Stanley Yates - Recordings



Fernando Sor - Second Sonata for Guitar (op. 15b)
© 2001 Stanley Yates

This article is based largely on discussion contained in my extended essay dealing with all of Sor's sonatas ("Sor's Guitar Sonatas: Form and Style") written for the Sor Studies project (a multi-volume collection of scholarly articles to be published by the Instituto Complutense de Ciencias Musicales a the Universidad Complutense, Madrid, edited by Luis Gasser).



1. Introduction
2. Early Publications
3. A New Modern Edition
4. Style and Composition
5. Interpretation and Performance
6. Notes
7. Online Edition

1. Introduction [top]

Though Fernando Sor is widely recognized as the principal guitar composer of the classic-romantic period, his catalogue is relatively modest in terms of large-scale works for the guitar - a result, at least partly, of the considerable energies he expended at various times in his career on vocal music and the grander genres of opera and ballet combined, no doubt, with the demands of his publishers for marketable items of guitar music. Consequently, Sor's output of sonatas - the principal compositional genre for instruments of the time - was sporadic and totaled only four named works (along with a fantasia in sonata form). Nevertheless, these works, which span Sor's compositional career, are ambitious compositions which allow us a glimpse of their their composer at two very different stages of his development.

The op. 15 sonata - a single-movement work falling at the beginning of Sor's career, and his most modest effort in the genre - is a lively, attractive work, and one from which springs, with a little digging on the part of the interpreter, the excitement of the Italian opera overture.

In this article, I discuss the early publications of Sor's second sonata, and their implications for a modern edition; Sor's compositional and stylistic models; a conjectural date of composition; and performance and interpretation for the modern guitarist.


2. Early Publications [top]

Although multiple editions of this work published during Sor's lifetime survive, two essentially distinct versions emerge: a presumably early version published as "Sonata seconda" by the Spanish expatriate Salvador Castro de Gistau in Paris between c. 1802 and 1814;[note 1] and a revised version, first published by Antoine Meissonnier as "Sonate" at sometime between c.1816 and 1821,[note 2] also in Paris. A transitional version (apparently generally unknown to modern scholars) also survives; it appeared in a "Collection de Morceaux Choisis des Meilleurs Autuers Espagnoles et Italians pour Guitarre ou Lyre, Rédigée par un Espagnol" as "Sonata di Sor," in Paris sometime between c.1814 and 1821 (publisher unknown).

Since we know nothing of Sor's involvement with Castro, we therefore know nothing about the authority of his editions of Sor's music. Early guitar works of Sor's circulated in Spain in a variety of manuscript copies, some of which have survived to the present day but none of which appear to be autographs (one such manuscript of the work, titled "Allegro," is presently archived at the monastery in Montserrat). The work could therefore quite easily have been published without Sor's involvement (especially since Sor was still in Spain when the Castro edition appeared, in Paris).

We can be more confident of the authority of the revised version published by Meissonnier, since the composer, no longer in Spain, provided this publisher with numerous works, both old and newly composed. Nevertheless, the first appearance of the Meissonnier revised version contained many errors and the piece was re-issued during the same period (c.1816-1821) with the same plate number (110), but in a corrected version. Although the first page of each are identical, the remaining pages were re-engraved with corrections. Sor presumably played a role in these corrections, and this second Meissonnier version may be regarded as the most reliable source. This same corrected edition was published again by Meissonnier, in partnership with Heugal, from the same plates, sometime after c.1839.

A further edition published during Sor’s lifetime appeared in Germany, issued by Simrock around 1824-5 (plate no. 2310); it is in this edition we first encounter the erroneous designation of the work as "op. 15" - something which has led to the work being cataloged as "opus 15b" by modern scholars. Unfortunately, the Simrock edition was derived from the first, error-ridden version of Meissonnier, though the German editors did attempt to correct some of the more obvious errors.

The work was also likely published in Spain in a relatively early version which hasn't survived: in October and November of 1806 the Gazetta de Barcelona advertised for sale in Madrid a "Gran Sinfonia" and "una sonata," both for solo guitar by Sor.[note 3] The "Gran Sinfonia" would appear to be Sor's Sonata prima (as it was called in the Castro edition), also known as Gran Solo, op. 14 (as published by Meissonnier): a manuscript copy of the piece titled "Gran Sinfonia" is housed at the monastery in Montserrat. It is therefore reasonable to guess that the "Sonata seconda" (Sonata, op. 15) and the work advertised as "una sonata" are one and the same. Additionally, it seems unlikely that a further sonata by Sor existed, since Meissonnier assigned the majority of the first 20 opus numbers to Sor's back-catalogue before concentrating on newly-composed works (when the back-catalogue had presumably been exhausted). Surely, neither Sor nor his Parisian publisher would have left a major earlier work unpublished.

The surviving sources of the work, along with their principal differences, are as follows:

Manuscript: Monastery of Montserrat.
Titled "Allegro" -- an apparently early version of the piece in an unknown hand.

Source A: Castro (Paris, c.1802-14) (plate no. 47)
"Sonata seconda pour la guitare, composée par Fernando Sor," in Salvador Castro’s periodical, Journal de Musique Etrangère pour la Guitare ou ("Journal of Foreign Music for the Guitar or Lyre"). Judging by the plate number (47), Castro may have published this work before he published the Sonata prima (plate no. 50). This version contains several extra measures not found in any later versions, along with numerous small differences in the detail of chord voicing, ornamentation, and notated texture (of which it is the most polyphonically-complete representation of all sources).

Source B: Anon (Paris, c.1814-18) (plate no. 81)
"Collection de Morceaux Choisis des Meilleurs Autuers Espagnoles et Italians pour Guitarre ou Lyre, Rédigée par un Espagnol," as "Sonata di Sor," published between c.1814 and 1821 (publisher unknown). This version is marked "Allegretto," like the Castro version, has the same coda, and contains several pitch differences otherwise found only in the Castro version. It does not have the extra material of Castro, nor the same ornaments or beaming, and has much in common with the presumably later Meissonnier versions. It is significantly close to the these versions as to suggest that Sor may have supplied the manuscript himself. Certainly, the revisions present are more of a compositional nature than of an editorial one.

Source C: Meissonnier (Paris, c.1816-21) (plate no. 110)
Although Sor almost certainly provided the manuscript for this edition, numerous errors found their way into print, including the omission of the final coda. It is interesting to note that the notation of this version (and all subsequent ones) is actually less polyphonic than the Castro version. The tempo now appears as "Allegro moderato," as it does in all subsequent versions.

Source D: Meissonnier (Paris, c. 1816-21) (plate no. 110, but with pages 2-3 newly engraved)
This version corrects virtually all of the errors found in source C, and reinstates the coda (though a slightly different one than that given in Castro or source B). The most definitive version of the piece.

Source E: Meissonnier-Heugal (Paris, after 1839) (plate no. H. & C. 110)
An exact reprinting of the corrected Meissonnier version, source D, from the same plates.

Source F: Simrock (Bonn and Cologne: c.1824-5) (plate no. 2310)
Based on the first Meissonnier version (Meissonnier sold many of Sor’s pieces to this publisher), with an editorial attempt to correct a few of the more obvious errors.


3. A New Modern Edition [top]

Two versions of the piece are presented here: the presumably early Castro (source A) edition and the corrected Meissonnier version (source D). The latter, however, may certainly be regarded as the more definitive version of the piece (since it can only have been informed by a corrected manuscript provided by the composer). It is not recommended that players mix together in performance passages from what are essentially two quite different versions of the piece. Some of the notational clarity of the Castro edition has been applied to the Meissonnier edition (and indicated) however, principally in the separation and realization of durations and rests in the lower voice (otherwise these are often indeterminate, sharing a stem with an inner voice -- see the development section, for example). Small differences (principally with Source B), which might appropriately inform a modern performance of the piece are provided as footnotes to both versions. Though none of the early editions contains a single dynamic indication, a few obvious markings have been supplied editorially (see Interpretation and Performance, below).


4. Style and Composition [top]

Sor's op. 15 sonata is a single-movement work modeled after the Italian opera overture of the 1780s and 90s. The style was most widely represented at the time by such figures as Spontini, Paisiello and, above all, Cimarosa -- composers whose music Sor almost certainly heard during his involvement with the Barcelona opera house (in 1797, the nineteen year-old Sor had his opera Telemaco produced there). In common with Sor's early style, the stylistic features of the Italian overture (at the time termed "sinfonia") is characterized by very direct, non-developmental forms in which sections are defined by clear points of arrival (rather than seamlessly connected). The opening typically consists of a short head-motive followed by an immediate launch into a lively first theme consisting of parallel-thirds (in the violins) supported by a repeated-note bass; compare the opening measures of both Sor's Sonata prima and Sonata seconda (opp 14 and 15b):

Both follow an identical scheme of a short assertive opening "head motive," followed by parallel "string" thirds over repeated-note bass (trommelbass), followed by dialog texture.

Phrase groupings in the Italian overture style, in common with Sor's style in the early sonatas, are symmetrical and periodic with literal repetition of phrases and periods as the principal means of developing sections. Transitions are highly rhythmic, and are usually built on one or more commonly-used motives. In Sor's op. 15 sonata, we find prominent use of two such operatic motives. The first is a figure ubiquitous to the overture since its earliest days (Pergolesi's overture to S. Gugliemo, Naples, 1731, for example, is based almost entirely upon it):

Sor uses this figure at each transition, in both codas, and to announce the "development" section.

The second figure is somewhat hidden in Sor's uncharacteristically imprecise notation:

This is one of the most-used clichés of the Italian opera overture, form the early overtures of such Neapolitan composers as Lampugnani to those of Paisiello and Cimarosa, and beyond. The figure is also found in many symphonies of the time, including some of Boccherini's, and is a particular feature of Pleyel's style, appearing with great frequency in his transitions (it is also used often by Mozart, though not by Haydn). The figure is notated unambiguously in both Sor's Sonata prima (Gran solo, op. 14) and Grand sonate, op. 22, by the way.

A further feature of the overture first transition (i.e., the rhythmic passage that "loosens" the tonic key and moves to the secondary theme area) is that it doesn't actually modulate but proceeds on static harmony - dominant becomes tonic (rare in the high-classical symphony and other highly-developed sonata forms); the new dominant is reached, but is not reinforced by it’s own dominant. The relatively weak arrival at the secondary area allows for the extension of the secondary theme group: the first theme of the secondary key-area is followed with a more emphatic second transition that does now create a well-marked arrival on the new dominant; this is especially the plan used by Cimarosa in his longer overtures, and is employed by Sor in all three early sonatas (opp. 14, 15b and 22).

This transition is followed by a second thematic statement in the new key, usually defined rhythmically rather than melodically. In the Sonata seconda, this second theme is similarly defined by its rhythmic level and texture, as is confirmed when the material is restated in the recapitulation with only vague pitch resemblance:

The development section of the Italian overture typically begins with a striking, "gratuitous" modulation, functioning as an area of contrasting tonality rather than as a place to develop previously heard material. Compare the opening development keys in Sor's early sonatas: in op. 14 the dominant A-major moves to a startling Db-major; in op. 15b we move from G-major to the very odd Bb-major; and in op. 22 we move from G-major to Eb-ma (actually a standard "operatic" modulation to the dominant-Neapolitan key).

The early sonatas of Sor are not without their peculiarities -- features that suggest an ambitious and intuitive young composer not yet fully in control of his forms. In op. 15, the theme that follows the first thematic period re-cadences on the tonic (at m.22) before moving off again to the dominant, having the curious effect of neither closing the first theme nor moving away from it. The core development key -- the lowered-seventh -- is very unusual. This tonality is entirely unrelated to the large-scale harmonic structure of the movement, not functioning (in the customary manner) as a dominant to the relative-major of the tonic-minor, nor as any kind of Neaplolitan dominant (i.e., bII to V or V / V). Sor does not resolve this tonality (as, for example, bII of V / V / V in the large scale tonic), but instead moves to the supertonic (the large-scale tonic minor) and a tonic-minor prolongation of the large-scale dominant.

On a more positive note, the early sonatas of Sor show distinct motivic relationships between the various structural areas of the form, as well as a very strong preference for repeated-note themes. Although melodic repeated-note motives are hardly unique to Sor, his apparent method of associating them, probably intuitively absorbed from the Italian overture, is a quite individual characteristic of his early sonata style.

Like the first sonata, an anticipatory retransition "leads-in" to the recapitulation with motives adopted from the forthcoming (first) theme itself:

First theme motive (b) appears prominently in the secondary theme area of the exposition (mm.37 and 45) and the recapitulation (mm.142 and 150):

The core development episode appears to be built on a lower-voice presentation of the opening motive of the piece (in the later editions this is somewhat hidden in the uncharacteristically imprecise notational texture):

And a second development episode relates to the second theme of the dominant area:


We have already mentioned Sor's prominent use of overture motives at the structural points of this work.

When might Sor have composed this piece? Since publication dates establish only an upper limit for the date of composition, in the absence of dated autograph manuscripts, letters, or other compelling evidence, the precise dating of the piece remains a matter of conjecture. All we know for certain, is that it must have been written before 1814 -- the earliest verifiable date of publication. But the work was surely composed much earlier than that. The likelihood of earlier publication in Spain, as advertised in the Gazetta de Barcelona, would place the work before 1806, while the stylistic details outlined above would indicate that the work could easily have been composed during the time of Sor's assimilation of the Italian opera overture style -- that is, during his Barcelona period of around 1796-1800 (as mentioned above, Sor's opera in the Italian style, Telemaco, was produced in 1797). There is one further bit of information, provided by Sor himself to Ledhuy and Bertini, compilers of the Encyclopedie Pittoresque de la Musique (Paris, 1835). In the extended and detailed article on Sor (no doubt written in the third-person by Sor himself) it is mentioned that Sor did not yet dare to compose an overture in the style of Haydn for Telemaco in 1797, but that he has begun a study of string quartets by Haydn and Pleyel before leaving Madrid. That is, before 1800. While clear elements of Austro-French style are found in Sor's op. 22 sonata (particularly those of the Italianate Viennese composer Pleyel), no such traits are present in the first two sonatas. In other words, if Sor composed the first two sonatas after his study of Haydn and Pleyel, neither work shows any sign of it. It is therefore reasonable to place the two sonatas before 1800, written at the latest by a student composer in his very early twenties.


5. Interpretation and Performance [top]

Without hearing Sor play, and without detailed evidence relating to his performance style, it obviously is impossible to state anything certain about the way he intended his music to be performed. Taking Sor's ouevre as a whole, however, in addition to certain comments found in his Methode, it is clear that Sor did not perform with a brilliant virtuosity. His playing must surely have been light, graceful, and tasteful; one imagines a refined phrasing supported by a beautiful sound, rather than speed supported by volume (virtuosity in the modern sense). With this in mind, and considering both the timbre and scale length of Sor's guitars, I enjoy playing Sor's music, on a modern guitar, with a capo at the first fret (an instant forte piano). This can be quite revealing, and is quite likely to inform an interpretation, even if a player intends to perform the piece without the use of a capo.

Sor rarely provided dynamics markings in his pieces; the op. 15 sonata contains none at all. While we can only guess at the reason, it is obvious that this music is not intended to be played mezzo-forte throughout! (Sor probably felt that dynamics appropriate to the style were obvious). In the op. 15 sonata, dynamics implied by the form and style of the piece are fairly obvious, and have been supplied editorially in this online edition. Dynamics resulting from natural phrasing remain the prerogative of the performer.


6. Notes

1. For information relating to the dating of the various publications of Sor's music I have relied upon Brian Jeffery, Fernando Sor: Composer and Guitarist (Second edition. London: Tecla Editions, 1994), and presently unpublished research provided by Erik Stenstadvold. [Return to Text]

2. This dating, which conflicts with that of 1817-22 given in Jeffery (1994), is based upon information provided by Erik Stenstadvold, whose research into Meissonnier suggests the publisher occupied the address Rue Montmartre No. 182 (the address indicated on the publication) from May 1815 until, at the latest, April 1821. [Return to Text]

3. The "Gran Sinfonia" was advertised on November 1, "una sonata" on October 4. Mangado, La Guitarra en Cataluña, 1769-1939 (London: Tecla Editions, 1998), 288. [Return to Text]



Copyright 2001
Last update: 8/01