What I think is what I am.

This is a backwards, mirror view of the sonatas of Domenico Scarlatti in relation to what you are likely to read in the literature on this great composer. Even these two names are backward in the title of this website: Scarlatti Domenico.

Scarlatti’s music speaks so directly that it needs no explaining. But I shall attempt to do so anyway, since I have developed many ideas about the music which I have not seen written down anywhere else, or if they have been, it was in a more contingent and skeptical way. My mirror view of Scarlatti is a credo as well as a cogito.

The chief Scarlatti mysteries which I will investigate are:

• for whom did he compose the nearly six hundred keyboard sonatas which have come down to us?

• when did he compose them?

• did he compile them into the major source collections himself?

• in what order were these collections compiled?

I start with a working premise that the main sources available to us for the sonatas, including their arrangement and details within these sources, originated in some direct way from the composer himself, unless there is external evidence to prove otherwise. I don’t state this as a fact, but rather as a starting point. Since it is possible that the overall arrangement and any particular detail originated from the composer, the arrangement and details must be considered for their possible significance.

Rather than assign historical identities to various persons who might be involved in the creation of the sonata sources, I could use more neutral terms such as composer, collector, student, scribe, but I prefer to see the composer and the compiler, at least, of the sources as one person, Domenico Scarlatti. This is the most simple and direct way of dealing with the mass of material to be examined (the "least action" principle). And this is what needs to be investigated first. If I were to worry in every case about whether Maria Barbara or Farinelli were the compiler, for example, I would complicate matters unnecessarily. If it turns out that the material cannot have been both composed and organized by the same person, I will have at least gained that knowledge.

The resulting investigations are contained in the files listed in the navigation links at the top of this page, particularly Catalogue, Collections, Card games, and Chronology. Some ideas which might benefit from further investigation are in Capriccio

Some of my conclusions follow, arranged in the order also used for the Catalogue notes; the introduction to the Catalogue pdf file expands each of these areas:

• source collections

• dates

• companions

• instruments

• difficulty

• form

• accidentals

• ornaments

• graphics

• influences

• metaphors

source collections:

The fifteen manuscript libri (volumes) in Parma (Biblioteca Palatina \ Conservatorio Arrigo Boito) are the most satisfactorily arranged.

The Parma version of a sonata is not always the first nor the last, but the collection as a whole remains the most complete and authoritative, so I use it, with reference to earlier or later variant sources for which I have information.

I have listed the presence of the sonata in other collections and editions for which I have contents listings, but have concentrated on those which seem to have been produced during Scarlatti’s lifetime. There are many later sources I don’t list which are important for tracing Scarlatti’s influence on other composers, but I am more interested in trying to find out which collections Scarlatti may have compiled himself.

The maps and notes in the Catalogue describe facsimiles of original (18th century) sources whenever possible. Modern editions have failed so far to provide all the necessary information; none of them even provides a satisfactory transcription of just one primary source for each sonata, but attempts to pick those details from different sources which the editor approves. Fadini’s edition made a good attempt to document which details come from which sources, but it remains incomplete and more sources have been discovered in the meantime. The Kirkpatrick edition (Johnson reprint corporation, 1972) reproduces photographically all the sonatas in the Parma libri and Essercizi, and at least one primary source for the remaining sonatas accepted by Kirkpatrick as genuine. Other facsimiles have been published for single sources such as Venezia and Lisboa but are not easy to find outside large music libraries or rare book dealers (see Citations file for the works mentioned)

See accidentals, ornaments, and graphics below for some of the other areas in which modern editions of Scarlatti are usually unreliable. The only solution is for the libraries which own the sources to provide good facsimiles freely available to all over the internet. (See Citations for the internet version of the Venezia facsimile) Perhaps some day this will be realized. In the meantime serious performers are under an obligation to seek out at least one of the original sources mentioned above to use as a text. All the ambitious recording projects recently attempted or presently underway are worthless otherwise.

With about 600 pieces of uncertain date, there needs to be a method of identifying the sonatas. The Kirkpatrick numbers have outlived their usefulness and should be replaced by words, letters and numbers identifying a primary source: I use E1-30 for Essercizi 1 through 30, P1:1 through P15:42 for all the sonatas in Parma; V42:3 etc for the sonatas in Venezia 1742 not found in Parma or Essercizi and so forth for the sources which have a few sonatas not found elsewhere. Adding the key signature also helps to identify the work and make the identification stick in the memory. Someone has to take the step of abandoning K numbers--I have volunteered. These primary sources are given in the right-hand columns of Sheveloff’s worklist for New Grove 1980, revised by Boyd for Grove music online 2000+. I use the same primary sources as they do with a few exceptions: Venezia 42:50, 47, 44a, 44c instead of Coimbra 1,2,3,4 ; Münster 5:22 instead of Paris 4; Parma 6:27-28 instead of Venezia 2:1-2. Sheveloff’s worklist also contains the Pestelli numbers; Boyd’s replaces these with Fadini’s numbers. Kirkpatrick and Fadini numbers can be searched in the Catalogue pdf file; I don’t use them as an ordering device, but they are included in the editions used line of the notes.

The Venezia order has a claim to authority also, and corresponds to the order chosen by both Fadini and Kirkpatrick in their editions, but the V libro and sonata number should be used, not Kirkpatrick’s or Fadini’s numbering systems, both of which are incomplete anyway. The chief problem with all extraneous numbering systems is that they do not contain useful information and so are not worth remembering; the more time passes, the more out of whack and useless they become, but they cling to the works like indelible tattoos. The composer’s original groupings and opus numbers are often ignored. Newly discovered works are left in limbo. The chronological implications of the numbers become rigid and difficult to dislodge.

The libro numbers of Scarlatti’s major collections work exactly like opus numbers: they let us know, in most cases, associations of works which originated from the composer, like anthologies of poems or short stories put together by writers at different times for different audiences. The association of several pieces in Parma (and Venezia) with calendar events is another incentive for using the Parma numbering system.

I have attempted to establish, in the Card games pdf file, the order in which the various collections and volumes within the larger collections were compiled. The results are given in a table near the end of the Collections pdf file.

dates:

There are at least two dates to be considered in connection with every Scarlatti keyboard sonata: the first is the date of composition, which is not known, but perhaps can be surmised; the second is the date of a collection in which the sonata first appeared (from among the collections and editions which have come down to us and can be dated). The first dates of composition in all likelihood go back at least as far as 1715, when the composer was 30, but primarily a composer of vocal music; the earliest date which can be associated with some certainty to a collection is 1737, when the composer was 52, and the bookseller Charles-Nicolas LeClerc obtained a privilege to publish Scarlatti harpsichord pieces in Paris.

In the Catalogue pdf file I assign about dates at five-year intervals from 1715 to 1755 for when I think the sonatas may have been composed. I also list the 18th century source collections known at present for each sonata (more are still being discovered apparently) and attempt to arrange those in order of compilation.

The question of whether there is any such thing as a valid linear chronology which can be expressed by a numerical system is discussed in the introduction to the Chronology file, as well as my attempt to establish a non-linear chronology based on certain stylistic features which I describe there in greater detail.

I think my chronology is a start in the right direction, or a further step after Pestelli’s start. The Characteristics pdf file lists some ways to establish similarities between sonatas; at the end is a comparison of my chronology with Pestelli’s. If some overall trends can be accepted, it should be possible to look at various categories, especially those which involve all of the sonatas (keys, themes, rhythms, keyboard technique) and decide an order of development within each category. Perhaps the puzzle will all fit together; it is worth attempting even if it doesn’t.

My chief conclusions from attempting to date the sonatas are:

Most of the sonatas were composed much earlier and within a more restricted time period than is usually assumed.

Almost all of the characteristics of Scarlatti’s mature style were in place by about 1737 when the Essercizi were selected for publication; certainly the first fifteen sonatas of the 1742 Venezia libro confirm this. If the Bologna manuscripts are accepted as prior to 1727, then this maturity was reached even earlier.

For whatever reason, the composer released relatively few sonatas to the general public but kept them for the use of his royal patrons and for himself. When he compiled them into libri for specific collectors toward the end of his life, he revised some of them to accommodate the instruments owned by those collectors, and made a few other, usually not extensive, sorts of revisions.

The association of sonatas in the same key was a habit begun very early in his composing career: these may have been suites of up to ten sonatas in the earlier years, later smaller numbers. Pairs were used both early and late and had a symbolic meaning: the royal pair who were his patrons after 1730. These associations by key were flexible; they only seem fixed in the last few years of compiling because the most complete collections were apparently compiled for those same royal patrons. There is often a stylistic resemblance between sonatas in the same key taken as a whole (seen from a distance) but a much less pronounced resemblance in the components of the pairs (looked at close up) as if chosen to provide maximum contrast rather than maximum similarity. Dare I say it? the composer made artistic decisions.

Since first preparing the Chart of sonatas (pdf file), I have played through several sequences of the sonatas in the order given there, backwards as well as forwards, mixing keys as well as tracking single keys.  I feel the overall order holds up well, whether or not the rough dates I have assigned are correct; several sonatas have been moved more recently as I find more patterns that seem convincing. However I have avoided too much reliance on single factors until there is more news on investigations into the dating of sources, hinted at in various Scarlatti conference announcements in 2007, but still not published.

The following patterns have emerged:

Before 1725: key structures are generally simple; modal keys appear often; when more distant keys are explored there is often little sense of going somewhere in a larger scheme (Alessandro Scarlatti’s toccate seem somewhat aimless in this sense also). Sonatas are often in either a primarily contrapuntal or a basso continuo style, but the typical Scarlatti imitative opening which is not repeated in the second part appears rarely. Sonatas are generally short. There are sonatas for different instruments which were included in the cembalo collections in 1742 and later. Full-chord and partial-chord endings are present in some sonatas, but their absence doesn’t necessarily make a sonata later; they could have been edited out in the late sources in favor of Scarlatti’s obvious preference for ending on a single note or octave in one or both hands.

1725-1735: newer french and german influences show up alongside specific contemporary regional italian ones (venetian and neapolitan); iberian influences are there also, but usually in separate sonatas (as if intended to be dance movements of suites) or isolated sections of sonatas. Andalusia is close enough to Portugal to make it futile to say that Scarlatti had to travel from one to the other before he could have heard the music of either. The key structures are usually fairly simple and direct in the first parts, more roundabout in the second parts but with a clear sense of development and direction. Sonatas written as fugues, rondos and minuets continue from the earlier period but are mostly abandoned by 1730 (there is a group of probably later sonatas which combine rondo and binary forms) Sonatas which explore distant keys in sections of extended ostinati or pastorales start to show up about 1730. There is a clear progression of difficulty in sonatas which seem to have been written primarily as technical studies, from easy ones close to 1725 (perhaps for Maria Barbara’s earliest training or for her uncle Antonio) to very difficult ones closer to 1735 (leading up to the publication of Essercizi) Exceptions could be taken as being for Scarlatti’s own use. The sonatas identifiable as being primarily studies tend to include a lot of rhythmic formulas, such as the 12345 bass endings I include in the Characteristics pdf file. There is a general absence of iberian features in these study-type sonatas, as if Scarlatti were really obeying the dictates of João 5, or as if he composed most of them when he was back in Italia, from about 1727-1730, and shipped them to the portuguese court. These studies also often have simple key structures in their first parts: tonic and dominant in various combinations of major only, minor only, or major and minor mixed, as if Scarlatti were also providing Maria Barbara lessons in composition, as he indicates in his preface to Essercizi (he says he participated in her education toward "Maestria del Canto, del Suono e della Composizione")

After 1735: the iberian influences are more prevalent and not isolated; both parts tend to travel through a large number of keys and have striking thematic contrasts; sudden change to distant keys are a common feature of both parts but handled with complete assurance. Technical difficulties are still there, but integrated within sonatas of great internal variety, such as the ones which are often given theatrical or orchestral metaphors.

Prozhoguin 2010 p112-113, 138 reveals that Scarlatti was granted salaries from both the portuguese and spanish courts until the end of his life; a pension from the portuguese court began in 1742 and from the spanish court by 1746 but perhaps sooner.  It is likely therefore that providing sonatas for his patrons was no longer a duty after the mid-1740's.

There are problems in positioning a few sonatas which have a simple first part and a wide-ranging but masterly second part; these may be early sonatas with late insertions or revisions in the second parts. I don’t think that Scarlatti composed new works in very simple key structures (such as first part tonic and dominant only) at the end of his life. One dateable late work by Scarlatti, the Salve regina in A major of 1756, confirms the move toward masterly handling of distant keys and theatrical contrasts.

None of this is provable or particularly scientific; I am being influenced by my own biases and assumptions. I’m just saying it makes a certain amount of sense and is a helpful way of organizing so much material.

companions: (pairs, groups of three and four sonatas, longer suites in the same key which appear together in source collections, sometimes in different combinations)

Many of the 18th century Scarlatti sources contain at least some sonatas grouped together by having the same key signature, major or minor. The majority are given separate numbers, but a few are under the same number (for example, Venezia 1749:2 and Münster 5:1-7 each contain at least two sonatas under the same number; other collections do not number any of the sonatas but use Suites in the overall title (Roseingrave-Cooke begins with two suites of five sonatas each in G and D major and minor; the related Boivin-LeClerc editions consist almost entirely of same-key suites).

A problem has arisen for some scholars because the sonatas which are sometimes associated with each other in the various collections are different (Boyd p160-166; Sutcliffe p367-375). I have made use of these differences to help determine the order in which the various collections might have been compiled. As I said earlier, I follow what is called the least-action principle: the line from one point to another is a straight one, unless there is an outside influence: the composer is the compiler of a collection of his own works unless it can be shown that someone else has done the compiling (the composer has died or editors acknowledge their own roles, for example). The composer as compiler selects and orders sonatas already composed. Through time he might change his mind about the contents of a collection and its order on aesthetic grounds or to meet the needs of the different collectors for whom the sonatas are being compiled. The evidence of the changes can be used as a tool to determine the likely order in which they were made.

I do not propose any rules for determining whether the groupings are intentional or not, other than using the tonic key for adjoining sonatas. I assume intention everywhere that accident is not known to have intervened (examples of accidents in compiling: notes to Parma 6:13 and 14 in D major and Parma 15:3 and 4 in d minor state that the order of the companions has been reversed). I have even included a few groupings in different but related keys, such as Parma 1:22-24 in G, C & G, and Parma 2:18-30 in a variety of keys with three or more flats. I don’t assume that a move from longer suites to simpler pairs can be used by itself as a tool to determine chronological order: pairs such as those within V42:42 in d, 45 in G and 48 in A have  been assumed by various scholars to be among the earliest sonatas.

In investigating the chronology of the sonatas, I purposely kept the companions separated. Nevertheless, I think many of them were in fact composed to complement one another, whether at the same time or across a span of years.  This can be seen in the Characteristics pdf file, where I have arranged the companions in the major sources by their about dates.  Of the 589 sonatas which I have mapped in the Catalogue (not counting variants of the same sonata) only 32 do not have same-key companions.

The association of sonatas by key is a fact; it can be ignored by the performer or scholar, but it won’t go away. Why not try to see what can be learned from it?

The two sets of Bach’s Well-Tempered Clavier were compiled from separate pieces composed over a wide span of years, especially the second set. Some of the preludes or fugues in the second set have more in common with pieces in the first set than with their own partner piece; an example is the Prelude in E major from the second set, which opens with the theme of the Fugue in E major from the first set. Players could mix and join pieces in the same key or same tonic from Bach’s two sets; or play just the preludes or just the fugues, etc. But most people would agree that Bach’s own order is the most satisfying. In Scarlatti’s case, the performer is able to choose from a variety of sources those different combinations which seem to work best together. Like Bach’s association of preludes with fugues, Scarlatti’s companion pieces shed light on one another, such as influencing a choice of tempo for each. For us as performers or listeners, there is the added satisfaction of finding larger structures into which these smaller works have been fitted by the person who created them.

instruments:

There are a few sonatas which require exceptional keyboard ranges, so editors have declared the manuscripts to be in error (see P10:23 E: high g#, P11:7 e: low E, P13:17 G: high a, in the Catalogue) There are also only four sonatas requiring a low F: this has been accepted by editors. Boyd p288 fn23 mentions the possibility of special tunings for the highest or lowest notes in P14:2 C, which requires both a low F and a high G.

Most of the primary sources state on their title pages that the sonatas are per cembalo; there is internal evidence that a few of these were originally composed for organ or for a solo instrument such as a violin with basso continuo accompaniment, and less certain indications that some? many? may have been composed for pianoforte. At the time of his death he owned a "clabicordio" which may have been acquired after his second marriage in 1741 (Morales p7-8; Prozhoguin 2010, p105, 119-27, 138-43). There is internal evidence that the ranges required to play some of the sonatas were altered at the time of compilation. Attempts to establish an order of composition based on the ranges required to play the sonatas have been counter-productive.

An example: the range of low G to high g doesn’t appear in the Parma & Venezia libri before 1754 (both collections are thought to have been compiled for the use of Maria Barbara). But the range A to e also shows up for the first time in those collections that year, and some of the sonatas in the same libri require only C to d (and are individually dated 1754 in their Münster versions). On the other hand, the Bologna 1 manuscript, which could be dated before 1727, since it is annotated as being for the use of Francesco Gasparini, who died early in that year, contains at least one sonata, a variant of P12:16 g, which requires the G to g range. There is also the appearance of 61-key (G to g) ranges in sources such as Lisboa (1751?) to refute the contention that all the sonatas were composed for Maria Barbara’s exclusive use and thus had to reflect what instruments were owned by her (these are only known from an inventory made after her death); she wasn’t billed for two 61-key harpsichords until 1757 (Kenyon de Pascual 1987 p74; who also gives some evidence for 63-key harpsichords being made in España in the 1740’s)  The range and date of Farinelli’s gift harpsichord are not known (Pagano 2006, p335-337) but it may be the cembalo expresso mentioned in the incipit of P9:29 in C (its companion P9:30 dated 1754 in its Münster version)

The ranges represented within a particular collection may be related to the instruments owned by the collector commissioning it, but the dates of composition have to be established by other means. Put in a different way: the reason wide-range sonatas don’t show up in the Parma and Venezia collections until 1754 is that the owners of those collections didn’t own wide-range instruments until that date, not that the sonatas weren’t already composed.

Scarlatti was mentioned in a newspaper report of 1728 (see Curriculum under that year for citations) to be in Halle (northwest Germany near Leipzig). He had left Lisboa for Roma the year before because of poor health, with a handsome gift from the king, and did not return until 1729, with his new wife and son.  What was he doing in Germany?  Halle is close to Freiberg, Sachsen (about 120 km) the location of a pioneering maker of pianos, from at least 1721, Gottfried Silbermann 1683-1753.  Shortly after 1725 Silbermann began building pianos modeled on those of   Bartolomeo Cristofori (http://www.baroquemusic.org/silblegacy.html)  For the literature on the connection between Scarlatti and Cristofori see Ogeil bibliography, especially p165, articles by David Sutherland.  And of course Leipzig was the home of J S Bach (see P10:8 d for an unexplained Bach quotation) Perhaps Scarlatti was acting as an agent between the two piano builders, or purchasing one on behalf of his employer, Dom Antonio, who was keenly interested in the piano and later received the dedication of Giustini's Sonate 1732.

difficulty:

It is possible that Scarlatti had at least three pupils (or composed pieces for them to play): Maria Barbara, her uncle in Portugal dom Antonio, and Maria Barbara’s husband Fernando. Maria Barbara was by report a highly accomplished player; both of the men were less skillful. See Prozhoguin in Sala, p90-92 (and below in graphics) for another possible pupil. Alvarenga, in Sala, p37, indicates that Scarlatti became music master to Maria Barbara only at the end of 1729, when he left Lisboa to join the spanish court in Sevilla (see the Curriculum file). In Portugal before that time he was royal composer (Alvarenga, p44); keyboard pieces composed in the 1720’s in Portugal and elsewhere may have been primarily for his own use. When he became music master to Maria Barbara, he may have provided her with daily keyboard exercises as well as performance pieces to be played at court gatherings for important visitors and select audiences. However his own statement in the preface to the Essercizi implies that he did indeed compose pieces for both her and Dom Antonio while he was at the portuguese court. And Scarlatti as performer of his own works would be master keeping far ahead of his pupils. The organization of many of the compiled collections, not just Venezia and Parma, clearly reflect the composer’s awareness of different technical skills needed, or to be developed, by the performer.

I have listed the Parma libri and Essercizi in general order of difficulty and integrated the more difficult Scarlatti sonatas into Schumann’s categorized list of etudes published in 1836; see the section Clementi, Czerny, Chopin and Schumann: Scarlatti’s Gradus ad Parnassum at the end of the Contemporaries pdf file.

Although the chronology of the sonatas (at least those of the toccata or study type) should show increasing demands on a pupil’s abilities, and my about dates are influenced by my assumptions about this, it would be necessary to know which sonatas were intended for which pupils and when. The compiled collections need not reflect the order of composition in this respect. If the libri are intended to provide daily exercises, there would be a variety of different problems offered as reminders of techniques already learned, not arranged in a strict progressive order.

form:

Analyzers of form often start by defining a number of models, then try to fit each sonata into one or more of them: Gerstenberg, Kirkpatrick, Foster, Chambure, Campbell are all examples (see Citations file). I think their models all collapse when specific sonatas are forced into these sometimes ill-fitting shoeboxes. I have tried to identify some of the same features they use in my own maps, such as thematic cells, but I believe I am starting with the sonatas themselves and trying to extract their often unique patterns, rather than looking for a pre-conceived pattern. Each thematic pattern can be expressed by a line of numbers (123:43 for example) and then related to other sonatas. As I say in the introduction to Chronology, I have not yet found these thematic structures as useful as the key structures in bringing out patterns of other kinds which might, when all are put together, provide a convincing picture of Scarlatti’s sonata form developing over time.

Scarlatti thought a lot about binary form and how to make it more perfect or at least more interesting. He often aimed for non-stop forward motion but also for symmetry and balance. Sometimes he omits the internal repeat (V42:51 C, P5:23 G); sometimes the second part repeat (several Essercizi lack the double dots at the end). Sometimes the introductory material within the first part is repeated separately (P3:27 A in its V49:17 version). As a general rule the two parts are of equal length.

Why do so many of the sonatas in Venezia 1742 and 1749 have "D.C." (da capo) at the end, after each part has already been repeated? The repeats are for the benefit of the audience more than the performer. The listeners are hearing the piece for the first time and may never hear it again. So the full da capo is their last chance to have the piece stick in their memory as something interesting and satisfying for a moment of their existence.

ABA forms have one kind of musical symmetry. AABB is less satisfying from the point of view of balance but a complete da capo, AABBAB, helps. If there are links between the parts, the sonata can be heard as A AB BA B (the BA link is the ending and beginning in the tonic key)

In the later collections the da capo indication disappears. Change and contrast take over from a striving for symmetry and balance, which are more appropriate to the visual arts. Suites, then smaller groupings of paired sonatas, provide a narrative structure: this, then this. The sections repeated within and between the first and second parts are shorter. There are more challenges to the listener’s memory.

The key structures of the sonatas are discussed above under dates, in the Chronology file, and in the introduction to the Catalogue (page 5-6 of the pdf file)

Classical rhetorical terms can also be applied to the internal features of the sonatas with profit: see rhetoric in the introduction to the Catalogue (page 6-7 of the pdf file)

I have been able to find golden section ratios (φ) in almost all of Scarlatti's sonatas, measured for the entire sonata within one or two measures, either the long-short or the short-long ratio, rarely both in the same sonata in my opinion.  Most of these φ points occur at the start of a theme or a change of key, often both coinciding.  Each has been indicated on the Catalogue maps; the Characteristics file has a complete breakdown by type.  I think a recognition of the golden ratio opened up a wealth of structural possibilities for Scarlatti, which he seems to have taken full advantage of.  In addition to using the φ point to start a passage of intensified expression (Parma 3:5 C is a good example) he seems to have been fully aware that the creation of an imbalance in the sections of the sonata would counteract the excessive symmetry caused by a double bar for repeats at the central axis and help unify the work as a whole.  There are sufficient good examples of telling events which occur at a φ point to convince me that this is one of the things Scarlatti meant by his reference to "lo scherzo ingegnoso dell'Arte" in the introduction to Essercizi.  But it is important to try to understand what he is up to at the measured point and not stray too far from it, rather than impose a pre-conceived expectation upon it.  Scarlatti sonatas are full of events: what happens at the φ point is just one of these, but can lead to a fuller understanding of the moments that have special structural and expressive significance throughout.

The use of linking devices (returns to beginnings of parts, alternate endings of parts, elision of measures, omission of downbeats, use of key relationships at ends and beginnings of parts) to make a continuous whole of sonatas (and often groups of sonatas) was another important concern of the composer.  See the Capriccio file for more thoughts on how the golden section and repeats might work together through the links at the central double bar.

accidentals:

The performer can take advantage of seeming ambiguities which arise from the conventions for accidentals used within most of the sources (see this section in the introduction to the Catalogue) to create expressive effects and should not feel constrained to obey vague concepts of historical practice, editorial advice or academic rules. Statements such as "the closing tonality is never thrown in doubt" (Kirkpatrick p263) should be disregarded when the sources provide for other interpretations. Some examples: see P5:27&28 B, P3:5 C and E20 E in the Catalogue. Since we cannot know for sure what Scarlatti or his contemporaries would have done about the so-called wrong notes in Scarlatti’s sonatas, why not accept them as happy accidents? There is surely nothing in a Scarlatti sonata which could offend a twenty-first century ear.

ornaments:

Summary of my system of ornaments, discussed in greater detail in the introduction to the Catalogue and in the notes for individual sonatas:

Shakes are usually short and sharp (biting, the literal meaning of mordente), principal note-upper note-main note, they can be longer or smoother in appropriate contexts or for varying a passage when it is repeated.

Trills are longer and usually begin on the upper auxiliary note.

Tremoli are usually short, the same note is rapidly repeated three or four times with finger changes, but can be longer and can be tied over measure lines.

Slurs indicate ties, elisions, or rubati.

Slurs on small notes (as found in the sources, not those used in modern editions) are often ties when the previous large note is the same pitch: they are not sounded but suspend or delay the following large note.

When the small note is a new pitch, the slur indicates playing the small note before the beat or crushing it with the following large note on the beat (the literal meaning of acciaccatura).

If neither of these solutions for small note slurs seems appropriate, the slur can be used to indicate stressing the next large note after the beat (rather than leaning on the small note itself, which is the literal meaning of appoggiatura). The slur on the small note also indicates tempo rubato in the sense that the value of the small note is not its exact duration.

Small notes without slurs are appoggiature (again, as found in the sources, not modern editions which usually add slurs to most of the small notes): they are played on the beat, are stressed, and the following large note is delayed for at least the full value of the small note (depending on the context).

I don’t believe that any of the interpretations in my system of ornaments would have been unknown in Scarlatti’s era, and they do seem to me to be consistent with his treatment of similar passages written out in large notes. The advantage of this system is that it allows me to take the indications in the Parma score and other primary sources seriously. It also leads to some surprising but convincing results.

graphics:

The handwriting in the sources needs to be investigated more thoroughly by scholars. Only a few examples exist which are thought to be Scarlatti’s own handwriting: in his early vocal music written in Roma and a letter written much later in Madrid. Some of the keyboard sonata sources show a marked resemblance to the handwriting in these.

I have not provided illustrations myself. It would be misleading to draw conclusions from reproductions; I have not attempted to gain permission from the libraries owning the sources. Most of the examples indicated below can be seen in facsimiles which have already been published. Scarlatti scholars seem surprisingly reluctant to investigate the manuscripts any further than Sheveloff and Choi had done more than thirty years ago, whereas whole volumes on the handwriting and watermarks in the Handel and Bach manuscripts have been produced.

The handwriting in Münster volumes 1-3 (see the Kirkpatrick/Johnson edition v15:452 & 453 for facsimiles from M2) shows similarities to that of Bologna ms FF232 & Torino ms394, as well as the 1744? or 1752? letter to Huéscar in Scarlatti’s own hand, and his autograph scores for the Miserere in e minor and Tolomeo sinfonia (Kirkpatrick plates 39, 21 & 22) See also Prozhoguin 2010 p118 ill3 for another example of Scarlatti's handwriting and signature.  Therefore it seems possible that the main scribe of Münster libri 1, 2, 3, 5F and 5H, might be Scarlatti himself (the title page of Münster 1b which refers to Scarlatti’s death in 1757 is in a different hand). Some of the characteristics of these examples: the letters are sometimes separated and sometimes connected; the bass clef sign is backwards, which Choi p46 says is typical of 18th century italian musical handwriting (and mentions Giuseppe Scarlatti, Domenico’s nephew, as an example); 8’s and g’s are open at the top; 2’s have a full loop (like capital Q’s but not connected at the bottom), d’s have the stem bent to the left; abbreviations use a colon rather than a period; there is a characteristic double bar at the ends with double diagonal bars at the top and bottom and strokes like = signs rather than double dots (approximately =//||//=); the similar writing of the words Segue and Fine at the ends of parts. There are variations of the same letters and numbers, explainable by parts being written carefully and somewhat decoratively (as an architect might on a building plan) and others in haste (as in a signature); also there is the lapse of time between the manuscripts to take into account. Thus there is a decorative 3 used for the meters in the M1-3 sonatas and index incipits, as well as a more ordinary three used elsewhere. I admit that the Parma scribe has some of the same letter formations but they seem quite different in the ensemble. All these music and text handwritings, as well as the watermarks of the papers used, ought to be checked more carefully by would-be experts.

Serguei N. Prozhoguin, Rileggendo la lettera di Domenico Scarlatti, in Sala p69-154, gives detailed graphic characteristics of the letter to the duke of Huéscar, p75-77, as well as for the Parma 1-15 &Venezia 1-13 scribe, p124-128. On p111 Prozhouguin mentions the date 1794 on the reverse side of the letter, illustrated on p72, the title page of the motets Scarlatti has transcribed for Huéscar, "Panegyricum Heroicum ... MDCCXCIV" -- but doesn’t state that it could be an error for 1744 (the last C might be a careless L), the year for which Prozhoguin offers other evidence, p84-92. (He also thinks the tone of the letter more appropriate if it were addressing the 11-year-old son of the 30-year-old duke, with other interesting reasons for believing the son may have been a pupil of Scarlatti’s at that time) The date 1752 occurs as a later marginal annotation on this title page (that is the date assigned to the letter in Kirkpatrick and other sources) See p70 of Prozhouguin’s article however: he thinks the title-page must have been written long after the letter--but how could 1794 be at all likely?

Some other graphic characteristics common to more than one source which might reflect unconsidered copying from an autograph by different scribes: key signatures which are only at the incipit or sometimes erroneous at the incipit (Bologna 1, Cambridge-Fitzwilliam 13, Venezia 1742: 3 & 52, Lisboa 33); whole notes in the middle of a measure (Münster 1; Parma 8-15); the use of roman numerals to number sonatas (Essercizi, Lisboa, Münster 1, Venezia 1742, 1749, 1, London-Worgan).

P14:7 D (Münster version) & P12:16 g (Bologna 1 version) contain canceled passages which use very similar means of crossing out.

Francesco Degrada, in his edition of La Dirindina (Milano : Ricordi, 1985), Scarlatti’s intermezzo of 1715, reproduces the first page from the manuscript source (Venezia, Biblioteca della Fondazione Levi ms 702b; p.vi of his prefazione) This manuscript has all the same handwriting characteristics as the sonata manuscripts in Bologna 1 and Münster 1-3 described above. The text letters are formed in the same way, the braces at the ends of ledgers, the bass and C-clefs, the common time and alla breve signs are identical to those use in M1-3. Degrada, p.ix & 129, discusses the manuscript volume as a whole, the watermarks, and the text used (this copy contains the earliest version of the libretto to survive). He concludes that it is a copy made for a collector or one of the singers. He does not discuss the handwriting.

Interestingly, the manuscript page reproduced by Degrada contains a joking solfeggia: Dirindina, probably sung by a tenor falsetto, sings a scale in whole-notes: d , r , m , f , sol, d ; the final d should be at the same pitch as the first one, but Dirindina lands a full step too high (Boyd p74 ex). Scarlatti’s name appears in the libretto p.xliv, fn17 (see also Sutcliffe p72) near the beginning of part 2 in a passage with a reference to gambling:

Terremo in casa il giuoco / quando sarem colà; / farem far delle riffe / a quella nobilità / d’oriuoli, d’anelli e di merletti / di vezzi, di scarlatti e d’orecchini...

(When Liscione and Dirindina get to Milano, they will open a gaming house and hold lotteries with the nobility to obtain jewels and costumes, among which will be scarlet silks). This is in a recitativo; Scarlatti sets his name to the notes re mi mi, referring to his nickname Mimo (page 83 of Degrada's edition)

I admit that calligraphic scribes are supposed to learn their craft by imitating their models as closely as possible. But if that is true in the cases discussed above, the scribes must all be working from the composer’s autographs. I find it easier to believe that Scarlatti sometimes prepared his own fair copies for collectors than that he had a personal scribe who followed him around Europe from 1715 to 1757.

influences:

Similarities between specific keyboard works of Scarlatti and those of his contemporaries are cited from various authors in the individual entries in the Catalogue.  These show that there are many resemblances with the keyboard works of Rameau and J S Bach, for example. Specific Scarlatti influences on later composers such as Clementi and Beethoven are indicated as well. A section at the end of the Contemporaries file integrates Scarlatti’s sonatas with Schumann’s list of etudes published in 1836.

Resemblances of Scarlatti’s own sonatas with one another are listed under the separate heading similar sonatas in the Catalogue pdf file.

metaphors:

In spite of what many writers state, there are programmatic ideas at work in Scarlatti’s music. See this section in the introduction to the Catalogue pdf file. Just one example: the juxtaposition of hunting motifs and birdcall motifs in the same sonata or in companions (P1:9-12 C, P6:6 C; P10:15-18 G; P11:29-30 C). He was a composer for the theater. This aspect of Scarlatti’s work needs to be brought into focus. Perhaps performers will start using their brain (both sides) as well as their fingers.

April 30, 2007; revised through February 22, 2012