University of Leeds



W. A. Mozart: Violin and Viola Duos — the 'Uppingham' collection - Part C

TITLE PAGE, Abstract, Acknowledgements, Citation 
Part A: INTRODUCTION (David Milsom & Clive Brown)
Part D: EXPERIMENTAL RECORDINGS (Clive Brown – violin; David Milsom – viola) 
Part E: CONCLUSIONS (David Milsom) 



C1. A Synopsis of Ferdinand David's Performing Practices (Clive Brown)
C2. Ferdinand David’s Hand-Written Annotations in the ‘Uppingham’ Copies (David Milsom & Clive Brown)
C3. Recording from the ‘Uppingham’ Collection: Practical Considerations, Problems, Difficulties & Decisions – Personal Perspectives of the Performers (David Milsom & Clive Brown)


C1. A Synopsis of Ferdinand David's Performing Practices (Clive Brown)

The implications of Ferdinand David’s printed and manuscript performance instructions in his edition of Mozart’s Duos for Violin and Viola can only be understood in the broader context of his style and performing practices as a whole. His training and early experiences already provide valuable insights into the forces that shaped him as a musician. These are also discussed in my article ‘Ferdinand David as Editor, but it will be convenient to repeat some of that material here.

After demonstrating his exceptional musical abilities as a child, the thirteen-year-old David was sent to Kassel where he spent the years 1823 to 1825 studying violin with Spohr and theory with Mortiz Hauptmann. Spohr’s tuition encompassed not only solo violin technique but also ensemble and orchestral practice; he encouraged his students to take part in chamber music and made them play as members of the theatre orchestra, which he conducted. It seems probable that many of David's stylistic precepts were forged at that time, for Spohr had very decided views about the style of playing that was appropriate to particular repertoires and instilled these into his students. As early as 1805 the influential editor of the Leipzig Allgemeine musikalische Zeitung, Friedrich Rochlitz, had recognised Spohr’s sensitivity in giving different composers’ works distinctive character, and considered him a great artist due to ‘his insight into the spirit of the most different compositions, and his skill in reproducing each in that spirit’ (seine Einsicht in den Geist der verschiedensten Kompositionen, und sein Kunst, jede in diesem ihrem Geiste darzustellen).

Furthermore, Rochlitz enthused: 

Er ist fast ganz ein Anderer, wenn er z. B. Beethoven, (seinen Liebling, den er trefflich behandelt,) oder Mozart, (sein Ideal,) oder Rode, (dessen Grandiose er sehr gut anzunehmen weiss, ohne mit ihm an das Scharfe und Schneidende zu streifen, und ihm nur Weniges, besonders in Dicke des Tons, zuvorlassend,) oder wenn er Viotti und galante Komponisten, vorträgt; er ist ein Anderer wie sie Andere sind.
[He is almost a different person when he performs, for example, Beethoven (his darling, whom he handles splendidly), or Mozart (his ideal), or Rode (whose grandiosity he knows so well how to assume without, like him, occasionally letting himself verge on scratching and scraping, particularly in producing a big sound), or when he plays Viotti and gallant composers; he is a different person, because they are different people.]
[Allgemeine musikalische Zeitung, vol. 7, 1804-5: cols 202-3]

Spohr’s seriousness of purpose clearly had a decisive impact on David’s future career, and when he left Kassel at the age of fifteen he seems already to have been a rounded artist with a well-developed technique capable of tackling his master’s most difficult works. During the years 1825 and 1826 he and his sister Louise (later, as Louise Dulcken, to become one of the most successful pianists in London) made a series of concert tours to Copenhagen and a number of important north German cities. At the Leipzig Gewandhaus in December 1825, and in Berlin the following March, David played Spohr’s 'Gesangsscene' Violin Concerto no. 8 (of which he was later to make an edition), while his sister performed works by Moscheles. A reviewer commented on David's performance that he played ‘quite accurately, with full and beautiful tone, and irreproachable bowing’ (mit ziemlicher Fertigkeit, einem vollen und schönen Ton und untadelhaftem Striche). During his stay in Berlin David first encountered Felix Mendelssohn and laid the foundations of the friendship that was later to have such a profound influence on his career. Their friendship deepened after David moved to Berlin to take up a position in the orchestra of the Königstadt Theater and became a regular participant in chamber music at the Mendelssohn house, especially playing string quartets with Mendelssohn (viola) and the brothers Edouard and Julius Rietz (violin and cello). For six years, from 1829 to 1835, David occupied an unusual post as leader of a private quartet for Carl von Liphart at Dorpat (now Tartu) in Estonia, where he had an exceptional opportunity to familiarise himself with the whole repertory of string quartets that were performed at that time. 

Although the basis of David’s violin playing was formed predominantly through his study with Spohr, he was also alert and responsive to the new and different approaches to violin technique that he encountered in later life. In March and April 1829, shortly before David moved to Dorpat, he had the opportunity to hear Paganini in Berlin and may even have met the Italian virtuoso when Paganini dined with the Mendelssohn family on 19 March 1829. According to David’s pupil August Wilhelmj, the experience of hearing Paganini was a revelation, which, although causing a momentary crisis of confidence, was to be a productive influence on David’s further development. Furthermore, while in Dorpat, David was able to make occasional concert tours to St Petersburg and other nearby musical centres, and although during his years in Leipzig his only major concert tours were to London in 1839 and 1841, he travelled widely in German-speaking lands and had excellent opportunities to hear and associate with the many celebrated European violinists who included Leipzig in their touring itinerary. As a player and teacher, therefore, he absorbed these experiences and allowed them to mould his own playing style and pedagogy. An account written shortly after David’s death explained it thus: 

Nachdem er den Erstgenanten [Paganini] zum ersten Male gehört, zwar hatte er – so erzählt Wilhelmj – das Violinspiel ganz aufgeben wollen. Doch ward er zum Heile der Kunst dieser Entschluss nicht zur That; denn geradezu epochmachend wurde sein Wirken für die Geschichte des Violinspiels dadurch, dass er das breite, sogennante deutsche Spiel Spohr's weiter ausbildend, die Errungenschaften eines Paganini und der französisch-belgischen Schule mit dem classischen alten Geigenspiele in Einklang zu bringen und eine Verschmelzung derselben anzubahnen suchte. Solchergestalt ward er zum Reformator, ja, wie Wilhelmj ihn bezeichnet, zum “Vater der modernen deutschen Geigerschule.
[After hearing [Paganini] for the first time, he actually – according to Wilhelmj – wanted to give up violin playing entirely. But luckily for art he did not carry out this intention; for his work was, in fact, epoch-making for the history of violin playing, because, while cultivating the broad, so-called German playing style of Spohr, he sought to unify and amalgamate it with the acquirements of Paganini and the Franco-Belgian School. Thus he became the reformer, indeed according to Wilhelmj, the “Father of the modern German school of violin playing”.]
[La Mara, 1878: 61]  

Comparison of Spohr’s 1833 Violinschule with David’s Violinschule of 1863 demonstrates clearly which aspects of the techniques of Paganini and the Franco-Belgian School were incorporated into David’s playing and teaching. The most important differences occur with regard to springing bowstrokes, the use of artificial harmonics, and the employment of left-hand pizzicato; the two latter practices, however, seem not to have been wholeheartedly embraced by David, for in a letter to Mendelssohn from London in 1839, discussing the violin playing he had heard there, he remarked ‘To my great joy harmonics and pizzicato are going entirely out of fashion, at last the most stupid person now knows that it is charlatanry and rejoices that he has noticed it; on the other hand many people will have to go without their chief effects’ (Zu meiner großen Wonne kömmt das Flageolet und Pizzicato hier ganz aus der Mode, der Dümmste weiß jetzt endlich, das es Charlatanerie ist und freut sich, daß er es merkt; da werden denn allerdings manche Leute ihre Haupteffecte einbüßen) [Eckardt, 1888: 110]. Presumably he included artificial harmonics and left-hand pizzicato in his 1863 Violinschule because they could hardly be ignored in an up-to-date method. In other important respects Spohr’s and David’s approaches remained essentially the same. Both warn against the overuse of vibrato and both give similar instructions for the execution of portamento.


In respect of bowing, the practical parts of the two Schools begin similarly, with sustained bowing on open strings. Stopped notes are then gradually introduced, but David diverges from Spohr’s method by introducing pairs of separated notes in a single bow (remaining on the string with a short break between them) much earlier. Also at an early stage he includes a page of specific bowing exercises aimed at familiarising the student with the employment of different parts and lengths of the bow, whereas Spohr gradually integrates these elements of technique into his much more extensive practice pieces, using detailed instructions for the precise parts and lengths of the bow that are to be used for them. Nevertheless, both authors confine themselves at this stage to on-string bowing, with a distinction between slurred notes and detached notes that are either played with well-connected or sharply separated short bowstrokes. One interesting difference between their approaches is that David introduces the concept of dynamic variation on single notes, as well as a wider range of dynamics in general, much earlier than Spohr. But this is a difference in method rather than practice, although David’s early exercises create an impression, perhaps illusory, of greater concern with nuances of bowstroke than Spohr’s, which seem more concerned to develop firmness and roundness of tone. The first distinct difference seems to emerge in David’s exercise no. 50, a Presto in 6/8 with many separate staccato 1/8-notes that are marked to be played in the middle of the bow. In Spohr’s Violinschule all faster detached notes up to this point (before the introduction of position changing) have been played in the upper half of the bow with as much movement of the forearm as can be accomplished in the time without moving the upper arm. Both authors familiarise the pupil with playing in the highest positions, with passages in octaves and tenths and instructions for the execution of portamento, before returning to the subject of bowing, at which stage an important difference emerges.

Spohr’s advanced instruction in bowing begins with the nuanced treatment of long bows (which David introduces much earlier). He then provides a series of exercises in 1/8-notes at a tempo of 1/2-note=104, beginning with the détaché, which he marks with staccato strokes. This is essentially the detached bowing that had been used in the earlier exercises, and he instructs that it is to be made ‘with a steady upper arm and as long strokes as possible, in the upper part of the bow’ (bey unbeweglichem Hinterarm mit möglichst langen Strichen am obern Theile des Bogens), and that ‘the notes must be perfectly equal in power and duration, and follow one another in such a manner that, in changing the bow, no break may be observed.’ (Die Töne müssen in der Dauer und in der Stärke völlig gleich seyn und sich so aneinander reihen, dass beym Wechseln des Bogens keine Lücke bemerkbar wird) [Spohr, 1833: 130]. Spohr also observes that this stroke is intended in a succession of separate notes when no staccato mark is employed. He then continues with an extensive catalogue of bowing patterns with different examples of slurred and separate bows, after which he introduces martelé, fouetté, and slurred staccato.

David, on the other hand, first introduces a bowing, designated by staccato strokes, which he instructs should be performed with a firm stroke at the point (Fester Strich an der Spitze) and another marked with tenuto lines, which he describes as performed with a ‘lying stroke’ (liegender Strich), but without specifying where in the bow this should be executed [David, 1863: ii, 37]. These are evidently the equivalent of Spohr’s martelé and détaché, although David never uses these terms. David’s following exercises, showing different patterns of slurred and separate notes, indicate that the bowstrokes marked with staccato strokes could also be executed in the middle of the bow, and the ones with lines could also be played in the middle or at the heel. These uses are paralleled in Spohr’s exercises, but rather infrequently. Again without using Spohr’s terminology, David describes what Spohr calls fouetté, for which he introduces a special sign. In the next exercises, however, he departs entirely from Spohr’s practice with what he describes as ‘hopping and springing bowstrokes’ (hüpfenden und springenden Stricharten), which, to distinguish them from the earlier on-string bowstrokes, he marks with staccato dots rather than strokes [David, 1863: ii, 38]. Finally he turns to the slurred, on-string staccato, which he marks with staccato strokes under slurs; this is predominantly indicated to be played up-bow starting at the point of the bow, but like Spohr he also illustrates a down-bow staccato [David, 1863: ii, 39-40].

It seems clear, therefore, that although David employed all types of bowstroke described in Spohr’s Violinschule (though perhaps not always in quite the same manner), he also included a range of springing bowstrokes of a kind condemned by Spohr as ‘unworthy’ of serious violin playing. What David’s Violinschule fails to convey, however, is where and when he would have envisaged these different types of bowstroke being employed in the musical repertoire of his time. Fortunately, his many published editions and, more particularly, his detailed manuscript annotations in the copies from which he performed help us to fill this gap to a considerable extent as well as providing us with a rich mine of information about his left-hand technique and the expressive effects he envisaged with his fingerings.

It seems likely that David’s early training with Spohr (who included bowing instructions and detailed fingering in the published editions of his own music) may have encouraged David’s habit of writing detailed performance markings into the copies from which he played. He certainly adopted it at an early stage. In a letter to Mendelssohn in 1844 he commented:

Weissenbornchen spielt Quartett, dass es eine Freude ist; er spielt aus den Grabauschen Büchern, die ich im ersten Winter hier genau bezeichnet habe, und macht mir jedes kleine Mätzchen genau aufs Haar nach.
[It is a joy to hear young Weissenborn play quartets; he plays from Grabau’s copies, which I precisely marked up in my first winter here [1836] and he imitates all my little traits with hair’s-breadth precision.]
[Eckardt, 1888: 204]

By this point David had also made his first published edition – of Bach’s Solo Sonatas and Partitas for violin. Over the next thirty years he was to publish bowed and fingered editions of the majority of major classical chamber-music compositions involving the violin by Haydn, Mozart, Beethoven and Schubert, as well as many Baroque and a few contemporary works.

In the published editions much can be deduced about the part of the bow employed, especially where a bow direction is marked on a long note after a passage of short ones; and in his published annotated editions of classical chamber music, where only the bow direction is specified, we see an overwhelming preference for using the upper half or upper middle of the bow for a succession of fast or relatively fast notes.

One striking feature in David’s extant personal performing copies is his inclusion of manuscript instructions for bowing styles and sometimes of the part of the bow he intended to employ. These confirm the impression gained from the printed markings in his published editions. He used the abbreviation spgd (springend/springing) not only in his Violinschule [David, 1863: ii, 38], but also in his personal performing copies of music. The instructions spgd or sometimes saltato (which appear from context to be synonymous and to precede his use of spgd) occur in his extant personal copies only in Baroque and contemporary pieces. He used spgd quite often in his Hohe Schule des Violinspiels; for instance in Biber’s C minor Sonata. Both spgd and saltato are found in his copy of Cherubini’s Quartet No. 1 where they alternate with tenuto lines, and saltato occurs in Volkmann’s String Quartet op. 14. Significantly, however, there is no extant instance of either of these marking in David’s published editions or personal copies of any works by Haydn, Mozart, Beethoven, Schubert, or Mendelssohn.

Sometimes David marked Fr (Frosch) or talon/tal to indicate the heel of the bow, where off the string, or short percussive stokes seem to be envisaged, for instance in Biber’s Sonata, and Volkmann’s String Quartet op. 9. On the other hand, these instructions also occur where a broad stroke near the heel is apparently envisaged; for instance, Fr in the Biber Sonata (score), or talon in the last movement of Mendelssohn’s String Quartet op. 44 no. 2 (the only occurrence of either of these terms in his copies of Mendelssohn’s chamber music).

Curiously, in his extant personal copies, David never employed the abbreviation hpfd (hüpfend, literally ‘hopping’), which he used in his Violinschule to designate a rebounding stroke in which ‘the bow must never entirely leave the string’ (Der Bogen darf die Saite nicht ganz verlassen) [David, 1863: ii, 38]. This is essentially the same as the bowstroke designated sautillé in French usage. There are, nevertheless, a few places where such a stroke seems to be implied by David’s markings. In the Scherzo of Mendelssohn’s String Quartet op. 44 No. 2 he marked M for ‘Mitte’, indicating the middle of the bow, where a bowstoke of this kind seems appropriate (see score); elsewhere he marked legg[iero], for instance in the Scherzo of Cherubini’s String Quartet No. 1, probably with a similar meaning. It is likely, too, that he frequently interpreted printed directions for leggiero/légèrement in the same light, for instance in the Moderato sans Lenteur section of the Scherzo of Cherubini’s String Quartet No. 3, or the last movement of Mendelssohn’s String Quartet op. 44 no. 2 (score).

Hermann Schröder, in Die Kunst des Violinspiels, commented that when a sautillé bowstroke, which he calls der leichte Bogen (the light bow), was envisaged ‘An indication for this bowstoke could be designated by the term leggiero’ (Eine Bezeichnung dieser Strichart würde durch den Ausdruck leggiero bemerkt werden können) [Schröder, 1887: 72]. Spohr, who objected to the use of such bowstrokes in general, admitted their appropriateness only in a few scherzos by Beethoven, Onslow and Mendelssohn [Malibran, 1860: 207].

Another bowstroke described and employed by David, which is rarely encountered in modern performance, was the one designated fouetté in Spohr’s Violinschule. David instructed in his Violinschule [David, 1863: ii, 37] that it should be performed by hitting the string with the point of the bow, and he devised a special sign for it. He used this sign occasionally in printed music. In his Dur und Moll, op. 39 it occurs quite frequently, and he even provides a piece (Gebrochene Akkorde) in which it is a prominent feature. In his personal copy of Die Hohe Schule des Violinspiels it occurs not only as a printed instruction but also as an autograph addition. In a piece by Leclair the printed sign occurs in the Gavotte (stave 8, bar 1), and several times in the following Allegro (staves 3 and 4). In a Sonata, wrongly attributed to Geminiani, he adds the mark on page 4 (staves 4, 5 and 10), and also on the following page. Whether he would have applied this technique in Mozart remains uncertain, but it seems unlikely.

Left-hand techniques

David’s left-hand technique appears to have been essentially the same as that employed both by his teacher Spohr and his pupil Joachim. All three masters were of a like mind about the use of vibrato and portamento; both were seen as essential expressive resources, but neither was to be used too often.


Spohr and David employed a wavy line to indicate notes on which vibrato might appropriately be employed, not only in their violin methods, but also occasionally in their published music. Spohr began to include it in works from the 1850s (beginning with the Duo for two Violins, op. 148); David employed it mostly in works with a didactic intention, such as Dur und Moll and Die Hohe Schule des Violinspiels, but he also included autograph vibrato signs in some of his personal copies, for instance of Beethoven’s Romance op. 40 and Biber’s Sonata.

Spohr instructed that vibrato could be executed according to context: slow or fast, as well as with increasing or decreasing speed of oscillation, the latter being associated with crescendo and diminuendo respectively. David associated the speed of vibrato directly with dynamics in his Violinschule [David, 1863: ii, 43]; the louder the note, the faster the vibrato. Joachim’s rather cursory treatment of vibrato in 1905 [Joachim & Moser, 1902-5: ii, 96-96a] consisted mainly of direct quotation from Spohr’s 1833 Violinschule.

In David’s editions, as in Spohr’s markings in his own music, natural harmonics and open strings are frequently employed on important melodic notes making it absolutely clear that a vibrato was not envisaged in these and equivalent places (except perhaps very occasionally, using the technique described by David in the footnote to his vibrato instructions in the Violinschule for producing sympathetic vibrations on an open string). This tallies with the warnings by Spohr, David, and Joachim, that, as David put it, vibrato must not be used ‘too frequently nor without insufficient reason’. Another feature of David’s instructions for the execution of vibrato provides persuasive confirmation that he expected it to be employed only occasionally, for he observed that in performing it ‘the 1st finger must leave its usual place on the neck of the violin [my italics], which must only be held with the thumb and the point of the finger which is touching the string’ [David: 1863: ii, 43 ]. In modern technique the first finger is not expected to touch the neck as a rule, in order to facilitate continuous vibrato. Carl Flesch also advised that the habit of keeping fingers down on the string when it was unnecessary to lift them, which was a standard part of 19th-century practice, should be abandoned, since ‘the unnecessary resting of the unoccupied fingers on the strings lames the vibratory capacity of the entire hand in slow movements in a manner injurious to vitality of expression’ [Flesch, 1923i, 126].

Of course, we cannot assume that the lack of an ornamental vibrato, produced in the manner described by Spohr, David, and Joachim, always meant an absolutely still hand. Unless the hand position is maintained very rigidly, a slight degree of tremor, which would not be regarded as true vibrato by modern players, may sometimes have occurred, producing minimal oscillation with almost undetectable variation of pitch, where no vibrato was consciously intended. This type of ‘ghost’ vibrato can be detected in Joachim’s 1903 recording of his Romance in C major. Currently unpublished research using spectrogram analysis of the Romance has shown that whereas there are fifty-three notes longer than an 1/8-note on which some degree of vibrato could be discerned aurally by the skilled researcher [Robin Wilson], the spectrogram demonstrated a slight degree of oscillation on seventy-two such notes. From a personal perspective, I have noticed when listening back to a recording of my own playing that I sometimes perceive a hint of vibrato on notes where I was sure I had maintained a completely still left hand.


With regard to position-changing and the concomitant production of an audible slide (portamento, glissando, Durchziehen der Töne), Spohr’s, David’s and Joachim’s treatises are again fundamentally in agreement; as with vibrato, Joachim cites Spohr directly. David, however, provides a more detailed treatment of the subject than Spohr, with an extensive series of exercises taking the pupil through a series of intervals from unison to tenths in which the destination of the ‘leading’ finger (the one stopping the note immediately before the change of hand position) is carefully indicated [David, 1863: ii, 33]; he instructs, however, that ‘the small notes indicate approximately the place the gliding finger has to reach; take great care not to sound those notes.’ Spohr had explained that on reaching the new position in which another finger was to stop the target note, that finger was to go down so quickly that the ear would be ‘cheated into the belief that the gliding finger has actually passed over the whole space from the lowest to the highest note’ [Spohr (trans. Bishop), 1843: 108]. There is no doubt that such position changes, executed in a slur, both up and down, were expected to be heard.

David provides rather more information about the mechanism of position changing than Spohr, with his instruction that ‘in moving from a higher position to a lower one, the thumb must glide down a little beforehand’ [David, 1863: ii, 33], and he provides an exercise for ‘the flexibility of the thumb’ in position changing, stating that the player must ‘hold the violin in the following exercise so as not to let the chin touch it. The thumb alone has to support the instrument and must in going down always glide a little beforehand into the next position.’ [David, 1863: ii, 34]. This action of the thumb is also addressed by Joachim, whose Violinschule provides a photograph of the position of the thumb just before descending from first to third position.

Although not directly addressed by Spohr or David, it is clear too that portamento could also be executed between bowstrokes, by changing the bow smoothly and rapidly immediately after the hand had arrived in its new position. An example of this is given by Joachim [Joachim & Moser, 1902-5: ii, 94].  

The speed and intensity of the portamento, however, was a matter of taste, which was not easily susceptible of being represented in notation, although David’s contemporary, Charles de Bériot, attempted to do so in his Méthode de Violon of 1858 (p. 219) [see also Brown, 2009].

David’s editions, and especially his autograph annotations in personal copies, provide rich evidence of the importance he attached to portamento and the care he took to find the most appropriate places for expressive position changing. His autograph annotations reveal some unusual practices. At one point in Biber’s Sonata in Die hohe Schule des Violinspiels he writes the instruction ‘Rutschen’ (slide), evidently referring to pairs of notes marked with tenuto lines and a slur, where portamento might not have been expected (see score). In a Leclair Sonata in G major he indicates portamento from an open string, where it might otherwise not have been obvious, by means of a slanting line over the fingering (see score).

Further discussion of the implications of David’s fingering and bowing instructions in the violin and viola parts of Mozart’s Duos can be found below in section C3, where the rationale behind our attempts to realise their aural implications in the recordings is articulated.

C2. Ferdinand David’s Hand-Written Annotations in the ‘Uppingham’ Copies (David Milsom & Clive Brown)

The annotated copies in this collection contain, as already observed, a number of hand-written markings, the authorship of which cannot always be identified with certainty. This discussion will confine itself to examination of the blue crayon, pencil, and ink additions that can confidently be attributed to Ferdinand David. His annotations can be classified as follows:· Bowings that re-enforce the edition’s printed markings

· Bowings that adapt/change the edition's printed markings

· Fingerings that re-enforce the edition’s printed markings

· Fingerings that adapt/change the edition’s printed markings

· Dynamic markings that change the edition’s printed markings

· Dynamic markings that augment the edition’s printed markings

· Annotations for procedural convenience (to facilitate page turns, etc)

· Annotations that constitute a significant addition to or change to Mozart’s original text (based on the first edition).

· Additions that alter or supplement the printed text

The expectations that lie behind printed and manuscript performance annotations may not always be easy to determine. Some have a clear purpose, such as fingerings that indicate necessary position changes for technical reasons, or bowings that direct the player to use a particular part of the bow or succession of bow-strokes; but there is seldom a single solution to technical challenges and the motivation for one solution rather than another is not always clear. In many cases, however, these directions can provide important clues to the aesthetic expectations of individual editors. This is examined more extensively in the CHASE articles, 'Bowing & Fingering Instructions in String Music During the 18th and early-19th Centuries', and 'The Evolution of Annotated String Editions'. Understanding the nature and purpose of the 'editorial hand', therefore, becomes an important factor in making sense of such editions and relating them not only to other editions by named editors, but also to developed scholarship into performance style and practices gained from other sources of information such as treatises and early recordings. For example, one might contrast Joachim's editions of Mendelssohn's chamber music (see String Quintet no. 2 and Piano Trio no. 2), which are extraordinarily sparse and sporadic in their quantity of editorial information, with the detailed provision of printed bowings and fingerings in the music of Louis Spohr, which represent not merely suggestions but essential prescriptions for practice, the absence of which (as we find in many modern commercial performances) can in many ways misrepresent his musical language. [See, for example, Brown, 2010b: 477-478.]

Spohr’s attitude to the performance of his own music is well attested; he was critical of musicians who employed practices that the composers never intended. According to his pupil Alexandre Malibran, he complained bitterly about the use of springing bowing in his music where it was never intended, and Malibran also reported that:

He was absolutely adamant that one should not play all composers in the same way; on the contrary, he wished the artist to adhere to the true tradition; so to say, to deny himself, and reproduce the composition just as it is. ‘But they,’ he exclaimed ‘care neither about the style of the man nor about the instrument, which in the time of the composer was an entirely different one than now; they depict Frederick the Great with a haircut à la Titus, in a black coat and trousers!' [Malibran, 1860: 208; quoted in Brown, 1999: 273].

On the other hand, Joachim's famed spontaneity in performance, which made him reluctant to prescribe a single way of playing a piece, is equally well documented. Such matters cannot so easily be verified for many other figures discussed on the CHASE website as a whole. David's editorial approach in comparison with the early published editors examined here, and indeed later representative editions of the 'classical' German school, appears to be quite intensive, presumably to convey his personal conception and perhaps also what he regarded as the 'classical' tradition.

With hand-written markings such as these, it is important to understand as much of possible of their context and purpose. As suggested earlier, it seems probable that these copies were used when David performed the Duos in Leipzig. The most persuasive evidence of this is his autograph of the pizzicato version of the viola part for the Andante of K. 424 stuck over Mozart’s original, and the similarly pasted-in violin cadenza on top of David’s pencil sketch; these suggest that the copies were used in the 1870 Gewandhaus performance.

<p style="Times New Roman" ,"serif""="">The lack of consistency in the extent of markings across the two instruments is noteworthy. The violin part in the set containing markings ('Set 1': items 1 and 2 described in section A3) is extensively annotated throughout, but the viola part much less so, especially K. 423. If these markings were made by David for another player, the matter seems straightforward; from the evidence of the Cherubini E-flat quartet parts (performed by the present-day Ferdinand David Quartet from these copies on 25 Sept 2013) it seems that David marked up other parts with essential information as if co-ordinating bowings within sections of an orchestra. This makes the connection of these copies to the 1868 and 1870 Leipzig performances reviewed in the Signale für die musikalische Welt (see section A3)  rather puzzling, because it is known that David was the viola player on both occasions. The extensively marked violin part gives every impression that David played from that part on other occasions. That David marked the viola part with only a few items of information (mainly to align with equivalent markings in the violin parts) is perhaps less surprising than the extent to which he annotated the violin part, which, in numerous places, include crayon reinforcement of printed markings. The violin part is typical of David’s thoroughness (exemplified by the Cherubini quartet), not leaving such matters to chance and, in a way with which many living players might identify, ensuring that the part provide all the necessary information for a secure and confident performance. This being the case, why would David put far fewer markings into the viola part from which he himself intended to play? 

C2. a) Crayon, Pencil & Ink Markings in the Violin Part (Clive Brown)

NOTE: It is intended that the reader will refer, for illustrations, to the downloadable pdf copies of Ferdinand David's performing parts available through the CHASE website [view]. Hyperlinks within the text lead to an online view that includes, pop-up annotations via clickable boxes. References to these copies are given in the form [p.1, s.1, b.1] (i.e. page 1, stave 1, bar 1).

[Some of the following text may also be found in the clickable red boxes on the scans of David's violin part.]

G Major Duo, K. 423 (Set 1 violin part)

I: Allegro

The first of David's blue crayon markings is merely a reinforcement of a slur in bar 4 [p.2, s.1, b.4]. David's habit of emphasising such printed markings may indicate that his eyesight deteriorated in later years. It is equally possible, however, that his apparent dependence on clearly marked performance instructions reflects the immensely busy life of a musician who was constantly performing an enormous range of music with relatively limited rehearsal time.

The crayon fingerings at bars 4-6 [p.2, s.1, bb.4-6] provide a good example of David’s supplementary markings. We see that he changed his mind twice. The printed copy suggests performance in first position, but apparently for the sake of string colour and portamento David changed it to third position from the upbeat to bar 5 for two notes then returned to first position; subsequently he rubbed out the 1st finger on 5ii, which can still be faintly seen, and indicated a harmonic, evidently to be taken with a 4th-finger extension, with an 'o' on 5iii, returning to 1st position with a 2-2 slide followed by an open A string in the next bar. The same changes (including the deleted '1') are made when the passage returns in the recapitulation. The purity of the harmonic and the portamento give a distinct character to this figure, which is well illustrated in the recording, throughout which I consistently observed David's final version.

At bar 7 [p.2, s.2, b.2] David's blue crayon substitutes performance on the D string for a simple execution of the figure in first position. This produces an audible shift back to first position on the final note. Two bars later the unusual printed fingering also serves a clear expressive purpose.

The <> marking in bar 15 [p.2, s.3, b.4] has a variety of implications in 19th-century notation. Here it invites a delicate but noticeable vibrato.

The harmonic in bar 38 [p.2, s.8, b.1] is typical of David and other 19th-century violinists, but the use of such fingerings gradually disappeared with the reform of fingering practices pioneered by Carl Flesch and others in the 20th century, which aimed to avoid what were deemed unnecessary slides and to enable continuous vibrato.

In bars 40-41 [p.2, s.8, bb.3-4] and the equivalent passage later in the movement. David changed the printed dynamic scheme to produce an echo effect.

Other markings in this movement are largely technical although a few, such as in bar 78 [p.3, s.5, b.1] have implications for tone colour.

II: Adagio

The portamento to a harmonic with the fourth finger, which has a very distinctive effect, followed by the replacement of the same finger a tone lower is typical of David's and other 19th-century violinists' practice. Equally typical is the slide with the 4th finger at bar 6 [p.4, s.7, b.6vii-viii], on which a portamento of the kind produced on our recording is clearly envisaged.

The blue crayon change of dynamics at bars 18-20 [p.4, s.10, bb.1-3] gives an entirely different, more subtle, character to the passage.

A particularly interesting change of colour and expression occurs in bars 30-35 [p.4, bottom stave] with a 4-4 portamento from first to third position on the A string at 30iii-iv, followed by a shift to 5th position on the D string at 30ix, leading to a passage of three bars on the G string. This is combined with the breaking up of long bows into smaller units, including a twofold change of bow on the sustained g' in bb.31-2 (evidently to be made as connected as possible, which is succesfully achieved on the recording), and then in bb.33-34, after an initial portamento to 4th finger, descent from 5th to 2nd position by semitone shifts, which is typical of David's practice.

At bar 35 [p.5, s.1, b.1] David's printed fingering involves a shift from the g' in third position to the g'' in fifth position, allowing the g''' after the rest in the following bar to be located securely. This shift, however, may have felt rather ungainly at the conclusion of the broken octave scale. David's crossing out of the printed fingering obviates the awkward shift in b.35, but demands greater accuracy in placing the g'''.

David's changes on the last two staves of the movement retain the idea of dynamic contrast on the ornamented repetition of the previous four bar phrase, but produce a very different musical effect.

III: Rondo – Allegro

[It should be noted that pencil markings in this movement come from quite recent pupils at Uppingham School.]

In the principal theme David's printed version shortens Mozart's 1/4-notes in the first and third bars to 1/8-notes, which are played portato in up-bows. This was undoubtedly to ensure a detached performance of a kind that, if he had retained Mozart's notation, would not naturally have been elicited from a violinist trained in David's tradition, who would have been accustomed to a more legato style of performance. It is plausible that he inherited this understanding of articulation in Mozart performance from his teacher, Spohr.

The use of the 4th finger for successive notes in a rapid tempo at bars 30 and 32 [p.5, s.8, b.1 & b.3] is typical of 19th-century practice. This was one of the practices, condemned by Flesch, that gradually became obsolete in the early 20th century.

The change from p to f at bar 35 [p.5, s.8, b.6] emphasises the contrast with the following, quieter repetition of the figure. This dynamic was not added in the viola part, nor did David make the change on the next appearance of this passage, but in the recording we decided to implement it both times.

The slurred staccati at bar 49 [p.5, s.11, b.1] and elsewhere in this figure are clearly intended to be played with a well separated on-string bowstroke towards the point of the bow, which is how we executed it.

David's change to the dynamics at bar 52-3 [p.5, s.11, bb.4-6], which are repeated when the passage returns later in the movement, creates an entirely different character to the music at this point.

At bars 155-80 [p.7, s.4, bb.1-4] David seems to have been undecided about the bowing but eventually decided to make it the reverse of what is marked the first time this passage occurs (p.5). From a practical point of view, this facilitates the string crossing.

The use of a succession of conjunct notes played with the same finger at bars 181-2 [p.7, s.9, bb.2-3] is very typical for David. It is noteworthy that in this case he originally intended it to continue for two more repetitions of the figure, arriving at the new position required for the following bar, but that he later changed his mind and remained in fifth position until the return of the theme. It is highly probable that, as discussed in section C3a, he executed this in a similar manner to Karl Klingler in the latter's 1912 recording of the Meunetto and Trio of Mozart's String Quartet K. 421 [Klingler Quartet, 1912].

Bb-Major Duo, K. 424 (Set 1 violin part)

I: Adagio-Allegro

The division of the printed bowing into shorter units at bars 3 and 5 [p.8, s.1, b.3; s.2, b.2] is quite often encountered in David’s personal copies. This may be associated with his preference for the full and powerful tone that was noted in his playing by a number of contemporaries.

The addition of '3-2' fingering at bar 5 [p.8, s.2, b.2] may suggest David’s intention to execute a particularly prominent portamento here. Since he ended in second position in the previous bar, he could quite easily have continued with a 4th finger on the first note of the next bar.

At the second note of bar 8 [p.8, s.3, b.2] David adds a g to provide greater fullness to the harmony. In the viola part there are many more instances of his adding double stops to Mozart’s original (see David Milsom's discussion in section C2b, below).

At a later stage than his blue crayon markings, David added a number of annotations in pencil in this Duo. At bar 20 [p.8, s.5, b.7] he anticipates the p that is printed in the next bar.

At bar 50 [p.8, final bar] he reinforces the printed fingerings in blue crayon. This fingering scheme, with its rapid shift from third to first position – typical of David’s tendency to avoid second position – seems less logical than executing the whole bar in second position (which is marked by Herrmann, Schulz and Laforge); however, both Rauch and Schnirlin also give David’s fingering.

The execution of the repeated c'''s at bar 52 [p.9, s.1, bb.1-2] with a slurred staccato, evidently to be executed close to the point of the bow, is characteristic of David. It produces a particularly crisp effect.

At bars 55-6 [p.9, s.1, bb.5-6] there is one of David’s typical passages played with a single finger, followed immediately by a 4th-finger portamento. Quite what effect the successive 1st fingers were intended to produce in this instance is unclear (but see my notes on our performance in C3a).

A particularly expressive portamento is indicated by the printed fingering, reinforced in blue crayon, at bar 91 [p.9, s.7, b.1] with a characteristic same-finger semitone shift back from fourth to third position between notes iii and iv of the bar. At bar 97 (on the same stave) David alters the ‘clean’ printed fingering to produce a same-finger portamento between the first two notes.

At the end of page 9 an intriguing issue is raised by the absence of any marking by David. A page turn occurs in the violin part after bar 123, just before the recapitulation. Without a page turner, this requires an elongated rest. In other cases where awkward page turns occur, David commonly marked either the need for a page turner with the instruction 'Umwender', or provided a notated alternative as at the end of page 7 in the viola part of this movement. Here it seems likely that he simply paused to turn the page, which is musically quite satisfactory; and we chose to do that in our recording.
At bar 165 [p.10, s.7, b.1] David removes the instruction to play on the A string, which involves a rapid descent from fifth to first position, for a simpler execution all in first position. In bar 170 (on the same stave) he ascends to third position earlier than his printed markings indicate, thus cancelling the printed portamento fingering in bar 171.
At the end of the movement [p.11] David made a number of dynamic and tempo changes, which would have entirely altered the character of the coda. For David, tranquillo seems always to have indicated a slower tempo. He later cancelled these changes but added > p just before the double bar instead of the dim. at the beginning of the penultimate bar. At bar 208 [p.11, s.2, b.6] he amends p to pp and, on the next stave, pp to ppp. The added up-bow signs in bars 217-18 [p.11, s.3, bb.6-7] allow the trills to be taken very delicately, right at the point of the bow. At a later stage still he deleted the printed < > signs in bars 208-9 and 212-13, perhaps to produce a distant, almost expressionless effect. 

II: Andante

In the first three bars [p.11, s.4] David refined the original dynamic scheme with < and > in bb.1-2 and pp in b.3. He also marked two portamentos: the first with the printed fingering in bar 1 and the second through an added fingering in bar 3; from a musical point of view it seems clear that the first portamento was expected to be more prominent than the one in b.3.

The sliding fingering at bar 11 [p.11, s.5, b.5] helps to intensify the climax of the phrase, where an ornamental vibrato is probably intended on the f''.

The dynamic markings < and > in pencil in bars 13-14 [p.11, s.6, bb.2-3] move the emphasis in the phrase from the c''' in bar 14 to the b' two notes earlier. Pencilled dynamics at bar 19 [p.11, s.7, b.3] and bar 21 [p.11, s.8, b.1] produce a substantial dynamic contrast, and the deletion of the blue crayon hairpins in bars 21-2 [p.11, s.8, bb.1-2] sustains the low dynamic level until the printed crescendo in bar 23, where division of a long slur into three enables a more powerful increase in volume.

Fingering changes in bars 26-7 [p.11, s.9, bb.2-3] are essentially technical in nature.

The pencil < in bar 28, before the fermatas, probably represents what David would have done in any case in the lead up to the expected cadenza. As at the end of the Larghetto of Beethoven’s Violin Concerto in his personal copy (British Library), David sketched his cadenza into the bottom margin of the paper. He also added several notes to the end of the cadenza in the margin at the end of bar 29 [p.11, s.9, b.5], allowing more time for an effective diminuendo to the p, changed to pp in pencil, at the beginning of the next stave. His fingering change at this point allows the more veiled tone of the A string to be maintained until the end of the second bar, rather than crossing to the brighter E string at the beginning of the printed crescendo hairpin.

David’s printed horizontal lines under the slur at bar 37 [p.11, s.10, b.7] imply a broader and weightier style of bowing than if there had been dots under the slur. At bar 38 [p.11, s.11, b.1] he changes the straightforward slurred staccato/portato to a series of cross-beat slurs in a single bowstroke. The down-bow staccato/portato marked in bars 39-40 are to be executed less sharply than might have been implied by an up-bow, as Spohr explains in his Violinschule.

III: Andante Grazioso (Theme and Variations)

David’s crayon dynamics on the first bar [p.12] elicit a more sharply characterised delivery than the printed ones. His bowing change ensures that the weight is laid on the second and fourth beats, with the notes on the metrically strong beats being delivered lightly at the point of the bow. David has also emphasised this by changing Mozart’s 1/4-notes on the strong beats to 1/8-notes.

David’s added dynamics in Variation III [p.12] mirror those in the Theme.

At the beginning of the second half of Variation V [p.13] David’s added fingering produces an expressive portamento and also retains the tone of the A string until after the beginning of the crescendo. At the end of this Variation David has added an instruction to introduce a ritard, but only on the repeat.

The bowing change at the beginning of the Allegretto [p.13] may be to elicit a lighter execution of the final 1/8-note in the figure.

At bar 202 [p.13, s.10, b.2] the c''' was perhaps originally envisaged as being taken by an extended 4th finger, but David’s added fingering makes the position-changes, with concomitant portamenti (which could hardly have been avoided in these figures in Mozart’s time), explicit.

In the last two bars of the Allegretto [p.13, s.11] David prefers security of intonation in the double-stops over retention of Mozart’s f’ to f’’ in both bars, keeping the hand in fourth position, thus avoiding the awkward position change. He makes similar alterations elsewhere, for instance removing the lower note of double-stopped octaves on pages 4 (bottom stave) and 5 (top stave) in his personal manuscript copy of Schubert’s Quartettsatz D. 703.

In the last eight bars of the piece [p.13] David evidently considered using a succession of up-bows, undoubtedly in the upper half of the bow, but later deleted his crayon alterations in the 2nd-time bar and the three following bars. He pursued the idea, however, in the next three bars, bringing the bow to the heel for the final forte chords. 

C2 b) Crayon, Pencil & Ink Markings in the Violin Part (David Milsom)

NOTE: It is intended that the reader will refer, for illustrations, to the downloadable pdf copies of Ferdinand David's performing parts available through the CHASE website [view]. Hyperlinks within the text lead to an online view that includes, pop-up annotations via clickable boxes. References to these copies are given in the form [p.1, s.1, b.1] (i.e. page 1, stave 1, bar 1).

Of the three different viola parts in the Uppingham Collection edition, the part from Set 1 [pages 15-24 of the CHASE pdfs] contains the most markings. Set 3 [pages 50-59 of the CHASE pdfs] does not contain any annotations. In the additional Set 2 viola part [pages 26-35 of the CHASE pdfs] the slow movement of K. 423 contains a few markings which replicate those in the Set 1 part; there is, however, an additional 'pizz.' instruction in Set 2 at bar 40; crescendo and diminuendo signs at 42-3 have been crossed out, and a p marking inserted at 46 which indicates a slightly different idea. The performance accompanying this article replicates the markings in Set 1, but the variability of David's approach to this section of the slow movement should be noted.

G Major Duo K. 423 (Set 1 viola part)

In the viola part to K. 423 there are no markings in the Finale (although the copy is heavily marked in a more recent hand) and all other markings are adaptations to the dynamic scheme except a pizzicato instruction added at bar 44 of the slow movement, and presumably applicable to bars 44-5 [p.4, s.9, bb.1-2]. These dynamic markings serve to intensify the expressive parameters of the dynamic scheme – such as the change from the printed p marking at bar 40 to pp, which creates an echo effect not explicitly indicated by the printed markings, but which arises as a natural and unscripted consequence of the repetition of the figure in bar 39. David’s interpolation of the arco bow scheme in the 2nd movement with pizzicato (bars 44-45, with an additional marking at bar 40 in the Set 2 part as described above) implies that he wanted to create further texture variation here – this further suggests that David’s markings were designed to transmit a more detailed (and maybe vivid) understanding of the work in performance.

B-flat Major Duo K. 434 (Set 1 viola part)

The B-flat Major Duo contains rather more information for the viola, indicated in David’s customary thick blue crayon. There are also some pencilled annotations that appear to be in David's hand, such as the NB marking at bar 40 of the 1st movement (probably to encourage the player to tarry a little for the sake of ensemble co-ordination with the violin, with its wide-spaced leaps in this passage – something we incorporate, by means of exaggerated rests, in the recorded performances in Part D below). Another pencilled change (which aligns, as will be discussed, with the kinds of notation changes David practised) is the addition of g' and e'' to form a triple-stop chord in bar 8 of the 1st movement [p.7, s.3, b.1].


In this movement bow direction indications make the score more explicit in this regard and, as in numerous instances in the violin part, these include crayon re-enforcement of printed markings, as at bar 48. The most interesting of these occurs at the end of the movement, where David indicates an up-bow for the penultimate chord rather than the more usual practice of re-taking down-bows [p.9, s.6, b.8]. This creates a probable difference between the two parts, since the violin has a triple-stopped chord here which is most likely to be executed (in modern and indeed historical practice) on a down-bow. Evidence that David would have taken this approach is suggested by the crayon up-bow in the viola part on the triple-stop chord at bar 8 (assuming that the pencilled chord is in David’s hand) which implies that such an up-bow for a triple-stop chord is the exception rather than the rule [p.7, s.3, b.1]. The opening of David's edition of Beethoven's 'Kreutzer' Sonata, op. 47, gives an example of his propensity for taking triple- or quadruple-stop chords on down-bows, with 'retake' bows included in his printed scheme, and with the up-beat chord in bar 3 marked with an up-bow as the exception.

David’s hand-written bowings sometimes indicate actual changes to the printed notation (such as mvt.I, bar 30 [p.7, s.5, b.7]), as well as a number of changes to and/or explanations of slurred staccato (bars 62-5 [p.7, s.9, bb.7-10] and the equivalent place in the recapitulation). Similar bowing traits to those observed above characterise the Finale.


Fingerings indicated in the B-flat Duo viola part are, for the most part, simply procedural, although, as one might expect in this performance tradition, they do not eschew the portamento (that is to say, they give rise naturally to the effect and do not appear to seek avoidance of it). This said, David’s decision to take bar 48 of the first movement in second position does preclude the device, which would not be the case if the other probable fingering pattern – beginning in first position and shifting with a 3rd finger on the f'' – had been indicated. As with bowings, these crayon fingerings sometimes re-enforce what is printed, as at bars 95 and 96 in the first movement where markings imply the existence of portamenti, albeit probably quite discreetly in this context. The Finale has fewer such additional or re-enforcing markings, and only three instances of fingerings added to those printed: at bars 80, 83-4, and 117. The marking in bar 80 [last line of Var. IVsimply indicates a position change (which must occur at some point in this figure). Bars 83-4 [beginning of Var. Vare more interesting: rather than remaining in first position, David keep the passage on the D-string (for tonal reasons), beginning with a characteristic harmonic and leading to an implied portamento in 84. Similarly, the addition of two 1st-finger markings at 117 [p.11, s.10, b.5] keeps the passage on the same string and includes a portamento – characteristic traits of David’s aesthetic.

With the exception of hand-written additions for simple practical reasons (for example, writing out the first few bars of the development section of the 1st movement to facilitate a page turn; the word ‘bleiben’ being added at bar 119 to confirm remaining in position; highlighting the beginnings and endings of repeat sections in the Finale, etc.), the remainder of the annotations (apart from David’s substantial/compositional changes, which will be discussed below) are dynamic indications, aligning with the violin part, and which appear (as in K. 423) to create a more dramatic or vivid performance style – intensifying softer dynamics by increasing their meaning – as in changing the p to pp in bar 210, and pp to ppp at bar 216 (first movement).

The most striking feature of the viola part, of course, is David’s re-composed pizzicato slow movement, in which Mozart’s original (predominantly double-stop) viola part is simultaneously thinned out (by means of the pizzicato sonority) and thickened (by means of the richer chromatic harmonies of the predominantly triple- and double-stopped chords as at bars 6-8 in particular). Here, the pasted sheet has itself been edited in David’s blue crayon – changing the dynamic markings in places and (as seen in the crossed-out messa-di-voce in bars 21-22 [Andante, 3rd stave]) clearly experimenting with such effects in performance. Comparison of the ink dynamic markings and crayon markings show that these are in the same hand, graphological differences being small enough to be explained by the different writing media. Such ‘re-composition’ is found again on a smaller scale in additional notes to accompanying chords (pencilled in bar 8 of the first movement [p.7, s.3, b.1], and in crayon at bars 211, and 214-219 [p.9, s.6]); these testify to David’s endemic confidence in adapting, revising and changing Mozart’s text (as indicated, at least, by the 1792 Artaria edition) – an attitude to the notation of such music rather unfamiliar to performers of recent times.

C3. Recording from the ‘Uppingham’ Collection: Practical Considerations, Problems, Difficulties & Decisions – Personal Perspectives of the Performers (David Milsom & Clive Brown)

The gap between score and sound is always a large one, entailing a complex web of practical, stylistic and technical decisions which go far beyond the possibilities of even the most detailed score notation. These limitations also apply to this project, despite the fact that David’s markings (and in particular, the crayon annotations) deliver a richer vein of score evidence than is customary, even in respect of 19th-century editions which generally contain more detail than equivalent documents in earlier musical epochs. We have attempted to interpret the annotations in Ferdinand David’s editions and personal copies in the light of the knowledge and experience gained from practical experiments, which have been informed throughout by our own and other scholars' work on historical performing practices.

The evidence that underlies the present practical experiment falls into two distinct categories: that which is embodied in David’s editorial markings; and that which is understood to be inherent in the notation. The former is reflected in our execution of dynamics, position changes with varying degrees of portamento, sparing, ornamental use of vibrato (which is, in many cases, exemplified in David’s use of open strings and harmonics), and the realisation of various types of bowing; the latter is evident in our employment of the kinds of tempo and rhythmic flexibilities that we believe to have characterised that school and period. The aspects of our performance that were suggested by David’s markings in the parts, as well as those that derive from external evidence, draw upon not only documentary sources but also early recordings by heirs of the classical German tradition (Joseph Joachim, Leopold Auer, Marie Soldat, and Karl Klingler and his quartet).

The practical and philosophical basis of our research is articulated in detail in our publications and informed by the experience gained from Milsom’s AHRC Fellowship in the Creative and Performing Arts (2006-9), which is articulated in the project’s scholarly and practical outcomes. More recently, we have continued to experiment with the practical implications of 19th-century performance annotations, as well as the relationship between notation and performance more generally, in numerous research-based performances by the Ferdinand David Ensemble. Much of the intellectual infrastructure that informs our approach is to be found in Brown’s Classical and Romantic Performing Practice 1750-1900 [Brown, 1999] and Milsom’s Theory and Practice in Late Nineteenth-Century Violin Performance 1850-1900 [Milsom, 2003], while some of the outcomes of our practice-led research are also reflected in other, recent publications by the authors, including:

Brown, C: ‘Performing Classical repertoire: the unbridgeable gulf between contemporary practice and historical reality.’ Basler Jahrbuch für historische Musikpraxis XXX 2006 (Winterthur, Amadeus Verlag, 2008), pp. 31-44

Brown, C: ‘Singing and string playing in comparison: instructions for the technical and artistic employment of portamento and vibrato in Charles de Bériot’s Méthode de violon’ in Zwischen schöpferischer Individualität und künstlerischer Selbstverleugnung ed. Claudio Bacciagaluppi, Roman Brotbeck & Anselm Gerhard  (Schliengen, Argus, 2009)

Brown, C: ‘Leopold Mozart’s Violinschule and the performance of W. A. Mozart’s violin music’ in Cordes et claviers au temps de Mozart / Strings and Keyboard in the Age of Mozart ed. Thomas Steiner (Lausanne, Peter Lang, 2010)

Brown, C: ‘Reading between the Lines: the Notation and Performance of Mozart’s Chamber Music with Keyboard’ in Mozart’s Chamber Music with Keyboard, ed. M. Harlow (Cambridge University Press, 2012) 

Milsom, D: ‘The Performance of Mendelssohn’s Chamber and Solo Music for Violin’, ed. S. Reichwald Mendelssohn Performance Studies (Indiana University Press, 2008)

Milsom, D: ‘Joseph Joachim: Evoking his Style and Practice in Historically-Informed Performance in M. Callela & C. Glanz (eds) Anklaenge 2008 – Wiener Jahrbuch für Musikwissenschaft (Vienna: Mille tre Verlag, 2008)

Milsom, D: ‘Practice and Principle: Perspectives upon the German ‘Classical’ School of Violin Playing in the Late Nineteenth Century’ in Nineteenth Century Music Review, Vol. 9, Issue 1, June, 2012 (Cambridge: Cambridge University Press, 2012)

Milsom, D: ‘Expressiveness in historical performance – 19th-century ideals and practices (on paper and in sound), book chapter co-authored with Neal Peres Da Costa, in D. Fabian, E. Schubert & R. Timmers (eds) Expressiveness in Music Performance: Empirical Approaches Across Styles and Cultures (Oxford: Oxford University Press, 2013)

C3 a) Performance Perspectives with Regard to the Violin Part (Clive Brown)

NOTE: It is intended that the reader will refer, for illustrations, to the downloadable pdf copies of Ferdinand David's performing parts available through the CHASE website [view]. Hyperlinks within the text lead to an online view that includes, pop-up annotations via clickable boxes. References to these copies are given in the form [p.1, s.1, b.1] (i.e. page 1, stave 1, bar 1).
[In this section, some repetition of ideas and information that have already been presented elsewhere in my published work will be inevitable, but for clarity it will be convenient to reiterate them here.]

General Observations

Our recordings of Mozart's Duos for Violin and Viola here are not to be seen in any sense as finished artistic products, nor, despite the scholarly credentials of the performers, are they to be regarded as authoritative recreations of a historical style, in other words, performances of these works as they might really have sounded in the hands of two of David's faithful, if not always technically reliable, students. It goes without saying that any trained string players of David's time would have expressed themselves musically in a language of performance about which we can know much, but by no means everything. 

These recordings are not even to be seen as a summation of the fruits of our scholarly research. They are part of a process, analogous in many ways to the scientific method of subjecting theories to experiments, on the basis of which knowledge may be incrementally advanced by validating some aspects of the hypothesis and raising doubts about others. Further experiments, taking account of the findings of the earlier ones, may subsequently refine the theory and even, perhaps, lead to a genuine breakthrough in understanding that has the potential to make a transformative impact beyond the academic world. Thus, an investigation through practice, of the kind we have undertaken here, may help to produce real advances within the related disciplines of scholarship and performance. In the following paragraphs I will attempt to articulate the parameters of our experiment, as I see them, and to evaluate the results.

Perhaps surprisingly, since the most obvious purpose of the research is to estimate the aural implications of Ferdinand David's markings in this remarkable source, I want to start with what he did not mark in the music. The experiment cannot be meaningfully conducted at all if we try merely to interpret David's performance annotations in relation to his known practices, while responding to other aspects of the notation in a fundamentally modern manner; that is to say, if we play the notes with 20th/21st-century attitudes towards rhythmic precision, and with the kind of phrasing, rubato and ensemble we learned during our training and earlier experience. The types of portamento, vibrato and bowing styles implied by David's markings would then simply be superimposed upon an otherwise modern performance style, creating an effect rather like a modern building incongruously embellished with 19th-century cornices, pediments and pillars.

In addition to understanding David’s markings, therefore, we must try as far as possible to understand how mid-19th-century approaches to phrasing, rubato and ensemble, which undoubtedly retained elements of 18th-century practice [see Brown, 2008a, 2010a, and 2012], differed from our own; we must try also to evaluate the extent to which particular practices were felt by 19th-century composers and performers to be inherent in the notation. For this purpose early recordings are of key importance, especially those of David's violin pupil Joseph Joachim (1831-1907) and the pianist Carl Reinecke (1824-1910) who partnered David in many chamber music concerts at the Leipzig Gewandhaus between 1860 and David's death in 1873. Reinecke's Mozart recordings are of particular importance for our experiment in this respect, because he was widely acknowledged by contemporaries as guardian and exponent of a genuine tradition of Mozart performance. 

Alongside these recordings of leading mid-19th-century musicians, we may also draw upon the recordings of two of Joachim's leading students, Marie Soldat (1863-1955) and Karl Klingler (1879-1971) in whose playing many of the conservative traits observable in the older musicians' recordings are apparent. 

The recordings of all these artists trained in the 19th century exhibit a degree of freedom with regard to written note lengths, but at the same time a strong tendency to keep these rhythmic freedoms within a fundamentally regular pulse. This is entirely untypical of modern playing, but undoubtedly closer to 18th-century practice, at a generic level if not in its specific application.


Prominent in the Mozart recordings of Reinecke, Soldat and Klingler (Joachim did not record any Mozart) [see Part F: List of References] is a propensity to lengthen strong beats and proportionately shorten weak ones. This is particularly the case in slurred figures especially pairs (reflecting Leopold Mozart's instructions [see Brown, 2010a]). They also demonstrate an elastic approach to ornamental figurations, taking time and hurrying in more asymmetrical ways. These practices have consequences for ensemble, which is often much less regimented than in modern performance. This can be observed to an extraordinary degree in Reinecke's piano playing, in which the right and left hands often seem remarkably independent of one another.

We have therefore attempted to emulate such practices alongside our realisation of David's markings. In our experimental performances during the last four or five years I have tried to internalise this approach to rhythm and apply it spontaneously and unselfconsciously, as the musicians of that time must have done. Listening back to the recordings I find that many of these rhythmic and tempo nuances convincingly reflect the early recorded sources and seem musically persuasive, even when they result in dislocation between a melodic part and an accompaniment; but others seem less convincing, occasionally stilted and clumsy, sometimes producing a type of disruption in the ensemble, which I suspect David would have considered faulty.

Dotted rhythms 

It is clear from a wealth of evidence that rhythmic figures in which there is a 3:1 proportion, such as dotted-1/4-note:1/8-note, were seldom expected to be executed as strictly as they generally are in modern performance. They will frequently have been over-dotted, or, in other contexts, assimilated to triplets. Instances of over-dotting can be heard frequently in our recording, in fact we rarely play such figures as written. There are no instances in the Duos where the triplet execution would be appropriate.

One distinctive interpretation of the reverse dotted (Lombardic or Scotch-snap) figure in the violin part occurs in the Rondo of K. 423 at bars 181-2 [p.7, s.9, bb.2-3]. In the recording I tried to execute them with a very short initial note, almost like an acciaccatura, thus placing the accent on the longer note, emulating the distinctive manner in which Karl Klingler performed these figures in the Trio of the Menuetto of K. 421 on the Klingler Quartet's 1912 recording. His execution of the figure, probably learned from Joachim, accords with instructions by some 18th-century theorists [see Brown, 1999: p.35f].

Left-hand techniques

The execution of David's fingering is in many ways much more straightforward. In the violin part I observed all David's fingerings, giving preference to his handwritten changes where these replaced printed fingering. Since his fingering in the violin part is comprehensive, there were very few occasions on which it was necessary to supplement it. I took the view, amply justified by 19th-century sources, that where a change of position occurred within a bowstroke, this would be audible to a greater or lesser extent. The only partial exception to this is where a single finger moves relatively rapidly by a semitones or sometimes a tone, where the shift might be barely or not at all noticeable. 

I attempted to make portamento effects appropriate to the musical context. As explained in section C1 above, it is clear that David took great care to locate position changes that were necessary for technical reasons in places where he believed they would be most effective musically. In circumstances where a passage could perfectly well be played without shifting the hand, he often contrived a position change to produce a purely expressive slide and in such cases I tended to make the portamento more prominent in the recorded performances.

Many of David’s portamento effects are connected with a desire to retain the colour of a particular string in the manner of Rode, Spohr and many other early 19th-century players; but David’s position changes in such cases seem almost always to be carefully selected to enhance expression. In this respect his printed fingering is generally less expressive than his handwritten alterations and additions, suggesting that David adopted a more conservative attitude when supplying fingering for the users of his edition than he was inclined to employ in his own performances. 

There are a few noteworthy fingerings, the intentions of which are not always clear. In K. 424 at bars 55-6 of the first movement [p.9, s.1, bb.5-6] there is a succession of 1st fingers followed immediately by a 4th-finger portamento. I tried making the passage as smooth as possible without actually making it a slow glissando, and also, on the first time through the exposition, introduced some rhythmic inequality. 

In some cases a portamento between bow strokes (carrying up the note with the end of one stroke and connecting it with the next by a seamless bow change) was undoubtedly employed by 19th-century violinists. There are few instances of this in the Mozart Duos, but some of the places where a delicate portamento connection of this kind can be heard in the recording in the violin part of K. 424, for instance in the very first bar of the opening Adagio between the first and second notes, in bar 7 of the Andante from the c'' to the f'' [p.11, s.5, b.1] and from the d'' to the a''' in bar 24 [p.11, s.8, b.4]. I am convinced that in later repertory David would often have introduced such effects more frequently and probably executed them in a more prominent manner.

In almost all cases of portamento in the violin part I believe that the speed of the shifts and the degree of bow pressure I applied has produced effects that would have been familiar to David, though in particular cases he would doubtless have had suggestions for improving their effectiveness. Finer nuances, of course, would have differed from violinist to violinist and performance to performance.

For David and his contemporaries, it seems clear that portamento was a more important expressive resource than vibrato. Abundant evidence demonstrates that David was quite sparing in his use of vibrato and I believe that he would have regarded the range of vibrato effects used in the recording as fundamentally tasteful even if, as a teacher, he might have quibbled over some particular instances. In accordance with Spohr's and David’s instructions I applied vibrato with varying degrees of speed and intensity at the high points of a musical phrase and on accents, or to vary the quality of sustained notes. On the fermata at the beginning of David's interpolated cadenza in the Andante of K. 424 [pasted manuscript at the bottom of p.11]  I attempted a vibrato of the type described by Leopold Mozart and also illustrated by Spohr, which begins slowly and increases in speed. 


Seamless legato was a fundamental feature of 19th-century performance, in which smoothness of bow-change was cultivated to a high degree. I aimed to sustain and connect notes whenever an articulation was not specifically indicated by a rest; in a few cases the most imperceptible bow change possible was required, for instance in the sustained g' in bar 31 of the K. 423 Adagio [p.4, s.12, b.2]. I did not necessarily regard staccato marks as indicating a significant separation between notes [see Brown, 1999: 168] and generally made no pronounced break. There is much documentary and circumstantial evidence to support the view that David would have understood the notation in this manner. His insertion of printed rests at the beginning of the Rondo in K. 423 and at the beginning of the last movement of K. 424, strongly suggests that he felt that by default the notes would have been given greater length. The use of rests rather than staccato marks to specify separation can be seen in the notational practices of many mid-19th-century composers, including Mendelssohn [see Brown, 2008b].

In choosing bowstrokes for more rapid figures I was guided by the instructions in David's Violinschule. We performed the majority of unslurred rapid notes with various lengths of détaché (David's liegender Bogen / on the string) or martelé (David's fester Strich an der Spitze / firm stroke at the point). Where notes marked with staccato dots or horizontal lines occurred under slurs, I employed various gradations of slurred staccato near the point or longer portato strokes depending on context. David indicated no hopping or springing bowstrokes in these Duos, and we did not use any.


On listening to the recording it was clear to me that, although we made many of the dynamic nuances indicated by his printed or handwritten annotations effective, we did not do full justice to David's dynamic scheme. Greater contrast between the extremes was sometimes needed; I think we often failed to achieve a genuine pianissimo.

Technique in general 

Until the end of 2011 my posture and technique were essentially modern. Since then I have been experimenting intensively with handling the instrument in a manner that accords more closely with known 19th-century practices (see my article ‘Physical parameters of 19th and early 20th-century violin playing’). This meant adopting the low right elbow position, much maligned by modern violinists, and the old German bow grip. It also involved a radical rethinking of the way in which I hold the violin and use my left hand. I play without a shoulder rest, using only a very thin (less than a centimetre) foam pad under my shirt to prevent bruising to the collar bone when I play in shirt sleeves (David recommended ‘a cloth or a small cushion’ [David, 1863: i, 5]. I have changed my modern, fairly substantial chin rest for a small and very low late-19th-century one. David, however, as far as we can tell, did not use a chin rest at all, despite Spohr having introduced one before David studied with him. Consequently, if one is to hold the instrument horizontally, as instructed in all the major 19th-century methods, it must be supported continuously by the left hand.

For the most part this means that the neck rests between the top joint of the thumb and the third joint of the index finger. In modern playing the base of the index finger is generally kept clear of the neck as much as possible to facilitate vibrato, while for David it is only removed from the neck if the player wants to introduce an ornamental vibrato. 

When changing position, the hand can easily advance to third position without disturbing the location of the violin, since it is being pressed against the neck. In descending, however, I make the thumb precede the rest of the hand into a lower position as instructed by David and Joachim; this is a technique I did not previously use, having moved the whole hand in a more or less constantly maintained shape at least as high as fifth position.

As yet I have been unable to find more detailed information about the movement of the left thumb, but I am coming increasingly to the conclusion that David and his colleagues, relying on the left hand rather than the chin to support the instrument in its playing position, must have employed the thumb much more as a pivot than is done in modern violin playing. This is strongly suggested by some of the fingering patterns that occur in 19th-century sources, particularly when upward and downward portamento shifts occur in rapid succession. To date I have not encountered sufficiently explicit documentary evidence to confirm the details of such procedures, but circumstantial evidence convinces me that the violinists of that period must have employed the thumb in something like this manner. When, for instance, there is a change from third to fifth positions followed shortly by a return to third, I have experimented with adopting a thumb location somewhat more advanced towards the body of the violin than usual in third position, and retaining that placement of the thumb while the rest of the hand moves to fifth position and then returns to third. One instance in our Duo recordings occurs at bars 55-6 of the Bb-Major Duo first movement [p.9, s.1, bb.5-7]; during these successive 1st fingers I kept the thumb in the same place on the neck throughout.

I do not yet feel entirely secure with these left-hand experiments, and this sometimes shows in unreliable intonation, but I am becoming increasingly comfortable with them, and find that they are particularly conducive to producing controlled portamento.

C3 b) Performance Perspectives – Viola Part (David Milsom)

NOTE: It is intended that the reader will refer, for illustrations, to the downloadable pdf copies of Ferdinand David's performing parts available through the CHASE website [view]. Hyperlinks within the text lead to an online view that includes, pop-up annotations via clickable boxes. References to these copies are given in the form [p.1, s.1, b.1] (i.e. page 1, stave 1, bar 1).

General Observations

Whilst the purpose of the experimental recordings made as part of this project is to illustrate, in as much as the evidence allows, how Ferdinand David might have performed these Duos in Leipzig in the 19th century, we acknowledge fully the ultimately unrealisable nature of this goal. This is underlined by looking at the viola parts of these duos, which raise at least as many questions as they answer. David, as observed above, marked up the violin parts in considerable detail, but this level of detail is not reflected in the viola parts. The G-Major Duo in particular has few markings, and none in the rondo Finale. The B-flat Duo is rather better in this regard (and of course, there is the fascinating and potentially important intelligence of the pasted-in pizzicato version of the slow movement), but in numerous places performance information is relatively scant. In respect of bowing and fingering – parameters that naturally give rise to a degree of ‘personalisation’ by performers in order to bring about an effective and individually-practicable performance – decisions need to be made, and the bases of these decisions engender an amount of (informed) guesswork. It is never clear, with printed bowings, for example, whether or not a scheme is intended to be comprehensive, or whether the bow indications that are given are simply indicative. Fingerings come into the same category, and where markings are absent, one must reasonably assume that something other than staying in one position might be practised. In the first movement of K. 423, for example, the first printed fingering occurs in bar 32 [p.2, s.5, b.6]; whilst it is possible that all of the movement up to this point (approximately two-thirds of the exposition) is taken in the first position, it seems (to me at least) relatively unlikely that a player of David’s calibre would restrict himself to such elementary solutions. This gives rise to the fingerings I employ at bars 5, 26, and 30, for example, all of which (and the sliding 4th finger to a harmonic in bar 5 in particular) are meant to reflect the fingering systems associated with David and this performance tradition.

The purpose of our recorded performance is, on one level, to put into practice the content of the printed edition and, where evident, the hand-written markings by David, the latter taking precedence where a disparity opened up. On another level, an attempt was made to present a creative response to the score, contextualised by our scholarship into the practices of this ‘school’. Obviously, there is much that can be learnt only from recorded sound and, to adapt a well-known saying, ‘a recorded note is worth more than a thousand words’ in this respect. Ferdinand David did not live long enough to make recordings, of course, and the absence of actual aural evidence (particularly one considers the gap between theory and practice, as I have done in respect of recording artists [see my AHRC Fellowship project] severely limits any claims of verisimilitude we may make. Having said this, we have tried to reflect a context of scholarship regarding all available surviving documents including, crucially, early recordings.

As some have argued, the attempt to project stylistic traits ‘backwards’ can be methodologically flawed, since it is impossible to quantify the extent and nature of changes over time [see, for example, Bowen, 1996]. Daniel Leech-Wilkinson argues that numerous tiny changes to style and reception can be cumulative [see Leech-Wilkinson, 2010). Nonetheless, I have always argued that, if the artists in question are selected with care and the observed features treated with all due circumspection, this approach has so much to offer our understanding that it would be perverse to discount it. As a consequence, the sound world (especially as regards two of the most obvious traits – vibrato and portamento) interfaces not only with the paper evidence of the period, but also with sound recordings made by relevant violinists who are likely to have retained at least something of the architecture of David’s sound, albeit filtered through subsequent stylistic changes and indeed personal preferences. In this respect, recordings by Joseph Joachim and his evidentially-faithful pupil, Marie Soldat are invaluable, even when transmitted via the murky portal of acoustic recording. (Joachim’s complete recordings of 1903 for the Gramophone & Typewriter company have been re-issued on OPALCD 9851; Soldat’s complete recordings of c.1920-1926 on the Union A label were re-issued some decades ago in James Creighton’s LP Masters of the Bow series, MB1019). In more general matters of musicianship – what one might term generically ‘phrasing’ and ‘tempo flexibility’ – these recordings can be supplemented by a significant further body of evidence, including the Welte-Mignon and Hupfeld piano rolls made in c.1905-1907 by Leipzig-based pianist (and colleague of Ferdinand David) Carl Reinecke (1824-1910). These are analysed and discussed extensively in Peres Da Costa, 2012.

My approach to playing the viola parts is clear from listening to the performances whilst following David’s performing parts, but it is perhaps useful to outline some examples of performance decisions I took in the recording sessions.

Style and Approach – Viola as Accompaniment or Duo Partnership?

A generic problem with attempting historically-motivated performance is that it is all too easy to read historical evidence (or indeed, listen to it) with the eyes and ears of the present. This much is obvious, but some of the limitations can be easily overlooked. A guiding principle in appraising 19th-century performance style from today's viewpoint, is that many aspects of taste have changed so much in the intervening period that what was considered 'good' then is often now deemed to be ‘bad’. For example, informality of rhythms, tempo flexibility (especially when this includes ‘rushing’),departures from the musical text, portamento slides, etc. are not generally embraced by present-day performers – at least, not to the extent perceived through early recordings. Epitomising this change of attitude, one might consider Joachim and Moser's criticism of Henri Vieuxtemps (1820-1881) for his inability to 'read between the lines' – a charge that implies an expectation of interpretational freedom from the performer [Brown, 2003].

Trying to meet the past on its own terms is an established philosophy in many facets of HIP. Despite this, there is still a degree of reticence to practice style traits that are (for reasons that strike me as unaccountable) deemed ‘dangerous’ (the portamento in particular [see Brown, 2010: 480]). Furthermore, modern players involved in HIP often still fail to question a number of principles in performance that have been arrived at mimetically through the pedagogic agency of the 20th century. These include:

  1. Precision of intonation (usually ‘equal temperament’ or received notions of Casals’ ‘expressive intonation’)
  2. Precision of ensemble
  3. In chamber music, a democratic ideal of music making with ‘give and take’ between the players, allowing melodic material to be always in the foreground with accompanimental material keeping time with it in the background.

In contention with point 3, however, it is clear from descriptions of tempo rubato in respect of accompanimental figures (which should maintain strict time) and melodies (which might be free) that synchrony was not necessarily the aim of the 18th- and 19th-century chamber musician. Indeed, melodies might ‘follow’ accompaniments rather than the other way round, making for a very different balance of ensemble priorities. Such matters are well accounted for in the surviving literature; for example, Spohr’s description of concerto playing and orchestral playing in his Violinschule assigns a lower status to orchestral players, who must avoid the solo embellishments (including tempo rubato) that he describes as ‘fine style':

The division of each part of a bar, according to the value of its time, must in Orchestral playing be strictly observed. The tempo rubato (a slight delay on single or more notes)…[DM – this in itself is an interesting definition of the tempo rubato]…in the Solo of great effect cannot here be tolerated.’  [Spohr, 1843: 231]

The recorded evidence of ensemble practices of, say, the Klingler Quartet [see Brown, 2006: 41-42], or the independence of left and right hands heard on piano rolls made by Reinecke [a particularly revealing example is discussed ibidem: 37] further supports this view. The implications of ‘rank’ (given the more hierarchical structure of past societies) also impinge here in a way that is conspicuously at variance with modern ideals.

As regards Ferdinand David himself, his high standing in Leipzig, as well as some less-than-complimentary accounts of his performance style, testify at the very least to a strong artistic personality who may have dominated the texture of Duos such as these. Thus, he was criticised for his ‘virtuoso’ approach to quartet leading which might imply that, for him at least, the role was imbued with greater liberties than for the others, as Otto Jahn in Der Grenzbote felt in 1855:

Unfortunately he developed an ever-forced manner, which is opposed to the self-effacement which is indispensable in Quartet playing. Moreover, he introduced all sorts of cheap (effects) to give piquancy to the works of Haydn and Mozart. The manner in which Herr David coquets especially in Haydn quartets, as if he wanted to show what he can make out of a Haydn quartet how […] for instance, he plays accompaniment figures, as if he wanted to say ‘Thus accompanies the first violin’ is boastful and in bad taste. [Milsom, 2003: 19]

Jahn’s criticism suggests that David was out of step with contemporary ideals of performance taste – one might assume that the reviewer subscribed to the current notion of good chamber playing in which such exposure of the first violin is deemed inappropriate, although, of course, such qualitative judgements are comparative and to suggest that Jahn agreed with modern practice is perhaps disingenuous. More than twenty years earlier, however, David’s teacher, Louis Spohr, indicated that chamber music playing embodied characteristics familiar to present-day estimations of good practice: 

The delivery of the regular Quartett [sic] demands a very different treatment. In such a composition it is not intended that one instrument should exclusively predominate, but that each should enter into the spirit of the Composer and delineate it accordingly.
The power of tone of the first violin and the manner of playing must be in keeping with the rest, and where it is not the principal it should remain subordinate. As the style of delivery should always proceed from the idea and spirit of the composition, it is required of the Solo player, in the Quartett to lay aside his peculiar manner of Solo playing, and accommodate himself to the character of the Music. Until he be capable of this, he cannot discern the character of the separate parts of the Quartett and give proper effect to the variety of style displayed in classical compositions. [Spohr, 1843: 230.]

Whether or not Spohr and Jahn are in implicit agreement is unquantifiable. Perhaps it is simply David’s embellishments that Jahn dislikes. Whether or not David’s re-scoring of the viola part of the B-flat Duo's second movement comes into this category is open for debate – whilst such an act would perhaps be unthinkable to today’s musicians, it must be noted that the reviewer of the 1870 Leipzig performance made particular mention of its enthusiastic reception. What all of this does imply (perhaps not surprisingly) is that David had a strong soloistic personality even when playing chamber music. As a consequence, and knowing that David performed the viola part in these Duos, it seems plausible to assert that the viola part is at least rendered equal to the violin part here. This allows for a slightly more ‘embellished’ approach to the viola part in terms of performance style (tone colouration, expressive fingerings, bowings, dynamic balance, tempo and rhythm flexibility, etc.) than might be implied by the scant markings in the viola part. This, of course, necessitates adding to David’s markings (especially in the case of K. 423), by a combination of reflecting the style and nature of the markings David made in the violin part as well as practising traits of style learned from previous academic and experiential scholarship.

Accompaniment figures

A striking feature of the performance of Mozart’s Violin Concerto no. 5 (first movement) by Marie Soldat [Soldat, 192?] is the inequality granted to running 1/16-note figures, which is corroborated by other early aural evidence by the first generation of players to make records. The relevance of this to the Duo project here is obvious: Soldat’s performance (in the absence of such repertoire being recorded by Joachim) gives us a valuable glimpse of how a musician trained in the stylistic approach of the late-19th-century ‘German school' might approach the material. Marie Soldat's genealogical connections to the 'classical German school' (and to Brahms and Joachim in particular) are discussed in Milsom, 2012: 31-52. As Clive Brown has argued, this has much in common with 18th-century ideals of inequality with which Mozart would have been familiar [see Brown, 2006: especially 34-41]. This approach is endemic in early recordings. Reinecke’s piano rolls are full of such inequality of parts, as well as manipulations of equal note values to favour harmonically-or metrically-significant notes, as in his performances of Field’s Nocturne no. 4 [Reinecke, c.1905-7, viewable on Youtube] or his own Ballade, op. 20 [Reinecke, c.1905-7, viewable on Youtube], for example. I have tried to reflect such matters in relevant passages in the Duo recordings; for instance, bars 114 and 118-119 of K. 423 mvt I, bar 2 of K. 423 mvt II, and bar 51 of K. 424 mvt I, as well as numerous other places. More generally, I attempted to perform with a degree of rhythmic informality: ‘swinging’ rhythms in a manner that is prevalent in Reinecke’s piano rolls, and varying the performance of dotted figures commensurately (as in Variation 1 of the K. 434 Finale).

Pizzicato 2nd Movement – Bb-Major Duo

The pizzicato movement presented a number of challenges. David’s re-scoring is technically very awkward, and the span of some of the quadruple-stopped chords (particularly in bar 8) attests to the flexibility of his hands. I decided to adopt Louis Spohr’s advice to place the viola in a ‘banjo position’ (under the right arm), given that the entire movement is pizzicato:

If only a few tones are to be played pizzicato, and the coll’arco following quickly after it, the Violin is to remain in its ordinary position. The bow then is taken in the full hand, and held at the nut by the three last fingers of the right hand; but the thumb is placed with the ball of the hand against the lower edge of the fingerboard, when the string is pulled with the point of the first finger.
But if the pizzicato continues for a time, or if a pause precedes the coll’arco, it is better to take the Violin down. It is then placed with its back against the right side of the body and supported with the right back arm. The bow is to be held in the above stated manner; in lieu of the thumb, however, the first finger is placed against the fingerboard (yet a little distance from the edge,) and the string pulled with the thumb.
This method, in long passages, is preferable, as the pizzicato with the thumb produces a fuller and clearer sound than with the first finger.’  [Spohr, 1843: 154]


TITLE PAGE, Abstract, Acknowledgements, Citation 
Part A: INTRODUCTION (David Milsom & Clive Brown)
Part D: EXPERIMENTAL RECORDINGS (Clive Brown – violin; David Milsom – viola) 
Part E: CONCLUSIONS (David Milsom)