisfying of Scarlatti’s pairs are slow/fast, a form also found in the previously-named composers’ works. Around 1732, Francesco Durante, an influential Neapolitan composer/teacher whom his near-contemporary Scarlatti certainly knew (he studied with Papa Alessandro), published six excellent pairs of slow (studio) / fast (divertimento) sonatas. The print may have nudged Scarlatti to action: his Essercizi followed six years later.
The least satisfying of the Venice/Parma pairs are fast/fast. There often seems no reason for such pairings, which offer little variety of mood. And it seems obvious on stylistic grounds that it some cases the movements come from different periods and were simply paired ad hoc.
A type of pair quite frequently encountered consists of an allegro alla breve and a sonata in triple time, either fast or tempo di menuetto. The contrast can be reasonable satisfying, especially when the first movement is a substantial fugato. But more often than not something seems to be missing – a slow middle movement and/or a prelude.
The Neapolitan overture, the Venetian concerto, the sinfonia in general (right up until Mozart’s youth works) almost always follow this pattern: fast duple / slow / fast triple or menuet. The slow movements are often simple chords or easy tunes; so it is with Scarlatti’s few preserved orchestral sinfonias. It would have been the work of a moment for Mimo, when performing his elaborate and difficult fast movements for the court or friends, to improvise a slow middle movement and/or a prelude, as an aesthetic/technical preparation, or a breather between ofttimes savage or folksy fireworks. An interesting parallel is offered by Bach’s Brandenburg Concerto no. 3, with its two Phrygian chords in lieu of an improvised violin or harpsichord slow movement.*
The prevalence of this structure evidently led to an anomaly in Parma XIII (1757), the last volume of the set. K516-17, one of the finest (incomplete?) pairs of all, appears there in reverse order, with a note saying that the first is to played second. In the correct order (as it appears in Venice), the sonatas are 3/8 Allegretto and Alla breve Prestissimo. The copyist at first sight apparently couldn’t believe this topsy-turvey set of time signatures, and must have been corrected by the composer before preparing Venice.
This is where Visente Rodríguez Monllor (1690-1760) comes in. Organist of Valencia Cathedral, he was successor to Cabanilles, and inherited the latter’s florid style. But his 1744 manuscript entitled “Libro de Tocatas para Cimbalo, Repartidas por Todos los Puntos de un Diapason” goes entirely new ways. Its 30 sonatas (a Scarlattian number) are arranged in ascending chromatic order of all major and minor keys, beginning with D minor like the old ricercar cycles, and omitting only E-flat minor, F-sharp major, G-sharp minor, B-flat minor and C-sharp minor. (D-flat major is present in last place.) Despite their prolixity and banality, which has led to their virtual oblivion as far as the Scarlatti literature is concerned, they show all the elements of Scarlatti’s style: the Italianisms, the virtuosity, the modulations, the fugal style, the prevailing binary structure, the hand crossings, the trills, and the more recent Spanish folk influences as well. Some sonatas closely resemble the Essercizi in their single-movement binary form and figuration. Others are clearly inspired by later works, and show different shapes and various numbers of movements. Hence, the 1744 20-key compilation reflects a later arrangement of Rodríguez’s evolution.
The date offers a stylistic milepost for Scarlatti. The gap between the time around 1735, when the Essercizi were being prepared for publication and Andalusia had been left behind, and 1752, when the great archival project was launched, is the most difficult for establishing a chronology of Scarlatti’s sonatas. 1744 falls right in the middle; comparing Rodríguez’s most advanced works with Scarlatti’s in the earlier Venice/Parma sets is illuminating. The pre-classical operatic influences which Scarlatti encountered when the court moved to Madrid in 1733, and which leaven his later sonatas, are almost entirely lacking in both groups. The operas serving as inspiration pre-1744 are those of Alessandro Scarlatti’s Naples, the next generation of Neapolitans, Gasparini’s Venice, and Domenico’s own (largely lost) Roman productions.
It can only be speculated how the relationship between the works of these two composers came into being, but – to finally arrive at the main point proposed here – I think Rodríguez heard Scarlatti live, and that some of the Valencian’s sonatas offer a clue to missing elements in the adoptive Madrileño’s. Four of the multi-movement sonatas have simple, quasi-improvisatory, slow middle movements, such as seem lacking in so many of Scarlatti’s pairs. One of these, as well as a fifth sonata, also have slow introductory movements which can be conceived of as preludes. A similar “prelude” is found in a toccata by another Neapolitan, Nicola Fago (1676-1745). A solid, duple allegro is followed by a quick triple time. Both are preceded by a shorter grave introduction.
But by far the best examples of how Scarlatti may have improvised preludes are the recercatas of Sebastián Ramón de Albero Añanos (1722-56). These works, all the more astounding for their being so little-known, are the closest thing to the French prélude non mesuré to be found outside of France. Another close associate of Scarlatti’s, Padre Antonio Soler (1729-83), also has eight remarkable Preludios in his Llave de la Modulación (1762). They are full of directions for quick tempo changes and contrasting free sections (arbitri). Incidentally, the six pairs in the best source of Soler’s sonatas (the Madrid Conservatory MS) are all slow(ish)/fast, with the second movements in triple time. They are what I would call true pairs – satisfactory in themselves, and dating from after 1752 when he arrived at El Escorial from Catalonia. The title page proclaims Soler discepolo di Domenico Scarlati [sic]. On the other hand, the 14 pairs in the court copyist’s manuscript of Albero’s 30 Sonatas (the other two are fugues ending each half) are of the shotgun-wedding type, which is not surprising since Albero had arrived in Scarlatti’s orbit no later that 1746 – eight years earlier than Soler and presumably that long before the great archival work began.**
One final hypothesis: late slow/fast/fast “triplets”, such as K478-80 and K481-3 work well as four-movement sonatas (the up-and-coming form in the 1750’s) with a grand, composed first movement and an improvised slow third.
***
I feel sure that many more examples of preludes and slow middle movements could be found in better libraries than mine, by more diligent researchers than this writer, who only scribbles for his own pleasure and to his own mental satisfaction. Thinking, as I do, that Scarlatti’s unsatisfying fast/fast pairs – and especially those in sinfonia manqué form – were elaborated in performance by improvised preludes and/or slow movements, I can accept their existence in the sources as enhancements of single Sonatas, instead of puzzling annoyances.
*To this contemporary example might be added one from the 20th century. Bach’s fragmentary harpsichord concerto BWV 1059 was extended beyond the autograph first page by Gustav Leonhardt for a 1964 recording. He added virtuoso passagework to the obbligato organ part from the Sinfonia to cantata BWV 35, upon which the original was based, and quite plausibly used the introduction to the second part of the cantata, another Sinfonia with obbligato organ, as a finale. His middle movement was a melancholic “improvisation” which thrilled me when I first heard it in the library of the Juilliard School in 1969. I took out the LP this evening and listened to what was boldly declared a Rekonstruktion. The Leonhardt Consort is not quite what it became in later years, but my teacher’s playing shows his style at the precise moment when he broke the code of how to handle the harpsichord – a solution which, from everything I hear nowadays, has drowned in a morass of tempo chaos. It is also noteworthy that the two BWV 35 sinfonias follow the pattern [Allegro] 4/4 / Presto 3/8, which might indicate their previous existence as a concerto for a different solo instrument – with or without a lost second movement.
**Parma I, the starting point, is undated, but usually assumed to be, like Venice I, from 1752. Given its character as an anthology of teaching pieces, it could be considerably earlier.
January 30, 2025
|