Heard by a duke

Immediately after its two sentences about Mühlhausen (p. 87), the Obituary continues with:

Does this mean that according to what he said later, it was on a chance visit that Bach was heard by the duke? This would excuse him, or be an attempt to excuse him, for taking another job so soon, not least for the sake of readers of Walther’s Lexicon who noticed that the date of one of Bach’s jobs (1707) was strangely close to the next (1708). The Obituary’s sentence need not imply that it was on the organ, or only on the organ, that he played for the duke. By convention, the court position he was offered would necessarily include duties as ‘organist’, but these would not be exclusively in chapel or on the organ, rather those of a ‘general keyboard-player’.

There was nothing unusual in musicians visiting ducal and other courts hoping to be heard, and a gifted young organist with a good position in an important church in the Free Imperial City of Mühlhausen had much in his favour. Yet while a duke would appoint without the need to invite applications, hold auditions in public or have candidates vetted by committees, a post would not have been offered unless it was vacant or about to become so. This suggests one of two things: either that the ‘visit he made to Weimar’ was undertaken in the knowledge that the court organist Johann Effler, an old acquaintance, was near to retiring; or that in visiting Weimar in the summer of 1708, he learnt of Effler’s intentions, put himself forward as successor and was auditioned there and then by one of the dukes, the senior or the junior. It is possible he was visiting Weimar to look officially or unofficially at work on the organ (completed 16 June), even invited to do so by Effler as a fellow organist. If he did visit to see the organ, perhaps this is a hint that as an enthusiastic organist he did so more often and to more organs than is now known about.

Since it is also possible that Bach’s original return from Lüneburg was with an eye to a likely vacancy in Eisenach, something of a pattern might be observed here: he heard of or anticipated a desirable vacancy and made sure to be available. His original move as a teenager from being Laquey in Weimar to organist in Arnstadt had come about when he was inspecting a newly completed organ, somehow finding it possible to stay on as its organist. He must have made a quite exceptional impression on the authorities at each of these rather different places (Arnstadt, Mühlhausen, Weimar), combining, so one imagines, outstanding professional abilities with an obvious charismatic energy. Emanuel would not know the full story of any of his appointments or have given much thought to just how many direct and indirect family connections there were in this general neighbourhood. But not irrelevant to the Weimar appointment was that decades earlier Effler had been a colleague elsewhere of Emanuel’s grandfather Ambrosius.

Recent marriage and impending fatherhood must have made better pay desirable, and it is quite possible that Maria Barbara’s condition had prompted Bach to look to Weimar in the summer of 1708. Or there were disagreements with his pastor in Mülhausen, J. A. Frohne, a defender of Pietism (qv), on what constituted a ‘well-regulated’ ensemble music in an important church. Or there were personal tensions between the two main churches’ clergy themselves, even perhaps a particular falling out that year over a cantata for 24 June, St John’s day (Petzoldt Reference Petzoldt2000, p. 187). Or Bach did not get on well with the town’s instrumentalists and preferred a court’s music-making to that of a parish church and its ambience. ‘Chamber and court organist’ was one who participated with others in a wide musical repertory, even in the theatre if the occasion ever arose, but how expected it was that Bach should compose keyboard, instrumental and vocal music in Weimar has to be guessed. Almost certainly not as much as he did compose. But the musical potential in Weimar was higher than anything he had yet known, and it is a striking fact that for Bach, promotion and better pay never led to less work and never freed him from that inner drive to compose and perform. On the contrary.

However, it is certainly the case that Bach, like a clergyman, was called by God only to higher positions. The request for dismissal from Mühlhausen has touches of self-justification and disingenuousness, or at least of a barb or two thrown by the departed, one leaving with a better offer.

The Weimar appointment

Whatever the reasons for leaving the city of Mühlhausen, personal and professional contacts there did not cease. The Marienkirche’s archdeacon and his daughter became respectively godparents to Bach’s first two children (in Weimar), and apparently his election cantata of 1709 again had its text and music printed and available in the city (no extant copy). Also, decades later he was welcomed there when accompanying his son Johann Gottfried Bernhard on the boy’s successful application at the Marienkirche in 1735. It looks very much as if he had always had better relations in Mühlhausen with people at the Marienkirche than at his own church, allowing one to speculate not so much why he left but why so soon: church doctrine or church personnel?

But Weimar, though much smaller than Mühlhausen, was a notable German ‘residence city’, seat of an absolute ruler, a domain that was subject less to powerful clergy and elected officials and more to the will of the reigning dukes. In Weimar’s case, the ‘Red Castle’ of the reigning duke, with the court chapel and Festsaal (banqueting hall), was supplemented by the ‘Yellow Castle’ of his half-brother, both of whom soon supplied Bach with a harpsichord (Dok. II, p. 41). So Bach could call on at least three locations for music-making. It was a city in whose cultural life music featured high, judging by a succession of eminent musicians who had worked there: Melchior Vulpius, Johann Schein, Heinrich Schütz (for a short time) and J. G. Walther. (Much later in Weimar, Bach seems to have been largely forgotten. George Eliot, who stayed there in the 1850s and wrote a long account of the city, makes no mention of him.)

Having ‘the opportunity to be heard by the duke’ in Weimar, probably not in the town church but in the duke’s castle, sounds as if Bach was suing for patronage. When the child Handel had been ‘heard’ by a duke as he played, it was probably on the spectacular organ in the duke’s palace chapel of Weissenfels, and as with Bach in Weimar, this was a story presumably told by the composer himself in later years. Clearly, it was a standard way of obtaining a ruler’s attention, in Handel’s case not yet for a job but for patronage or sponsorship. J. P. Kellner, an important copyist of Bach manuscripts in the mid-1720s, and probably a friend rather than a pupil, reports in his autobiography of similarly being heard by several princes ‘on command’ while he was still at student age, presumably having sought their interest first. Interestingly, although Louis Marchand said he had ‘the honour to be heard’ by the king of France, and consequently dedicated his Premier Livre to him in 1702, he still had six years to wait for royal appointment – interesting, because this foreshadows the several years Bach had to wait for royal appointment at Dresden in the 1730s.

At Weimar, the duke must have acted fast, as he was entitled to do, for by 20 June Bach’s salary was approved and he was paid as for the second quarter of 1708, at a rate nearly double that at Mühlhausen. An annual allowance of corn (18 bushels), barley (12), firewood and beer (plus freedom from alcohol excise) was additional, as was an initial grant for ‘procuring his furniture’ (Dok. V, pp. 113f.). Surely Bach had bargained for all this. On 14 July, also perhaps at his request, on entering service he received a sum of 10 guilders ex gratia. The phrase is used in connection with payment zum Anzugs-Gelde (‘for clothing money’: Dok. II, p. 35), i.e. he had to acquire court dress. So Bach wore uniform. His rôle in the chapel music has not been established in detail, but it is likely to have been modest, surely in line generally with the Weimar Church Regulation of 1664 which required the organist not to play too long before the chorales or ensemble music, and to observe a proper musical gravity (BJ 2006, p. 40). As at any court, hiring and firing of employees was not an open process, and the duke’s appointment specifies little more than terms of salary, including the payments in kind. But as for good behaviour and obedience: a duke had no need to specify.

The payment of 14 July 1708 implies that by then the Bachs (Maria Barbara pregnant with Catharina Dorothea) were resident in Weimar, and also that Bach’s successor in Mühlhausen was already in place. The new ‘chamber and court organist’ of Weimar was much better paid than most parish-church organists and indeed better than his predecessors at the court, with a substantial salary rise in 1711 and another on promotion in March 1714, surely a result of his importuning. There were also additional miscellaneous fees for extra events, for keeping the harpsichords in order (at least, in the earlier years), teaching Clavier to the duke’s page, and for engagements beyond Weimar (organ-testing in Taubach, visits to Weissenfels and probably elsewhere). Such a position could also lead to the acquiring of students, as it did. The Bachs seem to have lived in a house owned by a fellow court-musician, at least until 1713, and it will be remembered that already by March 1709, perhaps from July 1708, Maria Barbara’s sister Friedelena had been living with them.

In the professional sphere, one particularly kindred spirit (so one might guess) to visit Bach early on in Weimar was the violinist J. G. Pisendel, composer of some solo violin music, the details of whose influence on Bach one would very much like to know. There may also have been good contact with Telemann, then working in Eisenach and, as we might now say, tirelessly networking. Emanuel said that ‘in his young years’ his father was ‘often together with Telemann’ (Dok. III, p. 289), though whether this means already in Eisenach is unclear. Perhaps he meant Bach’s early twenties in Weimar, even possibly hinting that they were less ‘together’ later. Whatever the case, from Telemann’s music Bach could have continued to learn for decades what was fashionable in the musical world, both new conceptions (such as arias in 2/4 time) and newer genres (the scherzo movement, the affettuoso movement). Quite what Bach thought of the rather flabby lines of Telemann’s Concerto in G for two violins when he made his copy of it in 1709 (apparently with and for Pisendel) is unrecorded, but it does bring together three gifted musicians much affected, one can safely assume, by the dazzling new Italian styles. It was some such interest that also led at Weimar to the arrangement of yet another Telemann concerto, BWV 985, now with a connection to Prince Johann Ernst, the younger duke’s young half-brother and a composer of whom more needs to be said below.

Already resident in Weimar from a year previously was Bach’s distant relative J. G. Walther, teacher of one of the scions of the ducal house (the same Prince Johann Ernst) and organist of the town church, the church to whose parish the young Bach family probably belonged. Why Walther rather than Bach taught the young prince is unknown but may have been the result of tensions in the ducal family: Prince Johann Ernst was a member of the junior family, not the reigning duke’s, assumed to be Bach’s employer. In his later twenties Bach came to stand as godfather to a son of Walther (1712), of his own brother Christoph (1713), of the court organ-builder H. N. Trebs (1713) and of a musical colleague in Weissenfels (A. I. Weldig, master of the pages, and one of Emanuel’s godfathers, 1714). Meanwhile Maria Barbara was to stand for the daughters of a Weimar court trumpeter and of Sebastian’s pupil J. G. Ziegler. The range of godparents for Bach’s own children, three each, implies an active circle of acquaintance in the town and the province, whatever rôle godparents played in practice.

Although by 1708 Maria Barbara had her own box-pew in the nave of the duke’s chapel, behind the capellmeister’s wife, the family may also have had a pew in the town church where its baptisms took place, from Catharina Dorothea (1708) onwards, including those of Friedemann and Emanuel. A close connection between Weimar’s two organists Bach and Walther can be guessed from the interest they shared in certain techniques of composition, in which Bach’s greater adventurousness is self-evident. One interest was in using little patterns of notes to embellish the chorales, the so-called figurae in use for decades by German organists and now systematically applied by both composers. Another was working on ways to set Advent and Christmas melodies in canon, the extant examples of which even suggest a rivalry between the two composers. They also shared an interest in certain French keyboard music, both of them making copies of Dieupart’s Suittes de Clavessin and de Grigny’s Livre d’orgue. Doubtless less self-reliant than his cousin, Walther seems to have been a much more prolific copyist of other people’s organ music than Bach – though one cannot be totally sure of this – and preserved a great deal of it.

A big question is whether Bach, in his years as a regular organist before 1717, copied the work of other organists more than he is known from surviving manuscripts to have done. Even his own copious output as a composer cannot have covered every requirement, so the possibilities are that he improvised or repeated a good deal, and that he copied other composers’ music, German or French, more than we know. If he did not, was it because he rejected it? If he did and used it in chapel (something he was unlikely to do with the French music he know), did it disappear along with other chapel music when eventually he left Weimar, probably in some haste?

On the chapel itself, see below, p. 118.

Early years in Weimar

Very little is documented of the first five years at Weimar, and a fair assumption is that they, Bach’s mid-twenties, were very much taken up with keyboard activities: playing, teaching, composing, inaugurating (a small organ in nearby Taubach, 1710), involving harpsichord as much as organ, if not more. It would be strange if Taubach was the only organ he inaugurated at that period. But when the Obituary goes on to say that here in Weimar he also wrote most of his pieces for organ at least two questions arise. Did Emanuel overstate, affected by his father’s spoken enthusiasm for what had been his last position as a regular organist? Or did he know (or think) that his major works apparently written later derived from versions already composed in Weimar? Both could be the case, the second more significant musically.

For any composer, increasing familiarity with foreign keyboard music leavens the standard fare of local consumption, and Bach can have been no exception in this respect. All the more unfortunate, therefore, is that the earliest known choral-ensemble music of his Weimar period, i.e. the second and possibly third election cantata for his old community in Mühlhausen, has not survived, since it might show how far ‘beyond Buxtehude’ he was beginning to move by now. Whether a visit to the Gotha Court at some point in 1711 (Dok. V, p. 273) was for more than advising on the organ is not known, but if the outstanding ‘Hunt Cantata’, BWV 208, was performed for the Duke of Weissenfels’s birthday in 1713, not only had musical understanding grown exponentially in five years but his renown had already spread, apparently beyond Thuringia. This is suggested by one pupil, P. D. Kräuter, coming from far-away Augsburg in the same year to study with him. (For BWV 208 and Kräuter, see further below.)

There must be several reasons why the Obituary emphasizes the organ, almost to the detriment of any vocal works there were, few or none of which Emanuel knew unless they were revised later. For one thing, by the 1750s the music for organ was in wider circulation than most other music of Bach including the cantatas and Passions, and by then the harpsichord, chamber and vocal music of 1715 had been largely superseded. But the very uniqueness of the bigger organ pieces impressed itself on writers whose own period produced nothing comparable. The result is that the Obituary authors knew little of what was composed at Weimar. When it was composed is another big question: much organ and harpsichord music is dated today on the basis of how far it appears to have developed beyond provincial compositions of c. 1700. But such reasoning could be circular for a composer who accumulated and rethought.

By ‘His gracious lord’, it is not clear which of the two Weimar dukes Emanuel is referring to, Wilhelm Ernst the senior or Ernst August the junior (his nephew), and perhaps he was uncertain himself. The former was the controlling authority of funds and personnel, as events were to prove, but in 1775 Emanuel cited the latter as having particularly supported his father (Dok. III, p. 289). It was also Ernst August’s own father who had employed Bach for a time in 1703, as part of building up a musical establishment, and it was his own younger half-brother Prince Johann Ernst whose string concertos are found among Bach’s transcriptions. But whichever duke it was that Bach impressed, Emanuel is silent on one intractable problem of life in Weimar: the two dukes lived in such mutual enmity and territorial rivalry as would inevitably involve the musicians one way or another, leaving those favoured by one to be discriminated against by the other. There are few signs that the atmosphere was easy.

The situation gives some idea of life in an absolutist ruler’s court. Shortly before Bach’s arrival, the senior duke had decreed that his cappella musicians were not to play in the junior duke’s residence without his permission, on pain of a fine and incarceration (Glöckner Reference Glöckner, Hoffmann and Schneiderheinze1988, p. 137). The decree was re-issued two years after Bach’s departure, presumably because it had been defied – by Bach himself? – and by then the conditions were stricter: the musicians were not even allowed to discuss the matter. In response, the junior duke tried to compel them to choose to be his retainers as well, failing which they would forfeit any payments they (like Bach) had had from him, a forfeit which he would pursue through their children and children’s children. Although by then Bach was not involved, such nasty conditions throw some light on another, much later event, namely, his own incarceration in 1717 and release a month later. They also raise questions about other contacts he had made in 1713, at Weissenfels (probably a visit) and later the same year at Halle (certainly a visit), this latter for the vacant organist’s post. Five years after appointment at Weimar, he was looking to leave?

If, at about the time he became court musician to one or other Weimar duke, Bach was producing the organ Passacaglia, not only was his ability to create harmonic tension maturing but so must have been his organ-playing. So they were if, similarly, he was soon to begin to create the types of chorale now familiar from the album later called Orgelbüchlein (see below). On such ‘ifs’ hang a whole interpretation of the composer’s development, not least as a player. Specifically, the technical demands of many organ works are taxing beyond norms of the day, setting him on the path to the Six Sonatas for Organ twenty years later, and justifying the Obituary’s remark at the top of this section. While the Orgelbüchlein chorales and the later sonatas are both remarkably succinct, interesting work was also done towards creating much longer pieces, as with the Passacaglia. Spacious treatment, even when the pre-laid harmonic plan of a chorale is treated at great length, can result in more than mere length: the long Variation 10 of the chorale variations ‘Sei gegrüsset’, BWV 768, already achieves a new and convincing coherence. For Bach, success in creating both unusually succinct and unusually extended music remained a recurrent aim: he could do both length and brevity.

When engaged as organ-examiner, at Arnstadt in 1703, Langewiesen in 1706, Mühlhausen in 1709 (?), Taubach 1710, Halle 1716, Erfurt Augustinerkirche 1716, Leipzig University Church 1717 (the last three major organs), Bach may customarily have played a public concert of his grander works, various praeludia or Fantasien and longer chorale-settings. (At Halle, it is not known which of the organists present played before or after the sermon when the special new organ was inaugurated, whether it was one of the three examiners including Bach or the newly appointed organist Kirchhoff. Perhaps all four (Dok. II, p. 60).) A probable reason why the dating and purpose of Bach’s bigger organ works are such guesswork is that some were portfolio works selected and revised as such occasions required or as studied by qualified students. ‘Occasions’ would include inaugurations and demonstrations as well as lessons, and it can be misleading to pin down the dates of so many keyboard works of Bach: the works themselves were not fixtures even when gathered into a collection, as gradually happened now and then.

The Obituary’s emphasis here on the organ is one-sided. Given that Bach had virtuoso ability as an organist and explored certain kinds of organ music farther than any predecessor, and given that he continued to do so even as cantor and no longer a regular organist, it is still puzzling that neither part of the Obituary, Emanuel’s or Agricola’s, acknowledges the strides he also took in chamber and harpsichord music during the Weimar years. At Leipzig, did he never speak of having taken a leading part in the duke’s concert life? Learning ‘how to treat the organ’ sounds like the composer’s own phrase when later describing his priorities at that period. It implies that he was also imitating other kinds of music on the organ, composing chorales like chamber trios and fugues like concertos. Emanuel would know the extraordinary range between a chorale of 8 bars in the Orgelbüchlein (BWV 631) and a big toccata of more than 400 (the F major, BWV 540), and would find in works such as these a clear demonstration of ‘how to treat the organ’.

With the duke’s approval and at his cost, the chapel organ was being improved and enlarged over several years, and was in and out of commission over 1707–9 and again from June 1712 to May 1714 (Schrammek Reference Schrammek, Hoffmann and Schneiderheinze1988). The instrument could never become very grand, placed where it was, and its sound must always have been somewhat indirect in the chapel, if not actually dull or indistinct. For it was located at the back of an attic chamber above the chapel ceiling through which a balustraded opening of 4 by 3 metres admitted sight and sound, 20 metres above the chapel floor. Through this opening the organ, just visible, spoke down into the rectangular chapel below, as if from on high. For its services, the chapel, which was some 30 metres long and 12 wide, was occupied on the ground floor and two running galleries not by a parish-congregation but by court personnel, who looked towards the liturgical east end at an altar structure that was itself not unlike a stage-set or scena. This consisted of (in vertical order) step, altar rail, altar table, baldacchino, pulpit, a decorated obelisk pointing up to ‘heaven’s castle’, then the attic chamber balustrade. This chamber had a narrow space for performers and some way at the back of it the organ, with a ‘heavenly’ fresco on the plaster dome above. Although documentary reference to certain seats built in this attic conveys a picture of singers and instrumentalists stationed there during services or rehearsals, whether they always performed from on high in this way and more or less out of sight is open to doubt.

The organ, glimpsed at the top of the scena or ‘path to heaven’ (Weg zur Himmelsburg), had its back against the wall, with bass pipes and bellows-chamber at the very back. In 1658 it had one manual only, which probably was all that could be accommodated comfortably in a restricted space; but then a second was added, its pipe-chest placed to the side and played presumably with a complicated action that might never have worked very well. Shortly before Bach was appointed, an improvement was made by relocating this side-chest under the main pipe-chest. How room was made for this is hard to guess, but it probably made further work of improvement inevitable; this was done during Bach’s term of office and resulted in an enlarged chamber being constructed behind, presumably at his urging. A row of tuned bells was also made (probably positioned just behind the music-desk and played from the manual), and the whole organ now comprised about two dozen stops: not a great inspiration, barely more than adequate for realizing new ideas in organ music, but useful none the less. Quite why Bach thought, as it seems he did, that a row of bells was necessary (acquired at some cost from Nuremberg) is not explained, and one can only assume that they sounded at certain jolly moments, during chorales at Christmas and other festive times. An organist writing in 1742 specifically mentions the young Bach having a Glockenspiel or row of bells installed earlier at Mühlhausen (Dok. II, p. 405).

The duke may have ‘fired him up’ (feuerte ihn an) but many a court organist strove to please, and in Bach’s case there must also have been an extraordinary creative urge and practical ability matched by curiosity and industry. Fortunately for this new range of musical styles, it seems that his predecessor Effler preferred a more modern organ-tuning than elsewhere (BJ 2004, pp. 160–1), which might explain the appearance of the keys F minor and E flat major for chorales in the Orgelbüchlein. The tangible result of the duke’s support was that the organ music produced by Bach surveys a range of styles wider and on a bigger scale than had ever been achieved before by any organist in any European tradition, offering every subsequent composer a model to emulate as best he can. The Obituary authors surely realized this. After all, there is an important possibility to bear in mind here: that the first music Bach developed beyond precedents was not so much that for voices, strings or harpsichord but for the church organ.

The Obituary’s gracious acknowledgement of the duke’s support may have had another purpose: it was of a kind that Bach was not to receive later in Leipzig. But what purpose his organ music had at Weimar is not as clear as is often assumed, whether for voluntaries before and after the service or as items for private ducal concerts. Or as brilliant music shown off by a star employee for the duke’s noble visitors? A ruler of such known piety as the Duke of Weimar might well take pleasure in special organ music being played in his chapel by a gifted and well-paid employee, although compared to other court chapels in that part of the world (Eisenberg, Saalfeld, Sangerhausen, Weissenfels) the Weimar chapel and its organ were not the most spectacular. Whether the organist played after the services as the congregation retired is uncertain, but if he did, there was a choice of toccatas, fugues, concerto- transcriptions and improvisations as well as exceptional mixed genres such as a grand ritornello Fantasia on the Whitsuntide hymn, BWV 651a. By now pieces such as these had become genuine organ works, no longer suitable for other keyboard instruments.

To what extent the works as we know them represent the composer’s improvisations is not clear from the Obituary’s phrase ‘everything possible in the art of how to treat the organ’. Certainly Bach was expanding the repertory by absorbing other kinds of music such as sonatas and concertos, as he was also expanding styles in the cantatas, and such music would come about only after thoughtful deliberation. But how he improvised can be only hesitantly traced from surviving examples: first, in fantasias and toccatas, with their various runs, arpeggios, stopping and starting, clear cadences, distant modulations, moments of free recitative, etc; and secondly, in fugues, with moments of contrapuntal imitation interspersing freer episodes based as often as not on broken chords. A work such as the ubiquitous ‘Toccata and Fugue in D minor for Organ’, BWV 565, might give an idea of what some organists throughout the eighteenth century could improvise with thin harmonies, a few rhetorical gestures, dramatic pauses, simple shape, much repetition and virtually no counterpoint. But far too much is doubtful about the authenticity of this piece for it to have anything reliable to say about the young organist J. S. Bach.

Although it is clear that Bach did develop organ toccatas into big, subtle, fully worked concerto-like movements (C major, BWV 564.i; F major, BWV 540.i), the steps in this evolution are also hard to trace. As already suggested, early compositions or creative adaptations such as the Albinoni fugues are as likely to have been revised, rewritten and transmitted in various states; and this is true of some of the later big preludes and fugues. How unusual Bach’s practice was in this aspect is impossible to know, since no comparable fund of sources exists for any other composer. But not all versions now known are likely to go back in all their details to Bach himself, and few are likely to be the only ones that ever existed. The ‘Fantasia in G major’, BWV 572, could represent the kind of music that Bach improvised in his late twenties or early thirties, indeed for the delight of his duke in Weimar, either on organ or on harpsichord, and attaching it to one or more other movements. But many works are so original that there are few clear criteria for dating them: the G major Praeludium, BWV 541, which begins like an improvisation but couples two highly organized ritornello movements, could have originated many years before the composer made a copy of it in the 1730s, with or without its known fugue. Evolution of a genre is not reliably traced when it is as isolated as Bach’s ‘free’ organ works are, individually structured like string concertos but otherwise unlike them and unlike each other.

A down-to-earth question is how often Bach paired a prelude with a fugue to make a complementary pair, either on paper or in performance. This was almost certainly less often than supposed by anyone who relies on later editions or who has the completed WTC in mind, as all subsequent editors and players of the music have. But the earliest known versions of many WTC preludes did not accompany fugues either, and not even in WTC did the composer actually label them Prelude & Fugue (see p. 237). Nevertheless, once the idea of a pair became familiar from publications by J. K. F. Fischer and others, pairing must have seemed natural, particularly when the prelude was short and preceded a substantial fugue, as it did not always do. Prelude-and-fugue pairing is also there in vocal works early (Cantata No. 131) and late (Kyrie of the ‘B minor Mass’), as it is in the harpsichord works of Clavierübung I, II and ‘IV’. A fugue does not have to be ‘bigger’ than its prelude – it rarely is later in the WTC1 and WTC2 – and it could be that some couplings of doubtful authenticity result from someone else attaching a fugue to a free-standing prelude.

That Bach was constantly, so to speak endlessly, curious is suggested by music of other kinds he became acquainted with during his earlier Weimar years, making a most important item in his biography and output. Concertos of Albinoni (Op. 2, published 1700) and Telemann (Concerto in G, copy c. 1709) are certain, but often over his middle years one can only infer what he knew, such as one or other edition of Corelli’s violin sonatas, Op. V. While these are a worthy model for any musician, the musical quality of some other items varies, suggesting that Bach’s self-teaching was serendipitous and eclectic. He came to own copies of at least three important organ publications originating far from Weimar: the volumes of Ammerbach’s Tabulaturbuch (Leipzig, 1571), Frescobaldi’s Fiori musicali (Venice, 1635) and his manuscript of de Grigny’s Livre d’orgue (Paris, c. 1700). He appears to have possessed at least three exemplars, possibly more, of the Ammerbach, either paying homage to a Leipzig predecessor or investing in them to sell to visitors. Their antiquarian value was high, being the first printed volume of keyboard music in Germany, an example of tablature, published by the then organist of St Thomas. A note, probably autograph, on one copy said that it cost one gold louis d’or (Dok. I, p. 269), a very large sum. By the eighteenth century, however, Ammerbach’s actual musical influence had dwindled.

As for Frescobaldi, discriminating composers must always have recognized his quality. In Dresden, J. D. Zelenka also had a manuscript copy of the Fiori musicali dated 1718 (Beisswenger Reference Beisswenger1992, p. 285), while in Hamburg Reinken is known to have been familiar with a book of Frescobaldi toccatas. Exactly how Bach got to know the de Grigny and (two or three years later) the Frescobaldi publications is a guess, probably through J. G. Walther who habitually corresponded with other musicians. His own copies of the de Grigny and Dieupart books were perhaps a little later than Bach’s but surely suggest a shared interest. While Frescobaldi continued to influence Bach’s counterpoint into his maturity, de Grigny’s manner of writing for the organ is less in evidence. Did copying not lead him to grasp the beauty of the French styles, especially the lyrical solos for the left hand – the en taille solos? These can be occasionally glimpsed, as in the Weimar organ-chorale BWV 663a, but Bach seems never to have composed fully in this way, perhaps finding its peculiar lyricism too indulgent for Lutheran chorales, Lutheran organists, Lutheran congregations. One thing seems likely: the exceptionally high musical quality of Frescobaldi’s and de Grigny’s books was recognized by Bach and they were even more carefully preserved than usual.

Some full-length organ-chorales from these years took in an unusual range of styles, confirming that Bach, though a Lutheran organist, did absorb much non-Lutheran music that he came across in Weimar. De Grigny, for instance, gave him practical examples of writing richer harmony by adding a fifth part to the usual four, something found then in certain chorales and cantatas. Dance-types were also adapted for chorales, either because they were familiar to German organists already or because (in Bach’s case) André Raison referred to them in his Livre d’orgue of 1688. The very length of one Weimar chorale, ‘Schmücke dich’, BWV 654a, a work fittingly enthused over by Schumann a century later, enables the composer to amalgamate two very different genres: a true-to-life sarabande and an organist’s typical setting of the hymn line by line according to tradition. The result is continuous and tuneful. A gigue, on the other hand, has to be tempered when the chorale is for Communion (BWV 666a) but can be extrovert for Whit Sunday (the astonishing ‘Komm, Gott Schöpfer’, BWV 667a). The large scale of such organ-chorales, even before later revision in the Leipzig years, is also a feature of the recently discovered chorale BWV 1128, as if it were Bach’s ‘answer’ to the kind of long fantasia beloved by organists in the north.

If it is less certain than usually thought quite how the organ-chorales were actually used, how much more is this so for the harpsichord toccatas! Although five or so of these are now known in disparate copies, the Obituary lists Six Toccatas as if speaking of a fixed set. (A lost set known to Emanuel and intended for publication?) Copies of two of them, BWV 913 and 914, had already been made by 1708 or 1709, so shortly after the move to Weimar, and possibly for or by a new student. The Toccatas’ sections appear to follow the structures of earlier German organ toccatas, familiar in the Weimar chapel perhaps, and in which free virtuoso passages are interspersed between sections more in the style of sonata movements. Each section is characteristically ‘thorough’, exploring its themes with felicitous harmony and taking an evident pleasure in perfect cadences. If the counterpoint is still somewhat hidebound, there are flashes of melody elsewhere in which one might fancy the influence of Georg Böhm. Despite the traditional repetitions and short phrases in these toccatas, there are increasing signs of the maturing composer’s grasp of sustained length, of harmonic movement and of what works well on the keyboard.

Cantatas in Weimar, 1

Tracing step by step Bach’s experience with vocal works before his promotion at Weimar in March 1714 (see p. 155) is not possible, though there are some pointers. In Cantata No. 196 (see above, p. 67) the young composer had not gone much beyond imitating choral works by Weckmann and others he could have heard in Hamburg or Lübeck. But this might reflect the fact that a display of old-fashioned permutation counterpoint is not inappropriate for the text of No. 196, a formal Old Testament benediction from Psalm 115; see Example 7 on p. 131. Not for the only time, guessing which of two accomplished but disparate Bach works was the earlier (the Passacaglia or Cantata No. 196) cannot be based on simple comparison: they are too different in genre and purpose.

An area explored during at least the later Weimar years was the copies Bach made (or had made) of choral works, including Latin works, chiefly old-style Kyries, by Roman Catholic composers not now well known, such as J. Baal, F. B. Conti, M. G. Peranda, J. C. Pez (from the Missa San Lamberti, Augsburg, 1706) and J. C. Schmidt. Whether these were for private study or for occasional performance in the Weimar chapel is not clear from documents, although some separate performing parts make it possible that Kyries (often sung in major Lutheran churches) were intended for it. Bach and Walther shared an interest in Palestrina and the stile antico (qv) of old Latin vocal music, but again, whether the known copies of Palestrina’s original printed parts were made in order for singing in a service (unlikely), in informal performances (possibly) or in contrapuntal studies (probably) is not established. Judging by the accidentals consistently added at some stage to Peranda’s and particularly Palestrina’s modal lines, someone was interested in modernizing this music for modern performance, either in Weimar or in Weissenfels. At this period in Weissenfels, J. P. Krieger had a collection of the very works known in Weimar (BJ 2013, p. 85).

An assumption has been that Bach made use of the Weimar Palestrina copies later for performance in Leipzig. But used in this way or not, such vocal music was to be studied for much the same reason that it was studied in twentieth-century British universities: as a model for contrapuntal work, to learn the nature of intervals, to absorb the art of good part-writing and so forth. Its usefulness as a foundation to such a composer of keyboard music as Bach is also clear from many a fugue of his over the years to come. On the other hand, the cantatas from the Weimar years rarely allude patently to this old contrapuntal style, just occasionally (a countersubject in Cantata No. 54.iii). Characteristics of the style emerge much more often in the later cantatas for the Thomaskirche, Leipzig, suggesting that earlier in Weimar, for the solo singers of the duke’s chapel or because the dukes had other tastes, Bach was deliberately turning away there from the stile antico.

How unusual such copying of imported choral music was is difficult to know, for some of it survives because it has Bach’s name or writing on it. But other acquaintances also copied out French keyboard music. The young prince himself was to compose concertos imitating Vivaldi with a musical competence rare, perhaps unique, for a teenage nobleman. In the same way, one can assume that many details in the arias and recitatives in the Weimar cantatas were due to exposure to other kinds of up-to-date music in a cultured court, even the court of a duke anxious to observe his religious obligations by standardizing the duchy’s hymnbooks. This he did by means of the published Weimar Gesangbuch of 1713. In March 1714, the very month Bach became concertmeister and began a series of cantatas, the chapel accounts include the cost of five blackboards on which the hymn-numbers were to be announced (Jauernig Reference Jauernig, Besseler and Kraft1950, p. 71). The two genres, chorales sung by the congregation and chorales sung by professional soloists, remained quite separate in their musical language.

Bach’s original position of ‘chamber musician’ indicated duties in the general music-making at court and as member of a cappella of fourteen musicians, to whom were added a pool of seven trumpeters and timpanist as occasion required. Whether duties included playing harpsichord solos in court concerts is not documented, nor indeed whether solo harpsichord-playing was known at all to audiences bigger than a select few. Nor is it known precisely what the ensemble works were in which he presumably did participate: solo instrumental sonatas (playing continuo with violin, gamba, recorder or oboe soloists), string trios, mixed trios and broken consort music for wind, strings and keyboard (those works the Italians called concerti). One can suppose that it would have taken some years for all these activities to flourish.

A particular unanswered question is whether the set of parts Bach himself seems to have copied for an anonymous setting of the St Mark Passion a few years into his Weimar job (a work attributed variously to the northerners R. Keiser and F. N. Brauns) was made for performance in the chapel, or indeed anywhere else. If this had happened or was meant to happen, it would have probably been his largest performance to date, though in a chamber setting for a few performers: SATB, two violins, two violas, oboe and harpsichord. There were some new movements added (recitative, a prelude and fugue for instruments, two bicinia), the work of a still-young composer using and adding to music from elsewhere, interested in ways to write for voices and instruments and making a new, whole piece for local performance. This was a procedure also familiar to Handel. Bach’s copy is the only surviving manuscript of the work, which he revived twice in Leipzig, and is an example of what could well have been a common activity for him in Weimar.

In the chapel cantatas, and perhaps on other occasions, the six male singers and a handful of string- and wind-players made it possible to create various ensemble combinations, especially when other soloists joined them. This was an enthralling opportunity for a young musician and composer of cantatas. Familiar with the old tradition for mixed consorts, Bach was gradually able to create newer works in a variety of instrumental combinations, fresh springs from which the mighty stream of ‘Brandenburg Concertos’ was to issue later. Particularly after his promotion in 1714, a big ensemble of strings, wind and brass could rejoice in a Sonata for an Easter Day in Cantata No. 31 or introduce the cantata for Christmas Day in No. 63; other ensembles include a five-part ‘French’ string band for an Advent Ouverture in Cantata No. 61, an expressive string and woodwind Sinfonia for a post-Christmas cantata in Cantata No. 152 and a delicate violin solo in the Sonata for the Annunciation in Cantata No. 182. The four solo violas in Cantata No. 18 have an essentially old-fashioned sound and texture, but the work also includes an example of the new recitativo secco (qv), and does so with little sign that Bach was experimenting with something novel to him, except perhaps in the shortness of the phrases. Example 5 already includes both a characteristic change of pace and, because of this, draws attention to important words (‘falls’, ‘fructify’, ‘Word’).

Example 5 Cantata No. 18.ii: ‘Just as the rain and snow fall from heaven, and do not persist but fructify the earth, making it fertile and increasing so that it gives seeds to sow and bread to eat, so should the Word …’

Telemann’s setting of this same cantata text by Erdmann Neumeister is typically bland in its harmony and gesture, emphasizing what a colourful drama Bach brought to the music for the Weimar liturgy. This becomes clearer when on later occasions in Leipzig, Bach writes a cantata as if imitating the simpler harmony and phraseology of the ever-popular Telemann, for example Cantata No. 47 (1726). There is in the Weimar cantatas a lightness of touch in melody and timbre that remained distinct from the later works for Leipzig, and of course from the work of contemporary composers. At the same time, writing recitative with continuo in a cantata was itself a way of reconciling ‘church sounds’ with ‘chamber sounds’, as it had also been for a generation or so of German composers. It is well to remember too that one cannot be sure that such colourful cantatas were performed only in chapel or only as part of its services. Never anywhere else?

Although the intimate understanding of French manière evident in the opening ouverture et fugue of the Advent Cantata No. 61 (1714) may appear to have little to do with Parisian organ music, the way it paraphrases the Gregorian chorale-melody is as French as it is German. De Grigny too, in his book copied by Bach, had paraphrased the original hymn melodies in this way in his settings, though doing so with a degree of arbitrariness in his harmonies (why this, why that, what is the key?) unknown to the orderly Bach. There are other touches of French style in Cantata No. 61, adopted almost certainly as a means of matching the theatrical aura of the text it uses, i.e. one of Pastor Neumeister’s, associated also with Telemann’s ‘French annual cantata-cycle’ of 1714/15. Thus might a cantata – here from 1714, but also others before and after – gather together all and any kinds of musical idiom, not only for the Greater Glory but for that mixing of styles to which the composer was not alone in being attracted throughout his life. Doubtless deliberately, the first movement of a later Advent cantata, No. 62 (1724) paraphrases the same hymn-tune but now in the very different style, not a French ouverture but an Italian ritornello concerto.

Whatever their musical styles, the works being produced in Weimar are responding in a lively fashion to the newer poetry as celebrated in Erdmann Neumeister’s libretti, of which there were hundreds, and gathered as yearly cycles from 1704 onwards. Neumeister’s original title of 1700, Geistliche Cantaten statt einer Kirchen-Music (‘spiritual cantatas in the place of a church anthem’), uses the word cantata in order to allude directly to fashionable Italian chamber cantatas and their sequence of arias (new poetry) and recitatives (new or old prose), now in ‘spiritual’ form. Alien to the Italian conception of cantatas, however, were the additional movements, the instrumental overture, the big choral movement and the final chorale. So of course were any chorale-melodies introduced at any point within a cantata. Such cantata-texts, often published and treated as self-contained poetry, could be realized as recitatives, arias and choruses (these with biblical sentences), and an appropriate chorale could round it off even if the poet’s text did not call for one.

In general, the new texts would naturally encourage a ‘theatrical’ style of music at a certain moment in services, introducing conspicuously new sounds and words, and it is not surprising that court chapels such as Meiningen and Weissenfels soon made use of such texts. At Weissenfels, the composer J. P. Krieger may have set Neumeister’s volume complete. But already in 1709–10 such works were being criticized as theatralische Music by none other than the Leipzig cantor Kuhnau (DDT 58–9, p. xlii). Deliberately or not, Kuhnau was recognizing the difference between court chapels and parish churches: as Bach was to find, a parish church was sensitive to theatrical music. Pastor Neumeister supplied ready examples of these so-called madrigalian texts for the composer of any court or, eventually, of any parish church, and the idea was given impetus by what was acceptable in royal cosmopolitan Dresden. Texts from 1710 onwards that mixed biblical words and chorales with the new poetry were especially acceptable. At Weimar in 1709, a member of the court clergy published a Passion text based on St Matthew, ‘with intermixed devotional chorales and arias’, and it is possible that at least in the town church some such work was performed over the following years (BJ 2006, pp. 45f.).

From now on Bach seems often to have chosen Neumeister texts and, since he set far fewer of them than did Telemann or Fasch, those from other poets such as the court secretary, Salomon Franck, who had a similar way of incorporating biblical words in his poetry. Franck’s Cantata No. 172 for Whit Sunday 1714 draws from Bach an alert series of musical affects: first a rejoicing (big orchestra, catchy rhythms) then a quieter recitative, a tenor aria in the minor interspersed with a forthright affirmation of the Trinity (this with trumpets and bass voice), an exhortatory aria (soothing soprano) and finally a fully orchestrated verse from an Annunciation chorale. While such music lends itself to a sermon-like interpretation encouraged, so one might guess, by the composer himself, the sheer charm of so many melodies in Weimar cantatas should not be missed, as in Nos. 132.i and 63.v. Nor should the inventive timbres colouring the effortless counterpoint, as in the viola d’amore in No. 152.i and two solo cellos in No. 163.iii. These cantatas are not from the earlier Weimar years but follow on the composer’s promotion (pp. 155f.). However, that this was less of a watershed moment, a turning point in his maturity, than it has sometimes been taken to be, seems clear from the sheer scope of the fifteen movements of the ‘Hunting Cantata’, No. 208, from a year or more earlier.

Musical development: counterpoint, variations, concertos

Bach’s musical experiences in Weimar were much wider than is implied either by the Obituary’s emphasis on him as a virtuoso organist or by any other extant documentation. A still-open question is what music visiting artists and companies brought to Weimar, including excerpts from French or Italian operas. There were also his other contacts with neighbouring courts, such as at Gotha in 1711 when he was a ‘guest player’ (Dok. V, p. 273) and then, on returning to Weimar, a successful applicant already for a salary increase – not a coincidence, probably.

It was no doubt from private study rather than court-visiting that a fundamental technique in Bach’s maturing compositions for instruments or voices emerged and bore fruit: his mastery of invertible counterpoint, well beyond the examples offered by any theory-book. For him, this generated virtually any kind of music, from the earliest organ-chorales to the later Inventions (qv), from a Mühlhausen cantata to the ‘B minor Mass’. With this mastery, different themes or melodic lines are combined or exchanged with a bass-line (which is melodious in its own way) to produce music for virtually any wished-for Affekt, including the lightest and brightest. See Example 6(a), recomposed as Example 6(b). Even a canon can be made to work if other such lines are added so as to convince the ear that all is well, as in Example 6(c). Examples 6(a) and 6(c) belong to much the same period (1712–13) and are quite typical. An advantage of invertible counterpoint is that at a stroke the lines become available for re-using in different combinations, and the different keys have their own character. In such ways any initial effort that is required to create the invertibility turns out to be an economical way of generating movements and one that is practically infinite.

Example 6 (a) Cantata No. 208 (1713?).xiii, b. 5; (b) Cantata No. 68 (1725).ii, b. 5; (c) Organ-chorale, BWV 600, b. 9

Typical of the music for choirs had been a counterpoint based on the permutation principle. Bach’s look very like the best achievements of previous composers and give the impression of being ‘systematic’ or ‘calculated’ (see Example 7). This is an old way of creating vocal counterpoint, and while in this instance the lines are still melodious despite the ingenuity, there can appear many short phrases and a certain repetitiousness that arises in the service of the text. In the organ Passacaglia’s fugue, however, the short phrases such as those found in Cantata No. 196 (Example 7) have disappeared, despite this being a permutation fugue. Here, new counterpoint emerges each time and is unpredictable.

Example 7 Cantata No. 196.i, b. 18. Text: ‘He blesses the House of Israel, the House of Aaron’

A less pervasive technique than invertible counterpoint or fugue but still important was canon. The German organist’s custom of making canons at the octave (qv) from the melodies of Advent or Christmas hymns will clearly be challenged more severely by some melodies than others. Example 6(c), ‘Gottes Sohn ist kommen’, required ingenuity if the canon was to work, i.e. to make us think that the lines occurred naturally. Walther’s setting of this melody eases the problem by leaving it longer before the canonic answer, resulting in a movement nearly twice as long as Bach’s but therefore ‘less canonic’. The canon in another Weimar organ-chorale, ‘Hilf Gott, dass mir’s gelinge’, BWV 624, needs to have its answers at varying intervals if it is to work, whereas in the Ten Commandments setting (BWV 635) there are several ‘simultaneous’ canons as the melody accompanies itself. The painstaking thought (and imagination) required for such music is obvious.

Certain practices and techniques, such as creating a set of variations on a hymn-tune by exploring a different motif in each variation, barely survived into the Weimar period, and one can understand why. Such sets, generally called ‘Chorale Partitas’ today (though not with any certainty by Bach himself), seem to have been a speciality of Thuringian organists for a few decades around 1700, if only seldom beyond the second decade. In 1802, Forkel’s claim that Bach found writing sets of variations a ‘thankless task’ (1802, p. 52) arose because he (Forkel) knew that in most cases they merely reiterated the harmony by superficially decorating it. One sees this clearly in the ‘Chorale Partitas’, BWV 766–8 and 770, and perhaps Emanuel or Friedemann had heard the composer speak of the ‘thankless task’. But Forkel could have concluded this because there are so few examples of ordinary variations in Bach’s worklist compared to Handel’s or indeed Mozart’s, with which by then Forkel was also familiar. The remark does not quite do justice, however, to the way a chorale might be varied, reharmonized several times in a cantata or a Passion (as was the case): harmony can be imaginatively explored under a melody as much as the melody can be decorated. More to the point, perhaps, is that any such ‘partitas’ reliably attributed to Bach confirm that his early activity as a composer was typical of a Thuringian organist of the time.

All the same, Bach can be imagined to have reacted against the simplistic variations he came across in his youth, works that continued to influence Handel in his variations. One notes that the four big, outstanding and indeed unique variation-works that Bach did compose in the course of over thirty years – Cantata No. 4 for voices, the Passacaille for organ, the Ciaccona for violin, the Aria mit Veränderungen (‘Goldberg Variations’) for harpsichord – do everything but reiterate the harmony in some simplistic way. A similar point could be made about those Leipzig cantatas that are based, in rather different ways, on seven-verse chorales (see p. 283). Three late works, Musical Offering, Art of Fugue and ‘Vom Himmel hoch’, take the variation principle a step further by using the stated theme not for reiterating or varying the harmony but as a point of reference, one way or another, for an array of complex canons.

For the picture he is drawing of a serious composer, Emanuel described the Weimar job in terms of organ music and, after the promotion in 1714, of ‘mainly church pieces’, as he called them. And yet of huge importance to Bach, something even changing the direction much of his music was to take, was something Emanuel never mentions: sudden acquaintance in 1713 with a group of new string concertos from Venice. Whether or not this was as simple as evidence now suggests, the vivid and seductive effect of these spectacular pieces, perhaps glimpsed already in isolated examples previously making their way north, can be imagined: a voracious, energetic composer in his late twenties suddenly gets to know Vivaldi’s Op. 3, L’Estro armonico, and Op. 4, La Stravaganza. A revelation! Did Emanuel not appreciate this or did it detract too much from his picture of the self-made German master?

The closer Bach had continued to keep to tradition in his organ and harpsichord music, the more startling must have been the ‘Vivaldian effect’ in other kinds of music. When, probably in July 1713, the young Prince Johann Ernst returned to Weimar from a stay in Holland with copies of music by Vivaldi and other Italians (Corelli? Frescobaldi?), the prints of string concertos he brought took the form of part-books: not scores but sets of playing parts. So one is left to imagine the effect on players picking up these parts and trying them out; or on the transcriber as the concertos emerged in score and revealed their effects. Soon after the prince’s return, one imagines those musicians concerned (the prince, his half-brother, Walther, Bach, some string-players) gathering to play the concertos in the junior duke’s residence, the Red Palace or Rotes Schloß with or without the senior duke’s approval.

There could be various reasons why one Weimar organist, Walther, seems to have worked more with the older generation of Italian composers such as Torelli, while another, Bach, worked increasingly with the younger such as Vivaldi: personal preference, accident of sources, availability, even reciprocal agreement between the two organists. In any case, the sheer quality of Vivaldi’s L’estro armonico concertos is surely the reason for Bach’s transcribing at least half of them, either at that time or later: three for harpsichord alone, two for organ alone, and the unique one later for four harpsichords and strings. Two more each from Op. 4 and Op. 7 (1716, one for organ) must be a sign of someone’s continuing enthusiasm for such transcriptions, Bach’s or perhaps a student’s. Transcribing part by part was laborious, and it must have been difficult to resist modifying in some way what was being transcribed.

Whether Bach learnt from Prince Johann Ernst that at this period in the Protestant Netherlands people could hear such Italian concertos being played in certain public organ recitals is not known, nor therefore whether he or anyone else had the idea of producing concerts of this kind in Weimar. There is no evidence either way, unless one takes the duke’s support for Bach as a hint that he did play such (private?) recitals – which is possible. Transcriptions were not the only way to assimilate Italian styles, as one sees later in the same year, 1713, when Bach’s aria BWV 1127 (apparently for the senior duke’s birthday) combined old and new: the German organist’s traditional bicinium bass has become a continuo part of a kind familiar to Italian cellists, and the old German strophic form (different verses to the same melody) is supplied with new Italian string interludes. The combination made a suitable offering to the duke, one of many offerings at a typical ducal court, no doubt – but why Bach and not the capellmeister? Was it a gratuitous, even presumptuous, act of homage?

If anything of Vivaldi’s music such as the trio sonatas Opp. 1 and 2 (1705, 1709) had previously penetrated to Weimar, the young prince would have been alerted before his trip to Holland to search for more of their kind among the Dutch publishers and booksellers. Albinoni’s three-movement concertos were being published from 1700 onwards, and signs of Vivaldi’s influence are also there in movements of Handel’s Concerto in B flat, HWV 312, now dated c. 1710 in Hanover. In Bach’s case, it is hard to imagine how at least some of the ‘Brandenburgs’ and the solo violin or harpsichord concertos would have come about without the revelations offered by Vivaldi’s Opp. 3, 4 or 7. All these have a dashing quality beyond even Corelli’s concertos, with singable melodies and clear solo sections, plus certain stylistic earmarks such as beginning in bare octaves. Any composer would be excited by Vivaldi’s way of organizing instrumental pieces by the means of repetition, note-spinning, sequences and simple harmony, all handled inventively with great rhythmic vitality.

It was probably at a period when the young prince was composing his own works in this style that Walther also arranged many Venetian concertos. Telemann got to hear of the prince’s efforts and published some of his works. Sources also attribute to Bach two transcriptions of the prince’s Concerto in C major, one for organ (BWV 595) and another, different in detail and substance, for harpsichord (BWV 984). There is a commonsense logic to the concerto shape of three movements, fast–slow–fast (or lively–slower–quicker), and it occurs as if quite naturally in very different genres. Two examples of the shape are the earlier Toccata in C major for organ, BWV 564, and the much later double-choir motet ‘Singet dem Herrn ein neues Lied’, BWV 225.

Musical development: strategy, tactics

How to shape a piece of music, to sustain length and allow a movement to develop and come to a well-paced conclusion without repeating inappropriately or continuing boringly, was clearly a question of importance to any composer. So was another: how to create the faultless miniature.

In both the small and larger-scale works of the Weimar period, melodic flair merges with a harmonic logic in which a simple common chord can sound as striking as a complex discord. Seldom if ever does harmonic control fail Bach, though he comes close to it at one moment in the long Fugue in F major, BWV 540, in a precipitous modulation from C minor to D minor (an unreliable source or a textual crux?). When a movement of carefully organized length strikes one as not very inspired, as some movements in Cantata Nos. 12 or 31 do, it could be that too much attention has been paid to conveying a text. In the case of No. 31, for example, the idea of ‘Heaven laughing’ leads to something ordinary in the melody, harmony and rhythm, while the idea of the ‘Prince of Life’ leads to a merely formulaic rhythm. Conventions can be too automatically applied, as when without any further ado chromatic intervals convey something anxious, sad, regretful or in some other way negative. As in other works of the period, Cantata No. 12 can leave an impression of ‘going through the motions’, as too does many a later cantata aria.

Cantata No. 12 also has a slow, opening melody for oboe that is more touching than anything one is likely to find elsewhere except for certain moments in Handel’s Italian cantatas. Even using standard effects like the Neapolitan sixths (qv) in an aria of the same cantata – picking ingredients off the shelf, so to speak – can also have a touching quality. It is a formula of the kind that went through changes later and gradually became rare in Bach, but the conventional appearances of it were a first, necessary step. The sheer number of Neapolitan sixths in compositions from the Weimar years and earlier (especially in certain toccatas) is presumably a sign that he liked them and found them ‘expressive’, doing so right up to the ‘B minor Mass’, both in a new movement (Kyrie) and an earlier one now transformed (Agnus dei).

Chromatics were another time-honoured formula that could automatically embellish harmonies, and one much practised by Walther and Bach. When Bach writes longer or shorter phrases that include all twelve semitones, as in a recitative of Cantata No. 167 (1724) or passages in the later A minor Prelude WTC2, one can assume it was intended, a stretching of the older chromatic motif in order either to be more expressive or to do something new. The chromatic phrase was one of a whole catalogue of the motifs or patterns of notes (figurae) taught and used by many German composers, learning from precedent rather than books. But in Weimar the year of Bach’s arrival, Walther listed and briefly described many of them in an unpublished treatise written for the same Prince Johann Ernst, the Praecepta of 1708, in which he gave them Italian names and in this way marked their origin. (Another somewhat similar treatise probably of some influence on German teachers was Mauritius Vogt’s Conclave thesauri, Prague, 1719). Bach’s ‘Chorale Partitas’ play with Walther’s motifs and explore his own versions of them, and they lead gradually, by steps not now traceable, to a unique collection of short organ-chorales in which the patterns are applied with unheard-of sophistication.

These are the chorales for the church year, in an album or collection later called the Orgelbüchlein (‘Little Organ Book’), which it is possible to view as the peak of the tradition for creating music by subtly embellishing hymn-tunes and their harmony. The melodies are harmonized in such a way as to establish a unique mood for each of the texts, ‘depicting’ the words of the hymns with old and newer motifs. Yet the Orgelbüchlein’s settings are rarely single-minded in this endeavour, especially when compared with chorale-based compositions by earlier organists such as Scheidt or Steigleder, or later by Walther and Vetter (see Example 10 on p. 160). One has the impression that in this album Bach is deliberately using the very motifs outlined by Walther in his book and adding quite a few of his own. Did the two organists discuss such techniques, vie with each other, compete in composing certain types of music?

The two men were surely aware of each other’s activities, and Walther’s music often uses straightforwardly the very patterns handled more inventively by Bach. Bach’s earliest surviving puzzle canon, BWV 1073, is dedicated to Walther, 1713, and it is not unlikely that he knew Walther’s copy of Johann Theile’s book Kunstbuch, Naumburg 1691, a treatise on ‘special music and secrets deriving from double counterpoint’. Both of them also made use of the works of G. M. Bononcini, Walther his treatise Musico prattico, and Bach his Sonata Op. 6, No. 10 (Venice, 1672) – or so it seems, for a theme from the latter appears in the so-called ‘Legrenzi Fugue’ in C minor, BWV 574. Assuming this print was his source for his treatment, Bach condenses the original movement and omits its echo-passages, a credible sign of his technique for reworking his models.

Note-patterns are like tactics in need of a strategy. In a cantata movement or prelude based on a chorale, shape is no special problem, since the chorale-melody provides it. But what of substantial pieces of music without such props? It is not obvious how far Vivaldi’s concertos helped Bach shape movements of his own, which, in the big organ preludes and fugues, are lone works anticipated by no predecessor and matched by no successor. This is uncertain despite assumptions made by Forkel in 1802, and copied ever since, that Vivaldi’s ritornello movements were models for Bach. Forkel seems not to have known of Torelli and the possibility that his concertos had also been models. But he did know that Emanuel’s colleague Quantz had previously acknowledged the impression Vivaldi and his ‘beautiful ritornelle’ had made on him, and how he had for a time taken them as a ‘good model’ in his own music (Quantz Reference Quantz and Marpurg1755, p. 205). For Forkel, if Quantz was bowled over by Vivaldi, so must his hero Bach have been.

But Bach had already written keyboard fugues which tend by nature towards a kind of ritornello form, with a theme returning periodically, after episodes and with a final statement or coda. Returns or ‘local repetition’ are not only so natural, even necessary, to the transient thing that is music, but the principle of ‘returning themes’ is open to a huge variety of treatment. For nearly fifty years Bach shaped fugues, concertos, arias, sonatas, chorale-preludes and choruses around this principle, whatever other habits formed around each genre. The principle of such a shape, especially where the opening and returning section is short or very short, had been familiar with Kuhnau and Buxtehude before 1700, and Bach’s Cantata No. 131 (1707?) includes something of this kind. But the organ Fugue in G minor, BWV 535a, is cast as a clearly formed ritornello and, if its dating is correct, c. 1705–7, is already producing a shape rare for organ fugues on this scale. Whether one calls this work ‘a fugue in ritornello form’ or ‘ritornello form as a fugue’ is moot: the tight fugues of previous periods are now giving way to longer-paced movements that return as naturally to the theme after a special episode as any concerto movement does.

It is likely, though not documented, that in his early days in Weimar Bach had come across works of Torelli, and therefore seen further examples of movements planned and constructed so that a distinctive theme returned after distinctive episodes. In Albinoni’s Sinfonie (1700) he could already have seen examples of fugues in a clear ritornello form somewhat comparable to the Fugue, BWV 535a. And from elsewhere in Albinoni he could have learnt the effect of bringing back at some point in a movement the whole of the opening statement, as well as marking the sections with strong cadences. In this way he would find how to design a substantial movement without deliberately imitating Vivaldi’s breathless continuity. Nevertheless, although Bach’s watchful ways of proceeding could well have gradually produced the great structures without his knowing any Italian concerto, the vividness of Vivaldi surely left a permanent mark. It can be heard in such later concertos as the A minor for Violin and the ‘Brandenburg Concerto’ No. 4.

The free way in which Italian composers would treat the returns of a theme (now longer now shorter, now this segment now that) also left its mark. Not only do Bach’s ritornello movements have various ways of re-presenting the returning theme but there are moments in his concertos, cantatas and even organ-chorales in which one particular segment of a theme returns when it could easily have been another. Extant sketches for the mature cantatas show the composer similarly considering and reconsidering such details with great care (Marshall Reference Marshall1972). Any apparent ‘arbitrariness’ in this process has the effect and the purpose of avoiding the too-obvious and of surprising the listener: the principle of return is not denied but is constantly rethought. A rather theoretical approach to Bach’s invention in recent years has emphasized those ways in which he creates a structure by bringing back themes or snatches of melody already heard; but more fundamental still, and far less of an Italian characteristic, is the control of harmonic processes, taking listeners in whatever direction his sense of logic moves. Cantata No. 199 (1714) already displays ritornello shapes one is unlikely to find in contemporary cantatas by a Graupner or Telemann, and looks like a conscious swerving away from their simpler conceptions.

A firm harmonic control can be heard in compositions written during Bach’s thirties, such as the Cello Suites, well before the biggest mature organ preludes in ritornello form. In addition, the kinds of theme, rhythm and even the reiterated chords heard in cantata or concerto movements can often be recognized as Italian in inspiration, though sustained for longer than they might be in Vivaldi’s Opp. 3 and 7 or any other concerto transcribed in Weimar. The impression Italian ritornello forms made on Bach is hinted at in such a work as Cantata No. 31 (Easter Sunday, 1715), where the opening instrumental Sonata uses a full orchestra to imitate a Vivaldian opening theme in octaves. A powerful stirring of the spirit on the day of Resurrection, though harmonically not adventurous! What follows, however – arias with recycled episodes, a fund of distinctive melody, moments of canonic imitation – would not be mistaken for Vivaldi or Albinoni or any contemporary German composer.

The year 1715 was also the year in which Albinoni’s Op. 7 Concertos were published in Amsterdam, giving yet other examples for structuring instrumental pieces. Even if none of them had circulated earlier in manuscript, in print they could have easily sped on their way to Weimar or Berlin or Dresden. By then too Bach had also developed ritornello shapes in organ-chorales, with or without having regard to Vivaldi’s concertos.

Some Weimar music: more on Italian and French tastes

Although patchy, existing evidence supports what one would expect: that the years Bach spent at Weimar in his twenties and early thirties saw developments in his chamber music, especially no doubt during his later years there. Although the sources do not confirm it, some instrumental movements known as adaptations or arrangements for church cantatas at Leipzig (Nos. 146, 156, 188, 35) or in harpsichord concertos (D minor for solo harpsichord, C major for three harpsichords) might have originated in other guises and for other forces in concerts during the later Weimar years.

Doubts sometimes expressed about the authenticity and especially the date of strikingly original works such as the ‘C major Triple Concerto’ may arise only because no one knows what the original form was, or even whether it had more than the one original form often supposed for it, e.g. as a concerto for three violins in D major. No other composer of either c. 1715 (violin version?) or c. 1735 (harpsichord version?) springs to mind as likely to have been capable of its original and characteristic amalgam of subtle counterpoint, extrovert rhythms, purposeful harmony and pleasing melody. That goes also for both of the triple concertos now known, the C major and D minor. Sources do not exist to clarify the history of such works, which is most unfortunate since only with them could one trace the composer’s maturing style, how quickly indeed it had matured since the arrival of Vivaldi’s concertos.

Slow movements with prominent solo melodies become a hallmark. Instrumental movements prefacing the Weimar Cantatas Nos. 12, 21 and 182 do not seem to me to owe very much to Italian concertos but rather result from an imaginative composer’s way of building on earlier German consort music that had still been the model for the Sonatina prefacing Cantata No. 106. Nor, yet more importantly, can one be certain what the expressive or emotional impact of such slow movements was expected to be and how much this changed over the period. The likelier that the well-known and bewitching Largo of the F minor Harpsichord Concerto, BWV 1056, imitates a Telemannesque woodwind concerto of c. 1715, the likelier that in its Leipzig version (late 1730s) it had matured to become even more expressive: now with pizzicato strings and a cantabile harpsichord melody in A flat major. What was charming, light and fresh originally and again when re-used in Cantata 156 (in F major, c. 1729), becomes more seriously beautiful, more affektvoll, probably slower, exquisite, inspiring reflection through the concerto’s key and instrumental colour.

By no means of minor interest, though utterly different, are the transcriptions for harpsichord alone of Italian and other string concertos BWV 972–87, probably made in the later Weimar years and appealing to the taste not only for imported string music but for a new world of keyboard music, one far from German traditions. Whatever music had penetrated to Thuringia beforehand, now, at a stroke, there appeared groups of imported works revealing how to shape sustained movements when there were no words to help provide shape or organization. Some composers were also surely desiring to move on from the traditional genres, such as suites, variations, fantasias, fugues, concerti grossi (qv), and would have found Italian solo string concertos to be another world. Bach’s keyboard transcriptions made at Weimar create a strikingly different repertory and present yet another peak of achievement in one very particular genre: indeed, a conspicuous group of works. They offered the Thuringian organist new melodies, new movement-shapes, new effects and textures, giving the player, then and now, a welcome breath of fresh air after the older German idioms: Vivaldi after Buxtehude.

In many little details these transcriptions of Bach anticipate details in his own later works. They include the broken chords colouring episodes (‘Fifth Brandenburg’), the emphatic chords in 2/4 time (Italian Concerto), the slow-movement cantilena (the three string concertos) and even simple contrary-motion scales (‘Goldberg Variations’): all these appear in the concerto transcriptions. In the work of other composers such as Christoph Graupner and Handel signs of Vivaldian influence can certainly be heard, but Bach’s transcriptions stand as a distinct repertory. One can still find them more effective as music for public performance than most suites and most fugues, for unlike them, their origins lay precisely in this: music for public performance. Nevertheless, how far Bach understood the natural verve and rhetoric of Venetian string concertos is an open question. Perhaps he wished to temper it with ‘German seriousness’, for it is otherwise difficult to understand why he sometimes filled in Vivaldi’s rests with bits of busy counterpoint.

In at least one case, the Concerto in A minor, BWV 593, the transcription cleverly derives such ‘bits of busy counterpoint’ from one of Vivaldi’s own themes to fill the gaps, as if Bach could not help but play with little motifs (‘melodic cells’). But Vivaldi’s gaps are perfectly effective, characteristic of the natural rhetoric of lively, dashing string concertos, as Bach himself recognized by the time he composed the first movement of the Fourth Brandenburg. The result is that Bach’s transcription BWV 593.i misses a crucial feature of the original Op. 3 No. 8 by requiring the organist to play more slowly and in a more measured way. Similarly, the two beautiful Fugues on a Theme of Albinoni, BWV 951/951a, entirely miss or choose to ignore (it is not at all clear which) the vivid string-sound of Albinoni’s original trio as published. Assuming that (i) this trio version was the one Bach knew, and (ii) BWV 951 is later than the alternative version 951a, it looks as though he was increasingly attracted to expanding and extending, playing with motifs, producing in this instance a piece rather more than three times as long as Albinoni’s original.

A key difference between Italian concerto or ritornello forms and Bach’s are that the Italian forms are often loose and capricious, deliberately asymmetrical (half a theme here, a snatch of melody there) while Bach’s are long, thorough, each section going the distance and giving an impression of being exhaustive. Bach’s own (later) harpsichord and violin concertos have long movements carefully planned around various keys, breaking off a returning theme in order to shoot off in another direction, resulting again in a characteristic thoroughness, the very opposite of the loose and capricious. The handling of so-called aria form ABA in his vocal music, also accomplished by 1714 or so, is similarly inventive: the first section (A) is not always repeated straightforwardly after the second (B) but modifies it to allow different modulations. I do not think he had to learn this from other composers.

Although in a variety of works Vivaldi’s influence can be heard, yet the effect is quite distinct and decidedly non-Italian: the spirit and even some melodic details of the F minor Concerto’s Largo already referred to come closer to one or two of the Weimar chorale-settings (such as ‘Nun komm’, BWV 659a) than they do to any Venetian concerto. Moreover, it could be that melodic resemblances people have since found between certain concertos of Telemann and Bach’s E major and F minor Harpsichord Concertos (fair-copied in the 1730s) sprang from the various revelations Vivaldi had offered to both composers equally. His styles could be worked by both German composers and result in works not dissimilar in principle.

In discussing what Bach learnt in Weimar from Venetian concertos, in particular from their brilliant ritornello movements, it is easy to forget how revelatory must have been the seductive lyricism of the better of their slow movements. Those of two concerto-transcriptions, BWV 973 (Vivaldi) and BWV 974 (Alessandro Marcello), have a simple but stunning melos that is as striking as any other characteristic of Venetian concertos. This Marcello concerto also has another of the characteristics of Bach’s own concertos, as found in the E major Violin Concerto: a startling change of direction at a certain point, like a conventional interrupted cadence but no longer conventional, a cadence specially ‘invented’. Such changes of direction come at the end of a section with which the movement could have closed, and so allow it to continue much further. Bach needed to know Italian concertos such as these to have learnt both the idea and the placing of these effects, which, unsurprisingly, he treated more exhaustively than they did. His many original interrupted cadences were there to surprise and at the same time to produce longer movements than usual in Venice.

The Venetian conception of slow movements is surely the ancestor not only of those in several of Bach’s concertos but also of a mature aria like ‘Aus Liebe’ in the St Matthew Passion, now setting words in such a way as to produce a sad, desolated Affekt. Similarly, that hard-to-define yearning quality one can sometimes sense behind the Vivaldian allegro of a concerto like BWV 975 (see Example 8) surfaces again in works like the double concertos in D minor and C minor, BWV 1043 and 1060. More strikingly still, the repeated tonic chords opening Vivaldi’s concerto transcribed as BWV 980 – an ‘extrovert’ effect the opposite of intricate counterpoint – could easily have inspired the more complex, throbbing reiterations at the beginning of ‘Brandenburg Concertos’ Nos. 3 and especially 6. And then there are Vivaldi’s ‘preparatory chromaticisms’: the sudden switch to the minor, or the sudden appearance of a chromatic chord or two, just before a final cadence. Bach continued to seize upon this kind of momentary harmonic colour in all kinds of music over the years, not least in fugues and even canons; it is extraordinarily effective in the first movement of the Third Brandenburg, just before the main theme finally returns. How indebted he was to Vivaldi for this effect deserves more study.

Example 8 Concerto (after Vivaldi), BWV 975.i, b. 23 (NB an early instance of Italian 2/4 metre in works of Bach)

Such transcriptions as BWV 980 and 975 imply two more mundane details: first, that the available instrument was only a single-manual harpsichord; secondly, that there is a stately tempo when concertos begin, whether in 4/4 or 2/4. Either Bach assumed from the first that an Italian allegro was slower, less dashing than the Italians themselves assumed, not having heard them play; or he wanted string music to be given a more moderate manner when it was transcribed for keyboard, especially organ. This steadier tempo was traditional in Germany, the ‘standard 4/4’ found in so much music of the previous generations, when only exceptionally is the crotchet fast. Today, as they constantly search for excitement, many historically informed performers approach the ‘Brandenburgs’ and solo concertos at a speed not at all clearly appropriate to 1715.

Differences between Italian and French styles and manner would have interested a lively young composer, especially at a ducal court employing good musicians. The differences are clear not only in the Weimar cantatas but also in the solo works for violin, which might in part go back to 1712 or so and where the two, the Italian and the French, are complementary. Whether or how well Bach played such violin solos himself is not known, though his violin-playing is unlikely to have been neglected at Weimar and could have developed very well at this time, something the Obituary authors are unlikely to have known (see remark on Cantata No. 182, p. 159). Other organist contemporaries writing concertos, such as C. F. Rolle, C. Pezold and J. G. Walther, have left no evidence that they ventured into virtuoso violin works of a comparable kind. But one has only to glance at the viola part in the opening movement of the St John Passion to recognize a composer with a sure sense of string-writing.

When in 1713 Bach’s pupil Kräuter mentioned the French and Italian music he expected to hear when Prince Johann Ernst returned from Holland with a trunkload of music (Dok. III, p. 650), perhaps he was studying the French as much as the Italian. In Weimar, Bach was not the only organist to be interested in organ music from France. Walther and Vogler copied music by de Grigny, Boyvin and Nivers and were probably doing this in some connection to Bach: Walther’s incomplete de Grigny copy may be based on Bach’s, while Vogler’s accurate copy of Boyvin (before 1715) may have been supervised by his teacher. These Thuringian organists certainly seem to have shared an interest in what was for them rather strange music. De Grigny might be useful for introducing them to the F3 bass clef, the many spilling-over ornaments, the unusual harmonies, new ways to write for the organ, exotic and rarely practical registrations; and copying it out was a clearly valuable exercise for Bach, whose copy is as cramped as de Grigny’s engraving, as if intentionally so. But how much of it they could find useful in church is very uncertain. Simple curiosity, a wish to learn from the many available French publications, is probably the reason for these copies.

In Halle by 1717, Gaspard Corrette’s organ-mass of 1703 was for sale from Adam Sellius, a bookseller and one of the editors of Prince Johann Ernst’s concertos, which, like the ‘Brandenburg Concertos’, were given the French title Six Concerts. Copies of French organ-books, again Boyvin but also d’Anglebert, Lebègue and Guilain, were made by some much later organists directly or indirectly associated with Bach or under his influence, such as J. F. Agricola, F. W. Marpurg and J. S. Harson (Delpech 2015). The Berlin connection of some of these musicians is a reminder that the local organs of Joachim Wagner, like those in Silbermann in Saxony, did provide many suitable frenchified sounds. It is noticeable, and probably not a coincidence, that several of the original French volumes had instructive prefaces describing such details as registration as well as ornamentation – a novelty for any German organist, not only in Weimar. And French organ-books could be interesting for more than their exotic registrations, as become clear when Marpurg extracted several fugues from them for his book of 1753.

In the dance-movements of what is probably the earliest of the so-called ‘English Suites’ for harpsichord, in A major, BWV 806a, one sees another stretching of conventions picked up from recent books of suites, again in this case French. Which precisely these books were is conjectural, and none is likely to have monopolized the composer, but plausible ones are those of Dieupart and d’Anglebert, possibly Rameau (Livre, 1705/6) and even Couperin (Premier Livre, 1713). Since it seems that little of Couperin’s music had circulated before 1713, at most a few lighter dances, then all the bigger would be the impression made on Bach by his new book, had it penetrated as far as Weimar soon afterwards. While this is not documented, these years certainly saw Bach pay tireless attention to all kinds of contemporary music, especially those that took him beyond and away from older, native kinds. And not only those years: the rest of the six ‘English Suites’ followed and are discussed below.

That Bach made efforts to complete his coverage of this or that musical genre – as if ‘ticking off a list’ of ideas – is not always obvious but is there in many an individual work, from the earliest to the last. Thus (to take one example), although it does not say so, the three sections of the organ Toccata in C major potentially demonstrate three different ways to use the second manual which he and his organ-pupils had available at Weimar: first, for echoes and alternating themes; secondly, to bring out a solo melody, like an Italian concerto; thirdly, to provide contrast in a substantial ritornello fugue. This work then becomes both a didactic demonstration and a ‘pleasing use of resources’ for the player. But no copyist, and very few players today, seem aware that the work, which is known only through copies without rubrics, can be interpreted in this way. If the title or notation had specified two manuals, it would have excluded those who had only one.

Generally in Weimar by c. 1715, there was ample opportunity for studying, composing and playing a wide musical repertory. The reason the details are unclear could be rather banal: that the scores, sacred or secular, were the duke’s property, kept in his library and forbidden to be copied or (especially when Bach became non grata) removed; later, after his departure, neglected and discarded; and then finally burnt in the fire of 1774 that destroyed the castle and library. The paper on which the performing parts were copied certainly belonged to the duke (see Dok. II, p. 56), and it is unknown to what extent the composer had his own stock of paper or had a right to the duke’s for his personal use. A likely arrangement is that for later use he kept his own full scores, vocal and instrumental. Very few of his own copies of Weimar works have survived, including the keyboard music, while isolated copies made by others, such as a rather run-of-the-mill fugue copied by Walther (BWV 1026), serve only to tantalize further. But one can suppose that there was a rich portfolio of works ‘in progress’ or (better) in a form that could be adapted for various uses, revised, rescored, retexted, re-assembled or simply compiled over as many years as the composer saw fit.

The Halle audition

At Weimar, at some point in 1713,

This seems clear enough and presumably derives from Bach himself. But it does not quite tally with his original account written in Weimar, when he denied that he had ‘applied for the post of organist’ there (Organisten Stelle angehalten: Dok. I, p. 23). The church board approached him, he visited and he would then have left had he not been invited to perform a cantata, which he did (just possibly the Advent Cantata No. 61). More needs to be said about this below. Disagreement over pay and conditions was likely, as was the matter of status: ‘organist’ to such a major church would have to be understood to mean director of music, with appropriate salary.

One oddity of the Obituary’s reference to the Halle job and its audition is that the former organist Zachow is mentioned by name when, in the Obituary’s account of the Leipzig appointment ten years later, the former organist there (Kuhnau) is not. Zachow too was a Leipziger, son of a Stadtpfeifer, and another early orphan. Was he mentioned because, though not on the Obituary’s list of composers whom his father admired, he had been named in Walther’s Lexicon in another pertinent connection: as the original teacher of Handel, who is known to have visited Halle for his niece’s baptism late in the previous year (HHB 4, p. 55)? Perhaps Handel had been the preferred if quite unlikely successor to Zachow. And did Emanuel exceptionally mention Kirchhoff at Halle because Kirchhoff’s own successor in 1746 was Wilhelm Friedemann?

There was no hurry to replace Zachow at the Market Church (Liebfrauenkirche, Church of Our Lady) of Halle, both because work had barely begun on building the new organ and because his widow was being supported for the customary six months or so. As far as the Bach biography was concerned, turning down jobs was not a thing to keep quiet about, especially since appointments to big parish churches were such public events that names of candidates became generally known. Walther’s Lexicon reported that Kirchhoff himself, who was only third choice at Halle, had declined invitations from two prince’s courts to become their capellmeister, something Kirchhoff must have told him. Telemann made a point of listing important offers he had not taken up (1740, p. 366), and Emanuel himself claimed more than once that while working for the King of Prussia he had had several advantageous offers (Burney Reference Burney1773, p. 200).

Two big questions are whether Bach did apply in the normal way at Halle and why in any case he declined. Quite possibly he had been invited as an adviser on the building of the new, large three-manual organ of sixty-five stops at the west end of this fine gothic hall-church, contracted for only a month or so after Zachow died. This is only a guess, however, but it would indicate a growing prestige, since Halle was so far the most distant place from Arnstadt to which he was invited as organ-adviser. Moreover, this was a grand instrument of uncommon size in an unusually significant church and town. So once again, the question of taking a new position as organist was raised in some connection with the building of a new organ. Bach did become one of the organ’s eventual examiners in 1716, with Kuhnau and C. F. Rolle (a Halle native), invited by the same church-council president with whom he had been in correspondence over Zachow’s job and who cannot, therefore, have taken umbrage. The Halle instrument was conceived very much on the model outlined in Werckmeister’s Orgelprobe, not unlike the Mühlhausen organ except in being bigger and more up to date, and contained in a new single but large case at the west end.

Bach’s two-or-three-week stay in Halle over late November–December 1713 seems rather long except as an opportunity to consider the vacancy and its musical potential in town and church. He then explored the job fully and formally by composing and performing, on the first or second Sunday in Advent, a cantata as requested by the chief pastor J. M. Heineccius, who had backed the organ project. Had it not been for this request, Bach later claimed, he would have gone back to Weimar earlier. That the church paid for his travel expenses, his first-class accommodation and the unidentified cantata, does seem to confirm that he was treated as an honoured guest, and supports his later claim not to have actually ‘applied’ for the job but only to have ‘presented himself’ (mich praesentiret: Dok. I, p. 23) for it. A distinction was evidently intended between the two, again rather like those people today who speak of being ‘headhunted’. (Telemann too, when asking to be released from one job to take another, claimed not to have applied for it.)

Bach’s assertion, written in a letter of 19 March 1714, was part of what looks like a tetchy reply to an accusation made by the people at Halle that in presenting himself for the job he had ‘given them such a runaround’ (solche tour gespielet) in order to solicit a better salary at Weimar. This is not unbelievable. However, considering what was clearly a contentious situation in Weimar between the two dukes, and one bound to involve the court’s musicians, it is more than likely that Bach was having thoughts about looking for suitable positions elsewhere. Perhaps there was a genuine ambivalence in his mind, and any new responsibilities he was to take on at Weimar as Concertmeister from 2 March 1714 were proposed by him partly to get above the ducal fray there, should he remain. On 19 March he signed the letter sent to Halle.

What is clear is that Bach, whether or not after having made a bid for the job, was invited by the church council or collegium, travelled across from Weimar, performed a cantata, was offered the job and was thought to have accepted it there and then (Dok. II, p. 49); he returned to Weimar in mid-December and was sent a draft contract; he delayed a month and then declined, saying he had not yet received ‘total release’ from Weimar (völlige dimission: Dok. I, p. 21); he then asked for changes to be made at Halle to the salary (an increase) and duties (a reduction); he said he would reply further and in full, when current work at Weimar allowed (surely this was stalling?); and then, probably in February 1714, finally withdrew completely.

This was a week or two before the confirmation at Weimar that he had been promoted, with increased salary, a promotion it seems that he had requested (see Dok. II, p. 53). The Halle people must have learnt of this – how, is itself an interesting question – and accused him of playing tricks. But in reply he asked why he should relocate even at the same salary let alone a lower one, and particularly before Halle had clarified what his entitlement to other fees would have been (Dok. I, pp. 23–4). This last is an important point, and yet the story is surely not quite to his credit. There is tetchiness in saying that he had no need to travel to Halle for the duke to raise his salary in Weimar, pride in claiming not actually to have applied for the Halle job, astuteness in making no mention of the Weimar promotion and speciosity in saying that his Halle correspondent (‘a learned lawyer’, ein Rechts-Verständiger) could judge whether he was right not to change jobs for the same salary.

Such points are not so very different from those in the letter he had written to Mühlhausen after leaving, so it seems, sooner than it was quite proper to do so. All the same, it is easy to believe that Bach was highly valued in Weimar, and it could also be that the frequency with which money and pay crop up in connection with him is a misleading consequence of his being represented today chiefly by formal documents and business letters. These are bound to be partial, more about business than music: for instance, there is written documentation for the cash-shares he took in one of Freiberg’s silver mines (see below, p. 612) but not about whether he ever played Freiberg’s Silbermann organs. At Halle, only certain things are recorded, and the facts that apparently he was well received for the organ-test in 1716 and had set a text by the Halle minister (possibly Cantata No. 63 for Christmas Day 1714 at Weimar) suggest that by no means did he become non grata there, any more than at Mühlhausen. The men of church and town councils he was dealing with were familiar with professional ambition, and it could be that the committee at Halle, though genuinely disappointed, knew their offer was not good enough and wanted to avoid blame. Besides, even creative musicians need to be practical.Footnote ¹

Only on 30 July 1714 was Kirchhoff appointed organist, after Halle’s second offer (to Melchior Hoffmann of the Neukirche, Leipzig) had also failed, no doubt for reasons comparable to those that had led to Bach’s refusal. Such reasons must have been common enough: Sebastian’s elder brother Christoph declined to go to Gotha to replace Pachelbel when his current pay in Ohrdruf was raised, and nor, for similar reasons, did Sebastian’s son Christoph Friedrich leave Bückeburg in 1759 for a job in Hamburg. It is clear enough that Bach was not backward in wishing to establish himself as well financially as possible, especially when in March 1714 (less than a week after the promotion was confirmed) Maria Barbara had just produced another child, her fifth: Carl Philipp Emanuel. One wonders whether, when in June 1711 the duke had already raised his salary because his predecessor no longer required a pension, Bach had actually solicited on these grounds, and if so, why no mention was made of any obligation to Effler’s widow (Dok. V, p. 116). Later, his own widow in Leipzig, Anna Magdalena, was to live on modest pensions from church, city and university (BJ 2002, pp. 245ff.), and another organist in the family, his son Johann Christoph Friedrich, was careful to negotiate a half-salary pension for his wife were he to leave her a widow (BJ 1998, p. 159).

The events in connection with Halle raise a question about Bach’s priorities. Its new organ, then in the making, was apparently not enough to draw him away from his position at a duke’s court, any more than the organ in Hamburg Jakobikirche was when he was capellmeister a few years later at another court. It is also possible, however, that the Duke of Weimar would not release him and even forbade him to say so to anybody. And there was the question of how subservient a church organist in Halle, though composer of the cantatas, would be to the cantor as well as to the church’s Ober-Pastor, who had to approve the texts chosen for the Sunday cantatas. On high and other feastdays and on every third Sunday he was to prepare a ‘moving and well-set devotional music’ (bewegliche und wohl klingend-gesetzte andächtige Musique: Dok. II, p. 50), the French word presumably meaning a cantata. As the document makes clear, this would involve the cantor, choir students (Schülern), town musicians (Stadtpfeifer) and other instrumentalists. In addition, shorter ensemble pieces were to be heard on the second and third feastdays (Feyertage), with these same performers plus the town musicians.

Although for Bach playing the organ for services and composing cantatas and motets was hardly a problem, there may have been potential difficulties with the cantor or clergy, or with the quality of singers and players, or even with the Pietist implications of that word ‘moving’. As organist, he was to play the big organ in all main services, the smaller organ and the regals as required. Regular services included the weekly Catechism and the frequent public weddings, and for the Composition of the Musique for each of these the organist was to be paid a further thaler (Dok. II, p. 51). Usually, the Catechism service (Examen) was weekly, but exactly what was involved is unclear: a motet rather than cantata, perhaps, on various days and times when instruction was given, with prayers and chorales sung with the choir. In Leipzig, the Sunday Catechism followed the final hymn of vespers (Stiller Reference Stiller1970, p. 242), probably without the cantor.

In contrast to all this, the new duties of March 1714 in the court at Weimar, complete with salary increase (see next section but one), were more promising and to be an organist even in a prominent town church like Halle’s was no longer attractive. (Handel’s putative job-search in Dresden two or three years later seems to have been for a dual appointment: capellmeister, or cantor plus opera-composer.) Bach’s Weimar colleagues would also have been a positive factor, raising his musical horizons beyond those of any town-church musicians. Later on after Weimar, to have remained a court musician at Cöthen instead of becoming organist in Hamburg (if he really did have an offer there – see below) suggests the same kind of preference for working in a court with distinguished fellow musicians. Although opportunities for church music at Cöthen were more limited than they were in Weimar, its well-run court gave a wide scope for other kinds of music – concertos, orchestral suites (ouvertures), chamber cantatas, serenatas (qv), songs, sonatas à 2 (qv), sonatas à 3 and all kinds of keyboard music. Was this what Bach most wanted at this stage of his career, and if so, was it what he continued to want?

The Orgelbüchlein: for Halle?

The Halle ‘application’ of 1713 may also imply that Bach studied beforehand the kinds of music a particular job would require. For if in the album of organ-chorales later called Orgelbüchlein the earliest pieces were copied around Advent 1713, as usually thought, it could have begun as a chorale-book for the job in Halle and not for Weimar, though added to over several months. These chorale-settings suited the church in Halle, a city then becoming a well-known centre of Pietism, which rather suggests that they originated, all or mostly, for that purpose. Music for this form of Protestantism would not be vaingloriously complicated or flashy or impersonal but aid the individual believer’s faith by directly appealing to personal feelings in reaction to the words of the hymns. Many a listener will feel this is also what certain larger-scaled works do that may have been connected directly or indirectly with Halle, as in the dramatic changes from minor to major over the long course of Cantata No. 21 which are responding to the sentiments of the text.

In the ways the Orgelbüchlein sets them, most of the hymn melodies are more immediately recognizable than in some other kinds of organ music: almost all consist of a short harmonization realized by means of subtly integrated motifs, going straight through the melody without breaks between the hymn-lines. It is as if the listener is singing a verse in an appropriate manner, slowly and thoughtfully at Advent or for Confession, joining in boisterously at suitable moments at Christmas or Easter, or simply praying for the dying. The aim is to affect the believing listener through original harmonies, resulting in new sounds conceived with a new purpose. How and when this concept originated is quite uncertain, though it is somewhat similar to the last verse in sets of hymn-variations, including those attributed to Bach. But only ‘somewhat’, for here is another example of ‘Bach the Thuringian organist’ pushing beyond local tradition in new directions.

Guesses that the first examples collected for the book (whichever these are) go back to about 1710 remain guesses, nor is it obvious what their function in Weimar would have been. By the time the title-page of the never-completed album was written nine or ten years after it was begun, quite different aims can be specified: now the young organist (potentially including Friedemann, then about twelve) is to be taught how to set chorales and learn pedal technique. So a guide to both composition and performance. Inevitably, it is from this title-page headed Orgelbüchlein that the album is now always referred to, but this is misleading on several counts. It is a pity there is no easy alternative.

What in general Halle wanted of its organist can be guessed by its various moves towards less formality in its church ceremony, such as abandoning Latin in the Lutheran Mass and requiring generally a warmer expression of emotion than had been usual. Specific requirements for the organist, had Bach signed up to them in the contract, were the following (Dok. II, p. 50, here paraphrased):

• to accompany the chorales chosen by the minister ‘slowly’ and ‘without special decoration’ (langsam ohne sonderbahres coloriren);

• to play them in four or five parts (i.e. the harmony neither too thin nor too thick);

• to draw the organ’s basic or foundation stops (Principal, Quintaden, Gedackt, reeds – i.e. avoid glitter) and to vary them constantly (i.e. avoid tedium);

• to realize the harmony with syncopations and suspensions in such a way that the congregation feels supported by the harmony.

‘Accompany’ is a more difficult word than appears. It could mean playing along with the congregation while it sang the hymn; or it might instead refer to playing over the hymn-tune, interspersing its verses with interludes, giving the people a music matching the text. If it did mean playing along with the congregation while it sang, this could be part of a Pietist approach to the hymns, a ‘communitarian’ practice not yet universal in Protestant churches. Being perhaps apprehensive of the large new organ then taking shape in the church, the authorities were being careful to insure against wilful organists playing too fast, too loud and too complicated, as very probably they otherwise would. Somebody in Halle knew the dangers.

Although the church’s clerk was rather uncertain of the technical language, his meaning is clear, and it could certainly help explain something very striking about the Orgelbüchlein: how it is that its settings are so different from the usual collections of chorales – richer, shorter, distinctive, ‘warmer’, full of Affekt, a most sophisticated harmony in three to five parts discreetly realized with syncopations and suspensions. Four-part settings, as in the two hymns for Passiontide, can have astonishingly different end-results: a ‘beautiful suffering’ in BWV 622 but a dance (polonaise?) for a doggerel-like text in BWV 623. Five-part settings also vary, from the rich pensive hymn for Advent Sunday, BWV 599, to the less weighty hymn before the sermon, BWV 633. One setting in three parts leaves the hymn-tune even clearer: the F minor of BWV 639 is surely meant to convey the anxiety of ‘I call to you, Lord Jesu Christ’. All forty-six settings encourage the listener to feel that emotions are being expressed, so much so that as with Bach’s other major compilations, the Orgelbüchlein requires a book to itself if one is to do it justice.

That the composer was enthusiastic for his new project, at least at first, can be easily sensed. Many settings could serve as accompaniments to the hymn while it was being sung: is this why most begin with the first note of the hymn-tune by itself, isolated? Because of the detail in the harmonizations, the result would generally be a slow tempo if used as accompaniment, but that may well have been the Halle style at the time, as it was elsewhere in Germany. Hence the reason, perhaps, why the composer added to the book briefly from time to time afterwards: because other organists would find it useful? When in 1746 Bach’s former pupil J. G. Ziegler applied at the same Halle church for the job that went to Friedemann Bach, he specifically said that his teacher had instructed him to play the hymns not indifferently but according to the Affekt of the words (Dok. II, p. 423).

To play according to the Affekt of the words was something the Halle church’s ministers would have desired, as Ziegler must have known from having held positions in the town for many years. His was a remark aimed specifically at pleasing the Halle authorities. Indeed, appropriate hymn-playing ‘with beautiful harmony’ had been specified elsewhere in the town on a yet earlier occasion: in 1702 when the seventeen-year-old Handel was appointed organist in its modest castle-cathedral (see HHB 4, p.18).Footnote ² To take such care seems typical of the town’s clergy, and whatever trial piece Bach had performed there he must have convinced the authorities that he was able to match what they had required of the late Zachow decades earlier: that he provide some kind of ‘pious songs of penance and thankfulness’ (Serauky Reference Serauky1939, pp. 377ff.).

The original collection of settings is doubly revealing. In aiming to provide chorales ‘for the whole church year’ it is responding to the desires of a potential employer and showing an awareness of what other composers were doing. If the first dating from Advent 1713 is correct (Dadelsen Reference Dadelsen1958, pp. 79f.) and Bach began assembling the album of 164 chorale-titles in late 1713, he did so soon after Daniel Vetter, long-established organist in the Leipzig Nikolaikirche, had published the second part of his collection of nearly two hundred chorale-settings for the church year (Musicalische Kirch- und Hauss-Ergötzlichkeit, Leipzig, 1709, 1713). No document proves that Bach was encouraged by these publications to make his own compilation, but other collections were also being made, such as the so-called Weimarer Tabulaturbuch, a book of harmonized chorales with short preludes ‘for the whole church year’, a manuscript dated 1704 and attributed to Pachelbel, though not in his hand. Although the manuscript was not present in Weimar, it seems, while Bach was there (Eggebrecht Reference Eggebrecht1965, p. 116), it is nevertheless a sign of the needs of organists at a time when the hymnbooks themselves were normally without melodies and yet when the number of organs being built across the region was gradually increasing.

In addition to the Weimarer Tabulaturbuch, there were the ninety-seven chorale-variations in the Musicalischer Vorrath (1716–19) by J. S. Beyer, cantor in Freiberg. Some years later a complete yearly cycle of chorales by J. G. Walther was announced by Mattheson (Reference Mattheson1725, p. 175), though it did not appear in print. In Vetter’s books a simple four-part harmonization is followed by a simple decorative variation; in the Weimar Tabulaturbuch, 82 of the 160 chorales are preceded on the page by a short fugue; unlike Bach’s, Vetter’s were suitable for the spinet or clavichord, as the title-pages say, and the book must have been successful, for a second edition appeared in Dresden in 1716. Bach’s collection shows no sign of any chorale having a second movement, or to have been spinet music, or to have followed any regular hymnbook. Like those other collections it begins the church year with the main chorale for Advent Sunday and after other seasonal texts moves more selectively towards a miscellany of hymns. In the way they are actually harmonized the chorales are clearly striving for a sound beyond any found in any other local collection.

It is a truism that with these settings Bach created an exceptionally touching world of sound, with a quality obvious from any comparison one might make with other composers’ work. Thus, in Example 9(a), Walther’s little semiquaver patterns though literate are commonplace, while Vetter’s setting in Example 9(b), for spinet or clavichord, is actually less like a French prélude than is Bach’s in Example 9(c). (Bach’s prélude as emblem of Advent and the start of the Church Year is matched by the ouverture written a year later for the Advent Cantata, No. 61). The manuscript shows that the opening chorale’s expressive tmesis or semiquaver break in the melody of bar 1, imitated in the following alto and tenor, was the result of second thoughts – a tiny but clear example of creative thinking, and typical of the album. No known documentation links Bach to Vetter’s publication in Leipzig, but the latter’s final chorale does appear at the close of Cantata No. 8 (1724), and in the original date, scope and aim of the Orgelbüchlein there is a good case for supposing that J. S. Bach wished to surpass Vetter in his own work. A case of Halle’s potential organist versus Leipzig’s current organist?

Example 9 Advent hymn for organ, ‘Nun komm der Heiden Heiland’: (a) J. G. Walther, verse 1 of a three-verse setting, b. 3; (b) N. Vetter, Chorale No. 1, ‘Variatio’; (c) BWV 599 (Orgelbüchlein, No. 1)

In the event, alas, Bach completed little more than a quarter of the album. New duties at Weimar in March 1714 might mean that he had no need for small-scale chorales there, even that he no longer played for the hymns in chapel. Or he was now turning more to longer, more demanding organ-settings appropriate to a ducal chapel, longer chorales evidently known to some of his students and in part assembled later in Leipzig for a new, revised collection. Had he gone to Halle, one can speculate further on how Bach’s cantatas too could have become more ‘approachable’, more immediately ‘affecting’, less ‘doctrinal’ than many of those for Leipzig. On the other hand, it could be that Halle’s form of Protestantism was not a great attraction and was even part of the reason that he remained at Weimar. Having begun the album, however, there was no reason why he would not add pieces to it from time to time, to make a more complete collection, and to do this well into the Leipzig years, by which time he might have been considering how or whether to publish it. None of this, however, affects what is likeliest: that the compilation of the album began in connection with the Halle application and stopped chiefly because he did not take the job.

In Weimar itself, practice was distinctive, not least in the court chapel. According to a report made some twenty years later, Choralsingen was led by a group of eight choirboys located in a gallery behind the altar, much nearer to the court personnel on the chapel floor than the main organ was (Jauernig Reference Jauernig, Besseler and Kraft1950, p. 71). Choralsingen could mean either Gregorian intonations or the singing of Lutheran hymns, or both: the duke’s recent decree, applying throughout the dukedom, that his revised Weimar hymnbook be used testifies to the importance he gave to congregational chorales. (The level of literacy this implies was no problem for the congregation in a court chapel.) All the more striking, then, is that neither Bach’s selection of hymns nor his order follows exactly any Weimar hymnbook. This alone might suggest the book was not for the Weimar chapel.

A practical question remains, concerning any organ-chorale, short or long, whether played at Halle by a big organ up at the west end or at Weimar by a little instrument far away in a ceiling-gallery: was there first a prelude wafting down from on high, then the choristers singing, then a congregation joining in, with organ accompanying and/or separating the verses by an interlude?

Cantatas in Weimar, 2: the promotion to concertmaster

Concertmeister was a new title and may not have had all the connotations it gradually acquired, i.e. the violinist leading the band from the first desk. The force of the Obituary’s ‘however’ is that the term, which had become more familiar by the time Emanuel was writing, would not normally imply a director of cantatas, the overall responsibility for which, and the court’s music as a whole, remained the capellmeister’s. The term Concert-Meister appears also in the ducal records and could well have been suggested by the composer himself, so that the ‘however’ implies that he had not replaced the reigning first violinist.

The official minute in the records says that the promotion to Concert-Meister was made at Bach’s ‘most humble request’ (uf sein unterth[änig]-stes Ansuchen: Dok. II, p. 53) – a conventional phrase but making one suspect that Bach had been pressing for promotion for some months, chafing at the bit to be responsible for the cantata and thus to raise his standing, still before his twenty-ninth birthday. In Weimar, the cappella was a considerable ensemble fourteen strong: capellmeister and deputy, concertmaster, six singers, a bassoonist and a quartet of string players. (This last is important for one practical detail: the bass-line was not doubled an octave below.) The minute also specified that the court musicians were to appear for rehearsals at Bach’s request, and rehearsals were to be in the chapel and not ‘in the house or [his] own apartment’ (im Hause oder eigenem Logiament). Whether ‘own’ refers to Capellmeister Drese’s or, on the contrary, to Bach’s, is uncertain: if the latter, then someone (Drese?) was hindering the new concertmaster from taking over the musicians in his own lodging, the month Carl Philipp Emanuel was born.

It is fair to assume that Bach preferred to move rehearsals to the ‘neutral’ chapel and, fearing that his authority would be questioned by the older instrumentalists, requested this in writing. A year later, by 20 March 1715, Drese father and son were officially informed that Bach was to receive a capellmeister’s salary (in fact, it was a little higher), as if the duke were directing them, willingly or unwillingly, to accept his status. Was this again at Bach’s request? Clearly, both the Weimar promotion and the Halle offer are saying much about the composer’s ambition and astuteness, as well as the genuinely high regard in which he was held by current and potential employers.

Though concertmaster, Bach continued to be court organist at least by title (Dok. II, 63) and probably harpsichordist in the chamber music. How much he played himself in chapel is vague, but his position in the world of organs was high, even unrivalled locally, both before and after the promotion. Although the court document’s phrase monatlich neüe Stücke (‘once-a-month new cantatas’: Dok. II, p. 53) is not without ambiguity, something of the kind was previously the duty of the vice-capellmeister, as it had been of his predecessor. What the vice-capellmeister’s response to this new arrangement was is unrecorded. Music in the other weeks remained under the capellmeister, including arrangements for extra instrumentalists on festive occasions, but Bach was paid extra for the occasional special court event such as wedding or memorial services.

The Obituary’s phrase ‘and perform them’ seems to be there to make it clear (as it might not otherwise have been) that Bach was to direct as well as compose these cantatas, not the capellmeister. This was something recalled with pride in later years, perhaps. Not only does the Obituary make clear that Bach was in charge but so does the court secretary’s minute, making one suspect that for the Obituary Emanuel had some original document to hand, such as a contract or copy of the court minute. That his father would preserve any such document is likely, although in that case not to include its further details – i.e. that he was ranked next below the vice-capellmeister, and that the title was at his own ‘most humble request’ – means that Emanuel omitted them.

The ‘church pieces to be performed monthly’, ensemble works of several movements, were sung after the creed and before the seasonal hymn and the sermon. Composing once a month was surely less arduous, more a pleasure even, than composing every week. Whether Bach intended or planned a complete monthly cycle of such cantatas for the church year is not clear, and there are still doubts about when exactly some of the surviving pieces were composed. Half or so of the three years’ output 1714–17 is probably lost. Fortunately, in many cases cantatas were revised later, with their instrumentation updated for the different conditions in Leipzig, and in both places they were performed to congregation-members who had the text in front of them. But in principle, the new church-cantata forms of that period must always have been more at home in court chapels than parish churches.

How far the promotion to concertmaster was a reason to decline the Halle position is another open question, since there Bach would have had similar opportunities for church cantatas, even perhaps on a bigger scale. Perhaps the standards of performance in Weimar and its aristocratic context were an attraction. One cannot be far wrong to hear in the Weimar works a wish to delight through pleasing and novel sound, and Bach must have spent time and effort in seeking to set the words expressively in the way that could be understood by his listeners. Cantatas dating from after the 1714 promotion, composed for the period from Palm Sunday to Christmas 1714 (Nos. 182, 12, 172, 21, 199, 61, 63), open mostly with an instrumental introduction before the chorus, and in this introduction especially the new concertmaster gives every appearance of aiming to please with sounds not so very different from the court’s other music, outside chapel. The length of, for example, Cantata No. 21 might itself suggest that it is a compilation made partly from earlier work, and assembled to show the composer’s range of abilities.

Typical of cantata texts over the next, most fertile twenty years, 1714–34, is the mixture of new poetry and prose extracting clauses or sentences from biblical or devotional passages and reacting to them, the whole often (normally?) rounded off both verbally and musically with a stanza of a familiar chorale from the hymnbook. This final chorale would alone distinguish the church cantata from any other kind, and where an extant score does not include a final chorale, it does not mean that there never was one, the hymn for the day perhaps, either cued in or not. (If this were so, it rather implies that the final chorales in other cantatas were more consciously selected.) Just as ‘Neumeister’ texts, unlike Italian recitatives or arias, mix biblical words and chorale (Cantatas Nos. 18, 24, 28, 59, 61), so a final verse from an existing chorale-book rounds off Salomon Franck’s texts (Cantatas Nos. 31, 72, 80, 155, 161, 163, 164, 165, 168, 185). This ‘rounding off’ rather recalls the old, familiar sets of chorale variations for keyboard whose last movement would ‘summarize’ the work with a full sound, the plenum, a final statement of the original hymn. Cantatas ever since Cantata No. 4 seem to have followed suit. But this is something that had to evolve, and it is not always clear at what stage a cantata before the Leipzig period acquired its final chorale.

Cantata recitatives and arias do develop in idiom over the years but remain remarkably uniform, or enough to make dating on such grounds most unreliable. The recitative in Cantata No. 132 (1715) is distant from any simple Italian model: the text begins with a couplet, the phrases are variously punctuated (full stop, exclamation, comma, semicolon), an arioso (qv) section includes melismas and repeated words, telling modulations mark the following sections, and so on. Despite little melodic phrases characteristic of Italian recitative, then, there is a correspondence between particular words and their musical gesture that takes the listener into a distinctive Lutheran world. The result can seldom be mistaken for the pattering, secular recitative of Italy.

The body of the colourful Weimar cantatas is the result of a constantly inventive creativity that draws on no standard formulae in setting each libretto, large scale or small scale. In the cantatas up to Easter 1715, scorings and types of setting are more varied than afterwards, and the five-part orchestra widespread in the previous period (with two violas in the Parisian manner of Lully) is reduced to four, as in chorales. In 1714 Telemann had also begun his ‘French’ cycle of cantatas, and it is possible that musical interests passed from the one composer to the other, though in which direction cannot be assumed. Peculiarly Bachian are not only the thorough French/Italian contrast between the two Advent Cantatas Nos. 61 and 62 but also the original way he deals with both sections in a French overture. In No. 61, the chorus participates in both parts of the overture, the ‘prelude’ and the ‘fugue’; in No. 194 (1723) only the fugue. Handel’s anthems and cantatas too may well have a French overture, but none is earlier than BWV 61 and in none does the choir join. Weimar choruses where they exist can be much like arias cast for a solo group of singers, whose participation spills over from the instrumental overture to the vocal cantata, from one genre to another.

That Bach-the-organist was sometimes providing cantatas already before becoming Bach-the-concertmaster is implied by a salary increase in June 1713 and by certain details in extant manuscripts. Such cantatas are Nos. 18, 199 and probably 54, chamber-like works not exclusively attached to particular Sundays but useful in other respects. In addition, there was Cantata No. 208 at Weissenfels in 1713, an expansive unicum of a cantata whose position in the composer’s development can hardly be overrated (see below). Probably the first cantata to be composed for performance under the new dispensation, No. 182 (25 March 1714, Annunciation and Palm Sunday) is knowingly and elegantly setting a new tone, as the King of Heaven is welcomed in a bright cantata that unites chamber music, chamber choruses, recitative, a sequence of arias, a chorale and finally a choral dance in the manner of a passepied. (It is possible that the first word Himmelskönig or King of Heaven was an allusion to the Himmelsburg or Castle of Heaven at the chapel’s liturgical east end; see above, p. 119.) Moreover, it has an opening sonata that leads off with a very charming melody between recorder and violin (see Example 10(a)). Is it possible that the violin solo was played by the composer, his début as violinist-composer-concertmaster?

Example 10 (a) Cantata No. 182, opening Sonata for flute (recorder) and strings; text of following movement: ‘Himmelskönig, sei willkommen’ (‘Welcome, King of Heaven!’); (b) Cantata No. 199.viii, introduction: text: ‘Wie freudig ist mein Herz’ (‘How joyful is my heart’), C major version in Kammerton (qv), viola part omitted

The last movement of Cantata No. 199, which might just pre-date the new contract, also has all the bright melody for setting words like ‘How joyful is my heart’ (Example 10(b)), where Bach is unable to resist imitative counterpoint and a masterful bass-line. Perhaps Bach also played the descant-like violin line that he added above the conventional chorale that closes another cantata, No. 185 (July 1715). Such final chorales have a chamber scoring whose effect is rather different from those at the end of the later Leipzig cantatas, where the bigger ensemble gives a more congregational feel to a hymn, whether or not the congregation actually joined in. In principle, at Weimar chorales are not unlike finales in an Italian opera in which, after the vicissitudes of plot, the soloists gather around to participate rather in the manner of a classical choros.

Two months into the new contract, Cantata No. 172 welcomes Whit Sunday with a dancing chorus and instrumental antiphony suitable for any birthday or celebratory cantata. This is a work so different from the intervening cantata, No. 21, that no one, from duke to pageboy, could have missed the musical range of which the new concertmaster was capable, and which he showed again on Christmas Day that year, with Cantata No. 63. No. 172 welcomes the Holy Ghost much as it might welcome the Elector of Saxony and does so surely to have an impact. Though in these works the parts could at times have been doubled, they are essentially chamber music, colourful, written with specific instruments in mind. In their way, and perhaps for the same reasons, the Weimar cantatas appear to be as self-dependent and distinctive as Haydn’s early symphonies at Esterhaza.

The texts by Salomon Franck are often described as too didactic and admonitory for a composer, compared with the Leipzig texts of Picander with their straightforward sentiments, and yet No. 182 is sustained by a freshness of melody and apparently effortless invention that are by no means always the case in Leipzig. As with the serenades composed at Cöthen a little later, the dance-types in which arias or choruses can be cast are not totally unlike those in keyboard suites, if not the formal allemandes or sarabandes, then minuets and, especially, dotted jigs. Which movement-types can pass from genre to genre, and which in practice do not, is a topic to be considered more broadly elsewhere.

Some eighteen cantatas for the period 1713–17 exist and another five are documented, leaving open some questions, chiefly not only where are the other twenty-plus from those years but what other musical or career avenues was he exploring? Also how do the eighteen come to survive at all? The three months’ period of mourning after the death of the young Prince Johann Ernst in August 1715 accounts for some of the ‘unproductive’ time in Bach’s tenure, though strangely, no memorial cantata for the prince survives (or has been identified) from any of the composers associated with him, Bach, Walther or Telemann. There is also a question, when a cantata has an unusual shape, whether it is known only from somebody else’s copy (such as the short No. 54) and whether it is complete. The unusually sumptuous Christmas and Easter Cantatas Nos. 63 and 31, when the duke’s cappella seems to have been strengthened, are good examples of works for which Bach must have preserved his own copies for various reasons including their usefulness in other church employment, as became the case in Leipzig.

The big festive ‘Hunt Cantata’, BWV 208, in two versions for two dukes (Weissenfels 1713?, Weimar 1716) is exceptionally well transmitted, probably not having been secreted away in the chapel library. (Bach’s score calls it Cantata rather than Serenade or Dramma, as for comparable later works.) It too raises a question about logistics: when two horn-players were brought in from Weissenfels for a later performance, someone must have arranged this, sent messages, negotiated payment, got their parts copied, arranged rehearsals, etc. How Bach had come to write a piece for the birthday celebrations of the Duke of Saxe-Weissenfels, whether in February 1712 or 1713, is not recorded, but the Weimar court must have led to such contacts, via the court secretary Salomon Franck, author of the cantata’s text. There was also a former colleague by then at the Weissenfels court, with whom Bach shared godfatherly duties in 1714 (A. I. Weldig, Emanuel’s godfather along with Telemann). Here again is probably an instance of the composer making useful contacts.

In Bach’s Weimar works, Franck naturally appears as text-author more often than Pastor Neumeister who, moving to Hamburg in 1715, plays such a big part in Telemann’s output. Cantata 208 is a path-breaking work, the first of its kind for Bach, with elements both old (types of melody) and new (recitatives, horn solos), and developing without any obvious effort the topics of hunting and the pastorella as these were familiar at the time. Here is his first known pairing of an aria with its lead-in recitativo secco, which is immediately dramatic in its isolated note Wie ? (‘how?’). The cantata is on a bigger scale than the church pieces, thoughtful but characteristically full of tunes. Each of the seven arias explores its own particular topic, and the obligatory hunting-horns in F major provide a strong central key, to which a dramatic opening is given by the soprano’s curtain-raising recitative. There is a famously memorable pastoral movement which, though still totally conventional in its gentle cantabile throbbing bass and recorders playing in thirds, is developed towards something quite new (see Example 11(a)). The pastoral topic is particularly striking after the horns have, in their own way, introduced a very different element, the bucolic (Example 11(b)). The cantata’s plan resembles the act of an opera, with a series of arias in da capo (qv) or concerto form, all freshly tuneful, with a grand total of fifteen movements, at least three of which were re-used in Leipzig church cantatas. Again, in 1740, BWV 208 was revived in honour of the Elector of Saxony’s name-day, with a minimally revised text. Evidently, it was an exceptional work regarded by Bach as suitable for two dukes and a king.

Example 11 (a) Cantata No. 208.ix, introduction; text: ‘Sheep can safely graze’; (b) Cantata No. 208.ii, introduction; text: ‘Hunting is the pleasure of the gods’

The recitative is particularly significant as part of a work not intended for a church service, showing the composer mastering the technique, possibly in response to Prince Johann Ernst’s return from Holland with Italian scores, but now of course in German. By coincidence, Handel’s first vernacular (English) recitative, in ‘As pants the hart’, HWV 251a, also belongs to 1712. Bach’s skill in writing melodious recitative is something that persists into his last decade, and like Handel’s in his (earlier) Italian cantatas, the melody and harmony go beyond those of the Italians, who might even hand recitatives over to pupil-assistants to compose. It is difficult to imagine Bach doing this.

In its very range of styles, especially the hunting motifs in the first aria, BWV 208 is a work that seems to lead naturally towards ‘Brandenburg Concerto’ No. 1. This last may be either a compilation, its first movement originally prefacing Cantata No. 208, or a full concerto from the Weimar period later revised in order to open the Margrave of Brandenburg’s set of six, fair-copied in score in 1721 (see below, p. 219). No doubt the margrave was as much a hunter as the Duke of Weissenfels was, so the hunting topic was appropriate and indeed commonplace. In going on to explore special instrumental combinations, the ‘Brandenburgs’ build on the kind of sounds already approached in the Weimar cantatas, although the leisurely paced and repetitious harmonies of the third, fourth and sixth concertos are more mature, perhaps resulting from Bach’s moving away from cantatas and their restricting preoccupation with words.

If in 1717 Bach did write no cantatas, as seems to be the case, was it pique at not automatically succeeding the deceased capellmeister Drese Snr in the previous December? Or that on Drese’s death, all current arrangements simply ceased and needed to be renegotiated? The cantatas for the later part of 1716 are maturing in melody and harmony, hardly a moment for a composer to choose to stop. (Why the Advent cantatas for 1716 have no known recitatives is unclear and may be a result of unreliable sources – or unreliable singers.) The three cantatas BWV 70a, 186a and 147a were surely meant to impress the duke with their composer’s fitness to succeed Drese. Perhaps in 1717 Bach sometimes performed pieces by other composers, even by Drese Jnr, who had been in Venice and must have learnt something about recitative and aria-form, and for all we know was a thorn in concertmaster Bach’s flesh. (Nothing is known, however, of compositions by Drese Jnr.)

But Bach did not automatically become capellmeister on Drese Snr’s death in December 1716, despite writing cantatas for that period (Advent) and hoping, one supposes, to be preferred above Drese Jnr, whose father had in turn succeeded his own father and for whom, quite possibly, Bach was not a favourite colleague. When no appointment was made, Bach’s interest in Weimar could well have waned. Other possibilities are that in the interregnum the chapel music was deteriorating, and that Bach was more or less silenced for a year by the senior duke for his loyalty to the junior. The two dukes’ quarrelling certainly exposed the distasteful side of a court run on absolutist lines, although in a major church squabbling clergy were by no means unknown. But whatever the situation, Bach would have had no trouble in turning more to chamber music were he to seek a different position elsewhere. Besides, perhaps too much is assumed today when the ‘church cantatas’ for Weimar are thought of only as church music, exclusive to services in the court chapel as if their devotional texts suited no other occasion or location, but which they surely did.

A contradiction one might sense in Bach’s twenty-plus years of writing cantatas is less easy to pin down: the sensuality of his music at moments of marked piety. This is something that one comes across in cantata arias or in organ-chorales throughout his life. It is there already in some Weimar cantatas, such as No. 54 but reaches even greater heights some time after Weimar, in the mid-1720s. The harmony is often so new and rich that one can only assume that Bach’s piety somehow embraced the God-given delight in musical sound, and went on to develop that delight. To take a Leipzig work from the mid-1720s as example: the Affekt of penitence in two arias of Cantata No. 115 is calculated and explicit, and yet the rich harmonies of the first aria and the dissonant ninths of the second are positively sensuous, beyond anything necessary for the text (see Example 12). Even if one described the latter’s contrapuntal lines and very slow, hesitant tempo as ‘evoking shame’ or ‘picturing an ingratiating penitence’ or ‘praying hesitantly’, what I find most striking is the sheer sensuality of the ninths.

Example 12 (a) Cantata No. 115.ii, ‘Ah, sluggish soul, how can you rest?’; (b) Cantata No. 115.iv, ‘Pray, pray, even there in the middle of the watch’

The Weimar Cantata No. 54, in beginning unexpectedly on a throbbing discord, might be alluding to the text’s command to ‘resist sins’, but it is the beauty of the chord itself that invites one to speculate and wonder what exactly it is alluding to. The slowly unfolding harmonies are not quite as conventional as may appear at first, nor is their implied crescendo (see Example 13). But what is the repeated discord conveying – the power of sin (it is a seductive discord), the desired resistance to sin (the discord evokes and stands for effort), the sinful resistance itself (it persists without resolution)? Why is the ‘resisting’ chord made so pleasurable? Because we are embracing the pleasure of sin, reluctant to turn from it? Or is it that a musical work having no true referent can picture resistance-to-sin with any chord, if required to do so, and a good composer finds a beautiful one in order to draw in the listener, consoling and persuading, rather than berating or ranting? Repeated chords were quite familiar to German organ-composers, but not quite like this.

Example 13 Cantata No. 54, opening (harmonies only); text: ‘Resist sin above all, otherwise its poison grips you’.

Such a rich chord and its treatment suggest a sensual side to Bach’s personality not restricted to cantatas; one hears a different side from it in those vigorous, pulsating major triads (qv) opening two of the ‘Brandenburg Concertos’. Such sensuality cannot have been exclusive to music, and one wonders how he dealt with it in its various forms, whether in his personal life he achieved equilibrium between the cerebral and the sensuous.

Pupils in Weimar

Adding to the picture of Bach’s activities at Weimar in chapel and chamber is the glimpse, given by various sources, of a dozen or so pupils. Adding also to his income: throughout his career, teaching fees could have amounted to at least his church salary at any one time and even as much as double it. Emanuel would know of pupils in Weimar probably both from his father and from later acquaintance with one or two of them, including his own relations. If ‘training’ organists implied a formal relationship, such as a regular apprenticeship between master and (live-in) pupil, the Obituary was inviting further respect for Bach’s early achievements: pupils mean that he had become an accepted and even well-known Meister, which a court position did not hinder.

Only miscellaneous chance references reveal an organist’s teaching-income. Thus for Clavier lessons that Bach gave Duke Ernst August’s page in 1711–12, payment in kind was recorded in the court books (Dok. II, p. 44); over the same period, lessons given elsewhere by Vincent Lübeck appear to have doubled or tripled his church salary (Syré Reference Syré2000, pp. 109f.); in 1729 J. G. Walther made it clear in a letter that without such fees, organists were in trouble. One can assume therefore that there were at least two further supplements to a court-employee’s income: from payments for live-in pupils’ board and lodging (as for P. D. Kräuter in Weimar); and from higher fees for teaching well-to-do amateurs.

Various manuscripts, typical of what once existed, show Bach’s keyboard works being copied by J. T. Krebs and J. C. Vogler, pupils working, in at least some instances, directly from his autographs. Probably such copying increased during the composer’s twenties and early thirties, to serve as models for the student both as composer and as organist needing repertory. There was, after all, so little in print. Walther would make several copies of the same piece, perhaps to pass on or (like a contemporary of his, the London organist John Reading) to keep at different locations. How far the Orgelbüchlein chorales were known to pupils in the later Weimar years cannot be judged from the extant copies, nor can it be judged why these were made, i.e. whether as examples of how to compose fluent harmonies or as practice music (including pedals), whether as duplicates for the teacher himself to use in services or even because the copyist simply admired them. Two of the Orgelbüchlein’s chorales were themselves recopied and greatly expanded by the composer about then or much later, showing pupils how it could be done. Perhaps there were more of these, lost along with other Weimar work.

Although Bach’s pupils are often now listed in such numbers as to imply that he had a brilliant reputation as a teacher, it is unclear how exceptional such numbers were or how unusual were the abilities of some of them. Walther mentions having fifteen to eighteen students in good times (Beckmann Reference Beckmann and Schulze1987, p. 133). Boys would normally learn organ and the arts of registration only by attaching themselves to an employed practising organist, because of formalities of the time – responsibility for the local organ was the organist’s, there were no other accessible organs, a bellows-blower had to be present, etc. There being fewer formalities for students of other instruments could explain in part why it was so common for organists to claim Bach as teacher: they needed to establish their bona fide for church appointments more than violinists did for court bands.

Several useful pieces of information about the situation in Weimar emerge from a request by one P. D. Kräuter in late 1711 to his school board at home, who were sponsoring his study with Bach, lasting for eighteen months, from April 1712 to September 1713 (Dok. V, pp. 116–22). It seems that aged twenty-six, he travelled all the way from Augsburg to Weimar for lessons – so there was nothing so very exceptional in the young Bach having gone all the way to Lübeck – and reported back regularly on progress with ‘the famous musician Bach’ (dem berühmbten Musico Bach). Whether by 1711 any of Bach’s compositions had penetrated to Augsburg is not documented, nor whether Kräuter came to Weimar because of some other connection with the court there. Either way, his testimony gives some insight into life in the earlier Weimar years, though one may have doubts about the claim that Bach was already ‘famous’: it was conventional for admirers to say so, as is clear from Handel also being called so, in the same year 1711 and thus at the same age (HHB 4, p. 47).

In speaking of ‘the famous musician’, Kräuter, like many a later student, could simply have been trying to satisfy his sponsors, as his word Musico also suggests something more than simply Organist. Kräuter’s remarks are more restrained than the fulsome praise sometimes given composers elsewhere at the time, as when the librettist’s dedication in the opera Rinaldo (1710) refers to ‘Mr Handel, whom the World so justly celebrates’ though he was barely twenty-five. It would also be odd if Kräuter did not carry some music of Bach back with him when he returned to Augsburg to take up a position in 1713. In Hamburg by 1717, Mattheson was reporting having seen both choral and keyboard works by ‘the famous organist in Weimar’ (Dok. II, p. 65), so manuscripts were circulating, if only fitfully. Still in the earlier eighteenth century, however, transmission of a composer’s music via his students or colleagues clearly played a major part.

For a year’s lessons plus board and lodging, presumably in his own house, Bach asked for 100 thaler, which Kräuter got reduced to 80 or, for the sake of his sponsors, said he did. For the teacher to charge for lodging was normal; so did Buxtehude’s predecessor Tunder in Lübeck (Edler Reference Edler1982, p. 74). As to the musical instruction itself: the pupil received six hours a day of ‘instruction’ (zur Information), in composition, keyboard (Clavier) and other instruments. After some months, as a token of his achievement, Kräuter sent his sponsors a Kirchen-Stuck in the ‘newest church style’ (Dok. V, p. 119), presumably a cantata, though quite what form this took in 1713 is unclear. Perhaps ‘six hours a day’ was also to impress his Augsburg sponsors and included such things as copying parts for his teacher, though in this instance no copies seem to have survived. He said he was free to look through his master’s work, a detail confirming the right of a fee-paying pupil to do so, and shedding some light on the moonlight anecdote (see p. 22). Students in Weimar are known to have contributed to preparing full scores of choral works with their teacher, in Kräuter’s case the motet Anh. III 159, probably by Johann Christoph of Eisenach and later listed in the Old-Bach Archive.

The picture given by Kräuter is one of serious sustained study, comparable to that of a professional college today if its curriculum were to be pared down to core studies. Lessons as we now know them but lasting six hours are unlikely. Rather, Kräuter’s day-long contact was as an apprentice to a master, observing and being useful, accompanying him in his duties, hearing or participating in music that included newly imported works, learning the details of Italian and French instrumental music, getting to know an organ’s structure (and costs) from the work underway in the Weimar chapel, and altogether ‘seeing, hearing and copying a great deal’ (vil sehen, hören und decopirt). In formally requesting a longer leave of absence from his church, as Bach himself had failed to do at Arnstadt, Kräuter mentions in particular Weimar’s musical prince, Johann Ernst, his ability on the violin, and the chance for him, Kräuter, to learn the arts of Italian concerti and French ouvertures made possible by the prince’s anticipated return from Holland. As the prince was still away, somebody must have told Kräuter what to expect. Very likely this was Bach, supporting a good student’s studies and having encouraged the prince to bring back such music. The incident gives an idea of what was discussed and planned between musicians, and what their hopes were in the musically active court of an interested duke.

At much the same period as Kräuter’s but in London, Maurice Greene was articled to the organist of St Paul’s and eventually became his successor, as Bach was succeeded in Weimar by his pupil J. M. Schubart, who was succeeded by yet another pupil (J. C. Vogler). As the Obituary makes clear, Bach had not been regularly articled himself as an apprentice to a master, though as far as their musical education was concerned, his own children might have been treated in much the same way as any live-in pupils. These could also have included a family cousin or two from time to time. Some of the seventy-odd pupils documented from the Leipzig years were regular copyists for the performing parts of their master’s cantatas and instrumental works, extracting them from his fair-copy score. One ‘family member student’ was Johann Lorenz Bach, a cousin who had studied in Weimar for some years from 1713 and felt himself qualified to be a cantor, which he subsequently became at Lahm in Itzgrund in 1718. (The organ of 1732 at Lahm remains in reliable condition.) His younger brother, Johann Elias, became tutor for younger Bach children in Leipzig 1737–42 and secretary to their father, as other pupils probably did for a period.

Johann Elias is notable today for some surviving letters in which he gives a few details of life in the Bach household twenty years or more after Weimar but still useful for a general picture, up to a point. There are pitifully few details, alas, though from them we do learn that Anna Magdalena, Sebastian’s second wife, was a keen gardener (several letters about this in BJ 2001, pp. 173–7) and that Sebastian himself had a sweet tooth, and is reported as liking cider as well as (so other references say) beer, Moselle wine, brandy, tea, coffee and tobacco. The substantial bill for such things when Bach examined an organ at Gera in 1725 represents, one must hope, consumption over several days and by several people (Dok. V, pp. 137f.).

For a salaried musician, any live-in pupils and young relations who served as some kinds of personal assistant were unlikely to adopt the elevated title ‘treasurer’ or ‘secretary’, as J. C. Schmidt Snr came to do with Handel in London. Nevertheless, some assistance was necessary, if less in Weimar than in Leipzig. Emanuel is speaking of the Leipzig years with which he was most familiar when he called the busy family house a dovecot (Taubenhaus: Dok. III, p. 290), with people coming and going all the time, his father too busy to deal with necessary correspondence though happy to talk with visitors. But the Weimar situation cannot have been so very different, except that presumably there were fewer distinguished visitors. The visit of one of them, J. G. Pisendel, later to be concertmaster at the Dresden court, was recalled by him some years after he had been in Weimar (Dok. III, p. 189), and his name occurs elsewhere in this book.

Johann Elias seems to be speaking of the Bach family house in Leipzig when he mentioned concerts taking place there (bey uns: Dok. II, p. 366), played by family and visiting musicians. Again, one can suppose life in Weimar to have been similar in this respect, with musical gatherings involving pupils, court colleagues and a precocious young Friedemann Bach. When Emanuel mentions organist-pupils specifically at Weimar, while there must have been many times more in Leipzig, he is establishing the authority Bach already had in his twenties. Only an occasional chance document from later years, such as a job application by one ex-pupil in 1726 (Dok. II, p. 157), suggests that Bach had more students in Weimar than is now known. On his being too busy for correspondence, see a suggestion on p. 329.

Equally uncertain is where and what exactly students were taught in Weimar, especially but not only when the chapel organ was out of commission. Perhaps in the Leipzig years instruction for occasional pupils, university students and, as time passed, visiting admirers was less on the organ and more in music and musical grammar generally. (It was for one university law student, L. F. Hudemann, that the nine-note canon BWV 1074 was composed, published without solution a year later in 1728 by Telemann, who thereby became in effect the ‘first Bach editor’.) One guesses that well into the Leipzig years, ‘organ lessons’ were mostly on practice instruments and covered all keyboard skills, including continuo-playing. J. P. Kirnberger, adopting the mantle of fugue-expert after Bach’s death, wrote about studying fugues in such lessons, and found himself criticized by another mantle-wearer, F. W. Marpurg. Another pupil, H. N. Gerber, spoke of learning to play fugues, along with figured-bass realization. ‘What was taught’ is discussed further below, but in Weimar it cannot have been very different from how it was in Leipzig.

Existing variants of particular works, when not an accident of surviving sources, give the impression of a composer inventively extending his first thoughts as if he were actively instructing himself – and conceivably, pupils. As well as for the early Albinoni and Bononcini/Legrenzi fugues, the versions of the mature and spectacular ‘Chromatic Fantasia’, BWV 903, demonstrate how to keep improvising, and it would be immensely useful to know whether the variants of this piece represented the composer’s actual improvisations. All three works survive in various forms, and the history of the ‘Chromatic Fantasia and Fugue’, if only one were certain about it, could reveal at a stroke the composer’s development, or at least the way copyists interpreted its ambiguities.Footnote ³ The achievement in its two such totally different but extensive movements is far beyond the free fantasia and the strict fugue of any predecessor, though in being a pair of substantial movements, BWV 903 does rather remind one of the Passacaglia. Consequently, it is hard to place any version of either movement in time and place: late in Weimar, early in Cöthen?

If the fantasia originated in response to the description of ‘free fantasy’ given in a well-circulating book of 1719 (Mattheson Reference Mattheson1719, p. 225), it would be a perfect model of how to marshal the various scales, arpeggios, hand-alternation, broken chords, harmonic tensions, modulations, changes of direction, rhetorical gaps, striking gestures, recitative and pedalpoints characteristic of free fantasias. But it probably came earlier, and in any case Bach is unlikely to have needed suggestions by Mattheson. The Fugue offers another demonstration: how to create a keyboard fugue with an idiosyncratic range of textures, corresponding in weight to the big organ fugues. As in the WTC, Musical Offering and Art of Fugue, a three-part fugue can draw from Bach the most extended and sustained structure. Whether the absence of substantial variants means that the Fugue originated after the Fantasia or, as a fugue, would by nature be less open to second thoughts, is an open question. It is not surprising that the work as a whole seems to have been more widely appreciated by keyboard students and admirers in the following decades than any concerto or cantata, with more than fifty copies of it extant or now known about.

Mattheson’s reference to fantasias raises the question how far Bach reacted to writings that appeared during his lifetime. If it was already in Weimar that he conceived the idea of compiling a complete collection of fugues (the WTC1) it could have been in response to two particular recent publications: J. K. F. Fischer’s set of twenty petite preludes and fugues (Ariadne musica, 1715), from which he appears to borrow a theme or two, and J. H. Buttstedt’s treatise comprehensively discussing fugal tehnique (Erfurt, 1716), and encouraging the idea of a grand compendium. Bach must also have been aware of current interest in tuning and in learning to compose and play in all the keys. As to specific techniques for composing: in the XX Kleine Fugen of 1731, Telemann is still advising that one can create or improvise preludes to fugues by playing with such patterns as broken chords. But this was already the case for most of the preludes in WTC1, a technique applied there with unheard-of invention.

One vestige of the way in which so many young German organists learnt to write preludes from building up broken chords and fresh harmonies can be seen in a very different piece: the opening section of Handel’s ‘Zadok the Priest’ (1727). Its atavistic allusion to German tradition was barely if at all recognized in England, although (considered objectively) it is a ‘standard’ prelude in 4/4 time, in standard tempo. An interesting question for the performer of Bach’s 4/4 preludes quoted in Example 17 (p. 239), even today, is whether there is anything in the way they are notated to imply that they have a different tempo.

The competition with Marchand

After reporting Bach’s call, not accepted, to be ‘music director and organist’ in Halle, the Obituary picks up the biography with a lengthy anecdote:

There are many implications in this anecdote, giving it an importance not immediately obvious. There may have been a religious-confessional aspect to the story: the royal family was Roman Catholic, the court increasingly so, especially later when the crown prince (Friedrich August, future dedicatee of Bach’s Kyrie and Gloria of the later ‘B minor Mass’) and his Hapsburg wife succeeded in 1733. Tension between Lutherans and Roman Catholics is clear from recorded quarrels in Dresden over burial rites and the machinations of Jesuit functionaries. It could be that the Generalfeldmarschall Flemming (in whose house the competition was to take place, according to Forkel) was looking for a Lutheran musician of repute to counter the Parisian Louis Marchand, doing so via an invitation from Concertmeister Volumier that matched the royal invitation to Marchand.

Apparently, there was ‘more honour’ in being called to a competition in Dresden than to becoming music director in Halle. The story had been briefly told in print during Bach’s lifetime (Dok. II, p. 348), hence, perhaps, the Obituary now giving so much space to it, the biggest single biographical item, even twice as long as the moonlight anecdote. Of the gaps in the extract above, one concerns Marchand as a person (he did not ‘show himself a man’), one concerns him as a composer (most of his fame rested on his Christmas Eve noëls played to Parisian crowds) and one concerns Bach’s reward being deflected. How Emanuel knew that Marchand entertained with noëls is unclear – probably through one of the younger German keyboardists (such as F. W. Marpurg) who were more familiar with French tastes than the prejudices of other German composers allowed them to be. It sounds sarcastic.

Unfortunately, Emanuel does not take the opportunity to say more about his father at this point, neither whether he was a fine interpreter of Marchand’s suites, as was said later (Dok. III, p. 125), nor even whether he really was reluctant to talk about the competition, as was said later still (Dok. III, p. 443). The last is doubtful, for the differing details in the story as variously reported look like the result of a middle-aged man’s retelling of an old anecdote about past triumphs – his very words, perhaps, in such superfluous details as ‘by special coach’ (mit Extrapost) or the ‘courteous hand-written note’. Apparently, he also spoke about it to the Erfurt organist Adlung in 1728 or thereabouts, and gave him further details (Dok. III, 122, 125). Like any good anecdote this one implies a range of things about contests of the kind that were not uncommon at the time, and especially what published biographies and their sources felt to be worth recording.

Emanuel, probably knowing that the first reference in print to his father also belonged to the year 1717, merely surmised that he was indeed ‘already so famous’ by then, though it took ‘great audacity’ to write to Marchand. Or he could assume it from an account of the Dresden competition that had been published in 1739 by J. A. Birnbaum, university teacher of rhetoric, who mentions Marchand’s significant status, Bach’s courteous letter to him and the Frenchman’s flight. Birnbaum, who was also defending Bach against some recently published criticisms (see below, pp. 533–42), had presumably learnt about it from the composer himself, goaded into publicizing his earlier success in the kingdom’s capital. For Dresden was the home of the newer musical styles preferred, apparently, by the critic against whom Birnbaum was defending him in 1739. Birnbaum’s account of the competition appears not the first time he defended Bach but the second (Dok. II, p. 348), as if the composer had meanwhile brought it to his attention: to have vanquished a French virtuoso in Dresden was a double feather in anybody’s cap, especially for a composer being criticized by 1737 for too much artifice. Again, therefore, it looks as though Emanuel was covering a story already in the public domain, particularly as there was another tale then in print about an earlier contest in Dresden, also between a ‘local’ man, Matthias Weckmann, and a distinguished, admired Catholic visitor J. J. Froberger. (This story is in Mattheson Reference Mattheson1740, p. 396.)

Well-read musicians anywhere would know from another and very different publication (Titon du Tillet’s Parnasse, 1732), that Louis Marchand was considered by some to have been the best French keyboardist of his day, more naturally gifted than either Couperin or Rameau. Pierre Du Mage, his student, already says in the preface to his Premier Livre d’orgue (1708) that his own music is modelled on la savante école et dans le goût de l’Illustre Monsieur Marchand mon Maître, and it is possible Bach knew Du Mage’s book (it too was mentioned by Birnbaum) and had long ago read its praise of Marchand. Altogether, then, for the Obituary a story of Bach conquering Marchand was by no means pointless, in fact comparable to the stories of Handel personally instructing Corelli or surpassing Domenico Scarlatti, stories recounted just a few years later in his biography.

The salient points of a story not unique in the days of competitive meetings in public between professionals seem to be these:

• Marchand had been offered a well-paid position at the Court of Saxony.

• The Dresden violinist Volumier, concerned about this, invited Bach over.

• Bach heard Marchand play, according to the Obituary (not Birnbaum), and praised his playing.

• He wrote to him (in French?), suggesting an extemporization contest.

• Marchand agreed (in writing?).

• The contest was to be in a minister’s house, with the elector’s knowledge.

• Many people (of both sexes) waited expectantly at the agreed place.

• But Marchand had left early that morning.

• The king (elector) rewarded Bach, but the 500 thalers were taken by a servant.

Why the Obituary said the listeners were ‘of both sexes’ is unclear: because Bach had reported this himself, aware that the queen (electress) was there and that she had a vested interest in Marchand’s success or failure? What is pictured is a court appearance before people of fashion, not a church committee of the usual kind. No French sources give the story, but a Dresden document, unfortunately undated, shows that Marchand did receive a payment in the year concerned (BJ 1998, 14). Among the many embroideries given later versions of the story are that Bach’s modesty allowed him to speak of it only when pressed (unlikely), that the composers did meet (also unlikely) and that Bach improvised on a theme given him by Marchand (possible – see Footnote note 4, p. 176). Other conjectures are more plausible, some of them illuminating:

If the story of the reward is true, the occasion cannot have been before October, i.e. shortly after the crown prince returned from Italy. Since J. D. Heinichen was already court capellmeister, perhaps Marchand was invited by the crown prince to direct music for the Roman Catholic chapel or for the future queen. Or certain Dresden musicians were attempting to introduce J. S. Bach as a ‘Lutheran counterweight’, although Volumier, an old acquaintance, was himself Roman Catholic, as were Buffardin the flautist, Weiss the lutenist and two-thirds of the court musicians.

The contest centred on the harpsichord. Although there is no evidence to support the idea, it is possible that (parts of) the Fifth ‘Brandenburg Concerto’ were first drafted for this occasion, with Dresden’s instrumental specialities in mind: harpsichord, virtuoso violin and (for the first time in Bach’s instrumental music) transverse flute. The slow movement seems to use a theme by Marchand,Footnote ⁴ while the theme of the first resembles others by Vivaldi and Telemann (also in D major).

Some later writers have speculated that there were to have been two recitals, one on harpsichord that did take place, one next day on organ that did not, but again, there is no clear evidence for this. On the contrary, since there was probably no fine organ in Dresden suitable for a competition until Gottfried Silbermann built his first instrument there, in the Sophienkirche in 1718–20, some public virtuoso music for harpsichord ‘in a court minister’s house’ was more appropriate. Fine harpsichords were becoming of great interest in Dresden, thanks to court patronage, and one possibility is that Bach was responding to this, even that Marchand demanded it, not being used to German organs. By saying that Bach heard Marchand ‘in secret’, Emanuel or his source is implying that Bach did not make himself known, which seems rather out of character. Perhaps he tried, unsuccessfully, and this had somehow led to Marchand being called ‘arrogant’ (hochmüthig). A former and evidently charismatic organist to the French king might well have seemed so to a Thuringian Concertmeister and his Saxon colleagues.

Four other conjectures concern possible job-searches:

• Although Bach, anxious to leave Weimar, was by then committed to Cöthen, the possibility of an incomparably better job in Dresden, where he too could ‘let himself be heard by the king’, crossed his mind.

• The largesse expended on music and theatre at Dresden was well enough known, if not in budgetary detail, for any musician to cast eyes in its direction.

• Two years later, according to the same Count Jakob Heinrich von Flemming, Handel too was in Dresden, as possible successor to the royal opera-composer Lotti (HHB 4, p. 83). Perhaps Flemming was contriving to get Handel and Bach together in the city. (The main opportunity would have been at the six-week-long celebrations in 1719 for the marriage of the crown prince. The new opera-house in the Zwinger Palace was also opened, an occasion on which Quantz and Telemann were present, the former for his first opera, the latter doubtless soliciting.) It was later said that Bach had gone that year to Halle to visit Handel, but he had left that day (Dok. III, p. 443).

• If Marchand had already decided not to take the Dresden position and return to Paris, which is possible, perhaps he ‘missed’ the competition rather than shirked it.

The anecdote says nothing about any of this, but it seems entirely characteristic of a story told by J. S. Bach that it should include something about money. When the Obituary adds the following:

who is speaking, Emanuel or his father? It could be either or both, but sounds like the latter.

Writing in the early nineteenth century, Goethe’s friend Zelter had a different interpretation of the anecdote, while frankly admitting he had no evidence for it: Marchand did come to the meeting, won the competition by playing in a more galant way than Bach had and took the king’s present while Bach went back to Weimar (Dok. VII, p. 150). Zelter, typically for his period, was questioning the nature of historical evidence and pointing out that there was no better documentary evidence for the Obituary’s story than for his. After all, it is quite possible that a spirited, vivacious, up-to-date Parisian harpsichordist had known better than J. S. Bach how to appeal to a mixed audience of non-connoisseurs. Birnbaum’s report, though apparently corroborating the later Obituary’s, was written after Marchand, Volumier, Count Flemming, the elector and the electress had all died.

How Bach and the Dresden violinist Volumier were previously acquainted in Weimar is not totally clear from documentation, and by the time of the Obituary, only the flute-virtuoso Buffardin, who had been in Dresden at the time, was still living. It is not possible to know such things as whether the flute part of the ‘Fifth Brandenburg’ had anything to do with him or whether the violin part was written for Volumier. Both are possible. But tempting though it has often proved in the past to connect some part of the Bach oeuvre with particular performers, it is, with few exceptions, too conjectural for comfort. It is also uncertain whether Bach’s works for solo violin and the later sonatas for violin and harpsichord had anything to do with Volumier or with some other violinist.

Late in the same year of 1717, a reference to Bach made in print by Mattheson includes a request for further biographical information from him, a request repeated in 1719 but to no avail on either occasion (Dok. II, pp. 65, 75). Was the first request too close to Bach’s contretemps with the Duke of Weimar that surfaced in November 1717 (see the beginning of Chapter 4) for him to want to respond? Was the second too close to Telemann’s own autobiography of 1718, which repelled him by its show of learning? Both could be the case, as again when the generous publicity Mattheson gave to J. G. Walther’s library and his plan to publish some chorales (1725, pp. 175–6) evidently did nothing to persuade Bach to seek the same kind of publicity himself. A natural reserve, as well as its opposite, could be responsible for the silence, though it seems from the preface to Mattheson’s personnel encyclopedia of 1740, the Grundlage, that Handel too failed to send a full biography.

Whatever was behind Bach’s not responding to Mattheson, there was good reason for compiling over the years, if he did, three biographical ‘archives’: a file of cuttings about himself, intended as a kind of summary curriculum vitae; eventually, a genealogical table; and a collection of his family’s compositions.