Six Sonatas for Violin and Harpsichord, BWV 1014–1019


The six sonatas for violin and obbligato harpsichord BWV 1014–1019 by Johann Sebastian Bach are works in trio sonata form, with the two upper parts in the harpsichord and violin over a bass line supplied by the harpsichord and an optional viola da gamba. Unlike baroque sonatas for solo instrument and continuo, where the realisation of the figured bass was left to the discretion of the performer, the keyboard part in the sonatas was almost entirely specified by Bach. They were probably mostly composed during Bach's final years in Cöthen between 1720 and 1723, before he moved to Leipzig. The extant sources for the collection span the whole of Bach's period in Leipzig, during which time he continued to make changes to the score.

Origins and compositional history

Bach's sonatas for violin and obbligato harpsichord were composed in trio sonata form, i.e. three independent parts consisting of two equally matched upper voices above a bass line. Instead of playing the role of a continuo instrument, filling in the harmonies of a figured bass, the harpsichord took one of the upper melodic lines on equal terms with the violin, whilst also providing the bass line.
In the totality of Bach's musical output, the instrumental sonatas written in trio sonata form are small in number. Apart from the BWV 1014–1019, there are the six organ sonatas, BWV 525–530, the three sonatas for viola da gamba and harpsichord, BWV 1027–1029, and the three sonatas for flute and harpsichord, BWV 1030, BWV 1031 and BWV 1032. In each case the trio sonata texture derives from the compositional form and not the particular combination of instruments, which was partly a function of the musicians at Bach's disposal. This is well illustrated by the first movement of the organ sonata BWV 528 which originated as the sinfonia starting the second part of the cantata, BWV 76, with oboe d'amore and viola da gamba as solo instruments; and likewise by the trio sonata for two flutes and continuo BWV 1039 and its alternative version for viola da gamba and obbligato harpsichord, BWV 1027.
Although it had been believed for some time—and advanced as a theory by —that the sonatas BWV 1014–1019 must have originated in lost trio sonatas for two instruments and continuo, no prior versions have been discovered and it is accepted that only a few movements could have such an origin. The first known source from 1725, in the handwriting of Bach's nephew Johann Heinrich Bach, explicitly specifies an obbligato harpsichord; and, despite the fact that a later version in the hand of Bach's pupil Johann Friedrich Agricola has a marginal "Violin I" at the start of BWV 1014, the scoring of the upper part in the keyboard, especially in the adagio movements BWV 1016/i, BWV 1017/iii and BWV 1018/iii, uses figures that are idiomatic to a keyboard instrument but unsuited to other instruments. Although this compositional style became widespread in the late eighteenth century, in Bach's day it was unusual and innovative. Although all the sonatas are written in trio sonata form, each has its own distinct character—the third is an example of the Sonate auf Concertenart, a sonata written in the style of a concerto. Throughout his life Bach returned to the sonatas to refine and perfect the score, particularly in the last sonata, which survives in three different versions.
When Wolfgang Schmieder created the chronology for the BWV catalogue of Bach's works in the 1950s, the assumption was that Bach's musical output matched his responsibilities in each of the three distinct phases in his career: the period 1700–1717 when he was organist at Lüneburg, Arnstadt, Mühlhausen and Weimar; the period 1717–1723 when he was Capellmeister at Cöthen; and the period from 1723 onwards when he served as Thomaskantor in Leipzig. Accordingly, the chamber music works by Bach were automatically assigned to the Cöthen period. Later generations of Bach scholars have recognized that Bach's involvement with chamber and orchestral music continued in Leipzig, especially through the Collegium Musicum; and accordingly Schmieder's rigid chronology is no longer generally accepted. Nevertheless, even though there is no direct confirmation for the dating of BWV 1014–1019, Bach scholars agree that the circumstances surrounding the 1725 source probably point to the first versions of these sonatas being composed between 1720 and 1723 during Bach's last years in Cöthen. In the 1958 Neue Bach-Ausgabe edition, the editor and musicologist Rudolf Gerber was unaware that the 1725 manuscript had been largely copied by Bach's nephew, who was only a pupil at the Thomasschule at the time. In addition two of the three last movements in the sixth sonata copied by Bach himself were borrowed from the sixth keyboard partita BWV 830, movements also included in the 1725 Notenbüchlein for Bach's wife Anna Magdalena Bach. This suggests that the initial collection of sonatas, assembled for an unknown purpose, was probably copied from pre-existing compositions and hastily completed. This hypothesis is not only compatible with Bach's heavy compositional duties as Thomaskantor at the start of his period in Leipzig; but also agrees with the dating of the sonatas to Cöthen by Bach's biographer Johann Nikolaus Forkel: a letter to him in 1774 from Bach's son Carl Philipp Emanuel describes the sonatas as being 50 years old.
The history of the sixth sonata BWV 1019 is distinct from that of the five others. The three different versions of the sonata and its successive comprehensive modifications in Leipzig indicate that its role in the collection evolved only gradually. The two first movements, a large scale concerto allegro and a short largo, remained largely unaltered throughout these revisions and were copied by Bach's nephew Johann Heinrich into the earliest surviving manuscript from 1725. The originals, assumed to date from Cöthen, are lost; but it is probable that these were the first two movements of a three-movement Sonate auf Concertenart. In the 1725 manuscript the remaining movements were entered by Bach himself. The sonata took the following form:
  1. Vivace, G major
  2. Largo, E minor
  3. Harpsichord solo, E minor
  4. Adagio, B minor and G minor
  5. Violin solo with figured bass, G minor
  6. Vivace, G major
The solo movements provide a contrast with the other movements, which are duos for violin and obbligato harpsichord; moreover as dance movements they add variety and lightness to the set, making it more like a dance suite. The harpsichord solo was later published in Bach's Clavier-Übung I as the Corrente in BWV 830, the sixth of the keyboard partitas; before that it had already been entered into Anna Magdalena's Notebook. The violin solo, with the harpsichord providing a simple figured bass accompaniment, was an early version of the Tempo di Gavotta from the same partita. Only the harpsichord part survives, but the violin solo for the fifth movement has been reconstructed without difficulty from the score of BWV 830; the missing violin part for the short Adagio has been recovered from the second version of the sonata.
After the publication of Clavier-Übung I, probably in the late 1720s, Bach revised the sixth sonata by excising the two published movements from BWV 830. He replaced the harpsichord solo by a lengthy Cantabile for violin and obbligato harpsichord:
  1. Vivace, G major
  2. Largo, E minor
  3. Cantabile, G major
  4. Adagio, B minor and G minor
The third movement is considered to be an arrangement of an aria from a lost secular cantata, probably dating from Bach's period in Cöthen. There is no longer any indication that the opening Vivace should be repeated in performance; the lack of a fast finale returning to the original key has been taken as an indication of the unfinished or intermediate status of this version.
The sonata attained its final form some time between 1729 and 1741 and survives in a copy made by Bach's pupil Johann Friedrich Agricola. Now with five movements and matching more closely the earlier five sonatas, it retained the first two movements but had three newly composed movements after that: a dance-like harpsichord solo in E minor in binary form; an Adagio in B minor, modulating to D major; and a gigue-like final Allegro in G major.

Musical structure

The first musical description of the sonatas for obbligato harpsichord and violin BWV 1014–1019 appeared in. In the 1960s Hans Eppstein made a systematic analysis of all the sonatas for obbligato keyboard and melody instrument, including the six organ sonatas, BWV 525–530. He determined common features in their compositional forms; part of his aim was to investigate their possible origins as transcriptions of lost compositions for chamber ensemble. Because of the complex history of BWV 1019, with its five movements and two previous versions, Eppstein gives his analysis for the first five sonatas BWV 1014–1018, viewing the movements of the sixth sonata as hybrid forms. The movements of the three versions of BWV 1019 will be discussed separately in its own section below.
The five sonatas BWV 1014–1018 are all in four movements in the conventions of the sonata da chiesa, with a slow first movement, followed by a fast movement, then another slow movement before the final allegro, often having a joyful or witty dance-like character. pointed out a uniform structure in the fast movements. They are all fugal in form but can be divided into two distinct and readily identifiable types:
  • Tutti fugue. Contrary to Eppstein's choice of name, these do not start off with a "tutti" section: they commence with the fugue subject in one of the upper parts, accompanied by a non-thematic accompaniment in the bass, which can be a bare bass line or a figured bass. The fugue subject is then taken up by the other upper part and finally in the bass. These movements have countersubjects, solo episodes, fugal development sections and a ritornello at the close.
  • Concerto allegro. These follow the model of the fast movements of the concertos of Antonio Vivaldi. Like dances, they have a binary form, i.e. are written in two sections which can be repeated. All parts play together at the beginning and there are solo episodes; the subject and countersubject are in invertible counterpoint, so can be permuted between the parts.
In general the first fast movements of the sonatas are written as tutti fugues and the closing movements as concerto allegros. There are two exceptions: in the fifth sonata BWV 1018 in F minor, the first fast movement is a concerto allegro and the closing allegro is a tutti fugue; and in the third sonata BWV 1016 in E major both allegros are tutti fugues. Both fast movements are usually linked by the musical form of their subjects. Although the binary form of the concerto allegro is usually described as "dance-like", unlike other movements of this form discussed in detail by, no specific dance forms have been associated to individual movements.
The slow movements by contrast are united only by their diversity. The violin and keyboard play different roles and there are often more than two voices in the upper parts, which can divide in the keyboard part or have double stopping in the violin. Bach explored all possibilities in the slow movements: they can resemble movements from every variety of baroque musical genre, including concertos, chamber works, dance suites, cantatas or accompanied arias; and the textures in the keyboard and the violin were often new departures, quite distinct from previously known compositions.
Unlike the fast movements, there is no longer an equality between the two upper parts and the bass, which plays a continuo role. Sometimes the bass has its own theme, as in BWV 1014/1, where it produces a partial ostinato effect; in BWV 1014/3, BWV 1016/3 and BWV 1017/3, the bass line is a genuine ostinato.
The upper keyboard part can have an independent structure from the other voices: that happens in the broken chord semiquavers or triplets that give Bach's predetermined realisation of a figured bass in the slow movements of BWV 1017; and also in BWV 1016/1 where it is divided into three voices. In the accompanying keyboard ritornello of the first movement of the F minor sonata BWV 1018, the two parts in the upper keyboard and the bass line share the same material which is echoed imitatively between them; in the third movement of the same sonata, the filigree demisemiquaver scale figures in the right hand are responded to by demisemiquaver arpeggios in the left hand.
In a few exceptional movements the upper keyboard part is directly related to the violin part: in BWV 1015/3, the two upper parts play in strict canon over broken semiquaver chords in the bass; in BWV 1016/3 in trio sonata form, the two upper parts share the same material, with invertible counterpoint and imitation; in BWV 1014/3, the right hand part adds an imitative subordinate voice to the melody line in the violin, often accompanying in thirds. In BWV 1014/1 and BWV 1016/1, there are instances when the violin and upper keyboard respond to each other, with one borrowing the thematic material of the other. In these last two movements the violin and the upper keyboard are equally matched partners.
In the majority of slow movements, however, the role of the upper keyboard part is subordinate to that of the violin and—although composed with independent material—serves the function of providing an obbligato accompaniment.