Improvisation In Mozart`s Violin Concertos

This article is an investigation into Mozart s improvisation and ornamentation practice, with a particular focus on the relevancy of this research to the performance of Mozart`s five violin concertos. It is actually a shortened version of a dissertation I submitted as part of my Master`s degree, anyone wanting to look at the whole thing or discuss it further should feel free to email me.

Improvisation was a crucial skill for virtuosi during the 18th century, whether in the form of improvised embellishments, cadenzas and preludes or the improvisation of larger scale pieces like fugues, fantasias and even sonatas. Contemporary reports testifying to Mozart s skill as an improviser abound, for example this extract from the Gazetta di Mantova which dates from the 19th of January 1770 (shortly before Mozart s 14th birthday): "The incomparable boy Sig. Wolfgang Amadeus Mozart improvised two sonatas on three themes successively given to him by the leader of the orchestra, elegantly linking them together the second time And what is most to be esteemed, he composed and at the same time extemporaneously performed a fugue on a simple theme given to him, which he brought to such a masterly harmonic interweaving of all the parts and so bold a resolution as to leave the hearers astounded."

The role that improvisation played in violin playing during Mozart`s time is less clearly documented. The earlier Italian style epitomised by Corelli and Tartini, where improvised ornaments were a vital showcase for the player s judgement and wit, was being gradually superseded by the more proscri ptive style that increasingly became the norm in the 19th century. To simplify, the decline of the importance of improvisation for 18th century violinists ran in parallel to the rise of the German mixed style (defined by Quantz as taking the best elements of the French and Italian manners and blending them) over the Italian as the century continued, and the increased popularity of the French violin school over the Italian, the former eventually cementing it s preeminent position during the 19th century after the founding of the Paris Conservatoire. Despite this, improvisation was evidently still a sufficiently significant tool in the soloist s arsenal during the 19th century for Pierre Baillot to devote a large section of his 1834 L`art du violin to the improvisation of ornaments, cadenzas and preludes.

As improvisation was essentially restricted to concertos, sonatas, or solo passages in chamber music, the amount of material evidence of improvisation in Mozart`s creative process is more numerous and has been more carefully studied in his keyboard music than his violin output, so I will begin there. Mozart was in the habit of improvising extensive ornaments to his piano concertos when performing. Although for obvious reasons he never wrote his ornaments down, it is possible to establish their character because examples of ornamented slow movements, recapitulations and repeats were occasionally prepared by Mozart for his pupil s study, or prepared by his pupils under his tutelage for their own performance. One of these is a document prepared by one of his pupils, Barbara Ployer, to the slow movement of K488. The florid nature of these ornaments might be surprising and not entirely in keeping with our view of Mozart`s style, the running chromatic passages being more reminiscent of Chopin than the master from Saltzburg. Despite this, the importance of the document is not the quality of the ornaments, which are rather over-reliant on chromatic scales and stepwise motion, but the quantity. As Robert Levin points out, "it is extremely unlikely that Ployer would have written demisemiquavers if Mozart had only played crotchets and quavers."

If improvised ornaments were commonplace, are there generic places where they might they be expected or required in a composition? One location where they are so indispensable as to be practically compulsory is in the return of the principal theme in Sonatas, particularly in Rondos. An insight into Mozart s creative process in this area can be found by looking at the autograph of the Fantasy and Sonata in C minor, K475/457. In the autograph, the recurrences of the principal theme in the slow movement are signified with the direction `Da Capo 7 sieben Takt` they are not written out. On a separate sheet of paper, much in the same way a cadenza might be expected to be found, Mozart has written diminutions for each return on behalf of his pupil Therese von Trattner, the sonata s dedicatee. This document reveals a notational practice vital to understanding the question of ornamentation - the shorthand da capo marking does not indicate a note-for-note reiteration of the opening music, but implies an ornamentation of it, preferably an improvised one.

There are specific points in Mozart`s music that lend themselves to ornamentation to such an extent that it must be treated as an essential, much in the same way that cadential trills continue to be mandatory today. Firstly, in several of the piano concertos, the theme is unadorned at reprises in the solo part, but decorated in the tutti ritornello that follows. Levin describes this as "an essential contradiction of the relationship between solo and orchestra" - it is contrary to the underlying assumptions of virtuosity and invention that existed at the time for the soloist to present a less elaborate theme than that played by the orchestra. Levin insists that the soloist must play "something as active in the antecedent as will be heard from the orchestra in the consequent."

Secondly, movements with repeat signs invite embellishment of both the major themes and subsidiary material. It is highly unlikely that repeat signs imply a literal repetition of the text, and far more likely that they indicate a repetition of the underlying music with a variation of the surface material. Similarly, the return of the theme in Rondos invites variation, especially if the theme returns verbatim in the musical text. It is instructive to look at Mozart s own written out decorations from works like the A minor Rondo K511, where at each return the theme is ornamented more virtuosically than the previous appearance.

Thirdly, there are passages in the piano concertos where Mozart employs shorthand for example K491/i where scales in semiquavers are written out in full in bars 257-260, but only dotted minims appear the following 2 bars, indicating in shorthand the range that the scales must cover without the trouble of writing them in full. A famous example of Mozart omitting material in this manner is in his concerto K451. A letter from 1784 to his sister explains: "It is quite true that there is something incomplete in the andante of the concerto in D in the solo passage in C to which you refer, and I will supply the deficiency as soon as possible, and send it together with the cadenzas." The final phrase, `together with the cadenzas` is revealing it is universally accepted that Mozart improvised his cadenzas, whereas his sister Nannerl and most of his pupils preferred to have a cadenza prepared. The fact that the filling material for the passage was included with the cadenzas he sent to his sister strongly suggests that passages like the one in question, where melodic and rhythmic activity slacken for no obvious expressive or dramatic reason, do not have something inherently `missing`, but that in his own performances Mozart would have put flesh on the bones.

Although Mozart improvised his cadenzas, several examples survive of cadenzas he chose to write down, either for publication or for his pupils. Mozart`s cadenzas tend to be somewhat shorter than later ones, around 10% of the total length of the movement. Daniel Gottlob T rk, in his 1789 `Klavierschule`, laid down 10 rules for cadenzas which can essentially be summarised as follows: Cadenzas should be short, non modulatory, feature a selection of themes from the movement as a whole, and be charming rather than ostentatious. They should ideally be improvised, though the less talented would be better served preparing and memorising a cadenza according to his instructions than busking one that might go awry. Mozart`s cadenzas follow this model In general, early cadenzas (prior to c.1780) tend to be shorter and simpler, relying more on arpeggiated chord sequences than melodic material, whereas later ones use more of the motifs from the movement as a whole.

How representative are the published cadenzas of what Mozart might have performed in concert? It is instructive to consider at this point who Mozart intended the concertos to be performed by. Of the 21 piano concertos under consideration, six do not have surviving authentic cadenzas. Of those six, two concertos were never fully completed (K491 and K537) for example K537 has passages where the baseline of the piano is missing, and K491 has passages which are chaotic, occasionally even at odds with the orchestral parts. Of the other four, we know through Mozart s correspondence that two of them had cadenzas written to be sent to Nannerl, leaving two out of 19 that are fully notated but don t have a surviving authentic cadenza. It seems reasonable to assume that they once did but that they have gone missing. It is common knowledge that most of the piano concertos were written for Mozart himself to perform, or at least that the were premiered by him before being disseminated. Perhaps the incompleteness of K491 and K537 is indicative of his creative process and suggests that Mozart revised and fleshed out his concertos after he performed them to prepare them for publication. This might suggest that the cadenzas are not necessarily representative of his own performances, which would have had less codified material and would therefore have been more free. However, just because Mozart was perfectly capable of improvising a cadenza and no doubt did so, does not discount the possibility that on occasion he might have chosen to play one that had worked well before, either because it was particularly good or to spare himself the effort. Mozart did not have an either/or situation but a range of possibilities, as no doubt does the modern performer today.

Having got an idea of the lie of the land, let s move on to look at the violin repertoire. Improvisation for violinists of the period can be subdivided into four main areas, the variation of melody, the decoration of fermatas, and the extemporisation of cadenzas and preludes. I will examine the last three together, but begin with the first.

The addition of improvised ornaments was central to the Italian violin school of the late 17th and early 18th century, the most famous example being Corelli s op.5 sonatas, first published in 1700. Various ornamented versions of op.5 emerged soon afterwards, some claiming to be transcri ptions of what the composer played in performance. The fact that ornamented versions of op.5 survive in number is testament to the difficulties of improvising such extensive ornaments, especially if the musician had been educated in a more `straight` tradition like that of France or England. The Italian style was seen by many foreign theorists (for example Quantz, Mattheson or Schiebe) as somewhat arcane, Quantz defining "the Italian manner of playing" as "capricious, eccentric, affected and obscure, likewise oftentimes audacious and bizarre" conceding that because of its reliance on "a considerable addition of embellishments" it necessitated a "fair knowledge of harmony".

In contrast, the French style descended from the likes of Lully and Couperin was more strictly notated, with ornaments not necessarily written down in full, but more usually represented by a number of symbols which were often specific to the composer or even the work. The German style was most often described as a mixture of both Italian and French Quantz describes it as a happy medium. "If one has the necessary discernment to choose the best styles of different countries, a mixed style results that, without overstepping the bounds of modesty, could be called the German style."

During the second half of the 18th century, the three schools of ornamental theory gradually merged into some measure of agreement. The French system of symbols was clarified increasingly thoroughly by theorists including Quantz, Marpurg, Agricola and CPE Bach. For violinists, the celebrity enjoyed in Paris by Viotti and Tartini, who s Trait des agr ments (1771) was the first publication devote entirely to ornaments to appear in France, contributed to the emergence of a more international language as the century continued.

When and how to vary a melody was a skill largely passed down from teacher to pupil, but in much the same way as CPE Bach`s Sonatas with Varied Reprises do for the keyboard, we are allowed a window into what might have been expected through pedagogical publications. One of the treatises which treats the subject of improvised embellishment most thoroughly is Leopold Mozart s `On the Fundamental Principals of Violin Playing`, which for obvious reasons merits careful investigation. Anyone wishing to add ornaments, whether improvised or prepared, to Mozart junior s violin concertos would do well to use his father s guide as a starting point.

Longer improvisations took the form of Cadenzas, Eing nge and improvised preludes. During a trip to the British Library I found several pedagogical works containing model cadenzas and preludes some of the most pertinent are books by Carlo Tessarini, Stephen Philpot, John Tashenberg and Luigi Borghi. A survey of the cadenzas in these books reveals that they fit the mould discussed above, in that they are brief and non modulatory and predominately consisting of broken chords, arpeggios and short scale passages. For obvious reasons they are non thematic, but often contain fermatas in the middle to allow a performer to insert themes from the concerto.The Borghi contains one in A which would be a rather nice starting point for a cadenza to the first movement of Mozart s fifth concerto, as the second half (after the fermata) bears a close resemblance to the first theme of the movement.

Mozart`s early letters contain a few references to his violin concertos, the most interesting of which is a descri ption given by Mozart of a concert he gave in Augsburg in October 1777, at which he performed his 3rd violin concerto: "I the evening I played my Strasbourg concerto [K216], which went like oil. Everyone praised my beautiful clear tone. Afterwards they brought in a small clavichord and I improvised and then played a sonata and the Fischer variations. Then the others whispered to the Dean that he should hear me play something in the organ style. I asked him to give me a theme I put it through it s paces and in the middle (the fugue was in G minor) I started off in the major key and played something quite lively, though in the same tempo and after that the theme again, but this time arseways the Dean was absolutely staggered. I then played a another solo, a fugue in C minor, and then all of a sudden a magnificent sonata in C major, quite out of my head, and a rondo to finish with." Despite his assurance to his father that a `beautiful clear tone` was a priority (Leopold`s treatise outlines his opinions on the importance of a pure sound and a certain modesty of technique), it seems highly unlikely that the 21 year old Mozart, quite obviously very taken with his ability to improvise, would have restricted that skill to the keyboard. It seems much more likely, although impossible to prove beyond doubt, that the young Mozart would have added embellishments to his violin concertos in much the same way as he did in his keyboard concertos. It is certainly beyond all reasonable doubt that he would have improvised his cadenzas like any other violin soloist of his standard.

There are a number of tools at the disposal of the violinist when it comes to adding embellishments above and beyond what we might think of as ornaments, for example changes of bowing and articulation, addition of vibrato and subtle changes of dynamics. These are particularly important in the outer movements where quicker tempi limit the addition of filigree. Vibrato should be used cautiously when a continuous vibrato is avoided, it can be used at specific points to highlight phrasing or emphasise harmonically interesting notes. For example, in the first solo entry (adagio) of K219/i, the violinist could add vibrato on the first note (A) or the third (E), depending on whether they want to phrase away from the downbeat or phrase to the highpoint of the bar. Alternatively, they could refrain from vibrato for the first bar and use it first on the G natural on the downbeat of the second bar, which would give a more balanced shape to the phrase. Vibrato can also be used to give direction where the music is more static, for example in bar 43 of K216/i. If each of the three Cs are given a little more vibrato than the previous one it helps propel the phrase to the start of the next bar.

Dynamics can be used to clarify antecedent and consequent phrases. In K219/i, the conversation between characters in the solo part in bars 74 - 77 can be heightened if bars 74 and 76 are played more strongly than 75 and 77. This is applicable elsewhere, for example K216/i bars 64 - 67, where the more martial figure in 64 and 66 lends itself to a louder dynamic than the more lyrical one in 65 and 67. Bowing and articulation can be used to vary melodies, for example one of the returns of the theme in K219/ii could be bowed with three notes under a slur, followed by a staccato semiquaver, rather than all of them in pairs. String choice could also be considered a certain type of ornamentation. On a couple of occasions Mozart marks una chorda over a phrase, indicating that the player must shift into higher positions to play it on the same string rather than cross strings. Colour changes can be emphasised in a similar manner by the player on the spur of the moment.

The slow movements provide greater opportunity for melodic variation. We can take courage from the fact that Barbara Ployer`s ornaments to K488 which are much more florid than we might expect in a traditional view of Mozart s style. As I discussed above, Robert Levin identifies several types of passage which invite embellishment. One of the most crucial is where rhythmic and melodic activity slacken for no immediately obvious expressive reason. There are several passages like this in the slow movements of the violin concertos, one of the clearest being 39-41 of K211/ii, where the strings play one chord per bar in quavers and the violin has held notes where it previously had semiquavers. Other passages similar to this include K211/ii 47-56 and K207/ii 37-41 and 65-72. These passages have a tell-tale orchestration the lower strings play one bass note per bar and the upper strings play chords in quavers, creating a spacious and harmonically slow moving backdrop to the solo part. In the slow movement of K216, the theme returns several times in the solo part with practically no ornamentation there is certainly scope for the careful addition of some of Leopold Mozart s Ribattuta, Zur ckschlag, Circles and Groppos to vary each repetition. As I have discussed in previous chapters, it is unlikely that the same melody written out twice means that it is to be played exactly the same.

Cadenzas and Eing nge present a variety of options for today s performer, as they would have done in Mozart`s time. Mozart clearly marks with fermata signs where the soloist is to present one or the other, and the custom in the day would have been to improvise them, or at least to give the impression of improvisation even if the soloist had prepared something prior. A cadenza is an opportunity for the soloist to reflect on the music they have presented. It seems a shame for today s soloist to revisit cadenzas by Joachim, Ysa e or Franko, which tend to be far to long and more concerned with the demonstration of technique than interaction with the music, when the cadenza gives them a rare opportunity to comment on a piece mid performance. In my opinion, there are two directions in which to take the cadenza. Either to play (preferably improvising) a cadenza in an 18th century style, or to play something unashamedly modern, playful and unexpected, a thoroughgoing commentary by one composer on the music of another, in the manner of Stockhausen`s cadenzas to Mozart`s clarinet and flute concertos or Schnittke`s cadenzas to the Beethoven violin concerto. I would argue that these cadenzas are actually far more in keeping with the 18th century tradition than any of the standard ones by the great virtuosi. Cadenzas give performers the opportunity to surprise and delight, even shock an audience. By playing something well trodden and familiar, that opportunity is wasted.

Mozart`s violin concertos are staples of the violin repertoire. They are considered vehicles for violinists to show of their tone, good taste and musical expression. Even violinists who do not explicitly subscribe to the `historically informed` performance movement (or in the case of some like Pinchas Zukerman, actively buck the trend) give performances where the key goal is faithfulness to the text, delicacy and subtlety of expression. Faced with this backdrop, it would take a courageous violinist indeed to present the Mozart concertos with improvised ornaments, especially ones as complex as evidenced by the Ployer, Von Trattner and Stadler ornaments for the piano repertoire. Improvisation of this kind is now a very rare skill, as it requires a substantial knowledge of harmony and musicology that in my experience is rare even among today s composers, let alone violinists. However, nothing ventured, nothing gained the next stage for this research is to move it from paper to practice and from the eyes to the ears. Careful preparation, rather than improvisation, would be the first step, but with diligence and practice I am confident that this is a skill that could be reclaimed, given time.